Article(id=1322949806875878170, tenantId=1146029695717560320, journalId=1249024232475115590, issueId=1322949743743213946, articleNumber=null, orderNo=null, doi=10.11834/jig.250475, pmid=null, cstr=null, oa=null, hot=null, price=null, onlineType=0, articleFormat=0, articleType=null, articleTypeStr=null, receivedDate=1759075200000, receivedDateStr=2025-09-29, revisedDate=1764864000000, revisedDateStr=2025-12-05, acceptedDate=null, acceptedDateStr=null, onlineDate=1793345413178, onlineDateStr=2026-10-30, pubDate=1784131200000, pubDateStr=2026-07-16, doiRegisterDate=null, doiRegisterDateStr=null, onlineIssueDate=1793345413178, onlineIssueDateStr=2026-10-30, onlineJustAcceptDate=null, onlineJustAcceptDateStr=null, onlineFirstDate=null, onlineFirstDateStr=null, sourceXml=null, magXml=null, createTime=1793345413178, creator=13701087609, updateTime=1793345413178, updator=13701087609, issue=Issue{id=1322949743743213946, tenantId=1146029695717560320, journalId=1249024232475115590, year='2026', volume='31', issue='7', pageStart='2279', pageEnd='2686', issueExtLink='null', onlineDate='null', pubDate='1784131200000', pubDateStr='2026-07-16', beforeIssueId=null, nextIssueId=null, price=null, status=1, issueComplete=1, articleOrder=1, issueType=-1, specialIssue=null, createTime=1793345398125, creator='13701087609', updateTime=1793348592997, updator='13701087609', preIssue=null, nextIssue=null, articleTotal=null, ext={EN=IssueExt(id=1322963144217330568, tenantId=1146029695717560320, journalId=1249024232475115590, issueId=1322949743743213946, language=EN, specialIssueTitle=, coverIllustrator=null, specialIssueEditor=, specialIssueAbout=), CN=IssueExt(id=1322963144217330569, tenantId=1146029695717560320, journalId=1249024232475115590, issueId=1322949743743213946, language=CN, specialIssueTitle=, coverIllustrator=null, specialIssueEditor=, specialIssueAbout=)}, issueFiles=null, downloadFileDto=null}, startPage=2505, endPage=2525, ext={EN=ArticleExt(id=1322949807148507931, articleId=1322949806875878170, tenantId=1146029695717560320, journalId=1249024232475115590, language=EN, title=Survey of trustworthy medical image analysis methods on the basis of evidential deep learning, columnId=1322949764341445152, journalTitle=Journal of Image and Graphics, columnName=Foundation Models for Medical Image Analysis, runingTitle=null, highlight=null, articleAbstract=

The integration of deep learning into medical diagnostics has heralded a new era of data-driven healthcare,demonstrating profound potential in tasks ranging from radiological image interpretation and pathological slide analysis to genomic sequencing and clinical outcome prediction. These advanced computational models have achieved,and in some cases even surpassed,human-level performance on various well-defined tasks,promising to enhance diagnostic accuracy,streamline clinical workflows,and facilitate personalized medicine. However,the remarkable success of these models is fundamentally predicated on an implicit assumption: the availability of large-scale,high-quality,comprehensively annotated datasets. This assumption is in stark contrast with the reality of clinical data,which are intrinsically complex and challenging. Real-world medical data are typically characterized by scarcity,heterogeneity,and imbalance. The deployment of conventional deep learning models,which often behave as black boxes in such complex data environments,poses critical risks. These models are notoriously prone to producing overconfident yet erroneous predictions and lack a mechanism to express their uncertainty or to reliably signal when they are operating outside their domain of expertise. In the high-stake context of clinical decision-making,where a single incorrect prediction can have dire effects on patient safety,this lack of trustworthy uncertainty awareness is a critical barrier to the responsible adoption of artificial intelligence (AI). This review is motivated by the urgent need to address the challenge of real-world complex data in medical AI. We systematically structure the landscape of data complexities into three principal categories,which form the organizational framework of our analysis. The first one is data scarcity,which is common in rare diseases,underrepresented populations,and emerging pathologies; it leads to overfitting and poor generalization. The second one is data heterogeneity,which includes label noise,feature noise,multimodal fusion conflicts,and out-of-distribution (OOD) samples. The third one is data imbalance,where the dominance of normal over abnormal cases biases models toward the majority class,undermining their sensitivity for rare pathologies. A new paradigm—one that goes beyond mere prediction to incorporate principled uncertainty quantification—is required to address these challenges. This review centers on evidential deep learning (EDL),an emergent and powerful framework designed for this purpose. Unlike traditional models that output a single point-estimate probability via a softmax function,EDL reframes classification by placing a high-order probability distribution over the categorical distribution of classes. The core contribution of this review is that it provides a systematic,in-depth analysis of how the EDL framework and its associated uncertainty metrics can be specifically utilized to mitigate the problems arising from real-world complex medical data. We meticulously map EDL-based solutions to our three-pronged taxonomy of data challenges. For data scarcity,EDL models naturally express high epistemic uncertainty when making predictions for classes with insufficient training evidence,thereby providing a quantitative and actionable signal that the model’s output should be treated with caution. For data heterogeneity,EDL offers robust solutions. In cases of data noise and multimodal conflicts,the conflicting or low-quality evidence is reflected in the parameters of the learned Dirichlet distribution,leading to high overall uncertainty and preventing an overconfident but incorrect conclusion. EDL is particularly effective for OOD detection; an OOD sample,by definition,lacks supporting evidence in the training data,causing a well-trained EDL model to produce a prediction with high epistemic uncertainty and a“vacuous”evidence vector. This situation enables the model to effectively“abstain”from making a high-risk prediction and instead refer the case to a human expert. In the context of data imbalance,although not a direct remedy,EDL provides a more nuanced view than standard outputs. It can reveal when a model is uncertain about its predictions for minority classes,even if the predicted probability is low,thus helping identify instances that require further scrutiny. This study,therefore,provides a comprehensive and structured survey of the current state of the art in applying EDL to medical image analysis under real-world complex data conditions. We synthesize the disparate literature into a coherent conceptual framework,elucidating the underlying principles,core methodologies,and practical applications of EDL in this critical domain. In conclusion,we posit that evidential deep learning represents a pivotal step toward building safe,reliable,and trustworthy AI systems for medical applications. However,the field is still nascent and faces several challenges that open up avenues for future research. These challenges include the need for robust theoretical foundations for evidence modeling,improving the computational scalability of EDL methods,and developing techniques for effective uncertainty calibration to ensure that the quantified uncertainty meaningfully corresponds to the true likelihood of error. The clinical translatability of uncertainty needs to be explored,with focus on how to present uncertainty information to clinicians in an intuitive and actionable manner. Moreover,large-scale,prospective clinical validation studies are imperative to move beyond retrospective dataset benchmarks and rigorously assess the real-world effect and safety of EDL-based diagnostic systems. This review aims to serve as a valuable resource and roadmap,inspiring further research to bridge the gap between the theoretical promise of EDL and its practical deployment in future clinical workflows. The codes link: https://github.com/oscarab/Medical-EDL.

, authors=Xinhao Wen1, Wei Liu1, Xiaodong Yue2, Yufei Chen1, *, authorsList=Xinhao Wen, Wei Liu, Xiaodong Yue, Yufei Chen, authorCompany=null, correspAuthors=Yufei Chen, authorNote=null, correspAuthorsNote=null, copyrightStatement=null, copyrightOwner=null, extLink=null, articleAbsUrl=null, sourceXml=null, magXml=null, pdfUrl=null, pdf=null, pdfFileSize=null, pdfExtLink=null, richHtmlUrl=null, mobilePdfUrl=null, reviewReport=null, pdfFirstPage=null, abstractGraph=null, abstractGraphContent=null, abstractVideo=null, citation=null, cebUrl=null, magXmlContent=null, mapNumber=null, fund=null), CN=ArticleExt(id=1322949813100225316, articleId=1322949806875878170, tenantId=1146029695717560320, journalId=1249024232475115590, language=CN, title=应用证据深度学习的可信医学影像分析方法综述, columnId=1322949766128218658, journalTitle=中国图象图形学报, columnName=面向医学影像智能分析的基础模型, runingTitle=null, highlight=null, articleAbstract=

随着人工智能技术在医疗领域的深度融合,基于深度学习的影像分析模型已在多项任务中展现出显著潜力。然而,这类模型的性能严重依赖于大规模、高质量的训练数据,而真实临床环境下的医学数据常面临三大类挑战:数据稀缺、数据异质性以及数据不均衡问题。传统深度学习模型在处理此类复杂数据时,往往表现出过度自信的错误预测,且缺乏对认知不确定性的表达能力,从而为高风险的临床决策带来潜在安全威胁。证据深度学习(evidential deep learning,EDL)作为一种新兴的不确定性量化范式,通过将狄利克雷分布引入到神经网络的输出对预测进行证据建模,能够一定程度上建模偶然不确定性和认知不确定性,这种能力为应对真实复杂医学数据提供了强有力的工具。本文从数据稀缺、数据异质性及数据不均衡这三大挑战的视角切入,系统性地综述了证据深度学习在复杂医学数据下的影像分析应用,并阐明了证据深度学习凭借其不确定性量化和内在的可解释性,在利用无标签数据、检测异常样本、解析并融合冲突证据、识别尾部数据以及仲裁噪声数据等方面展现的核心优势。最后,本文对该领域当前面临的挑战进行了总结,并对未来的研究方向进行了展望,为研究者们提供了一个清晰的路线图,以推动构建更安全、更可靠、更可信的下一代智能医学影像分析系统。本文提及的相关算法已汇总至https://github.com/oscarab/Medical-EDL。

, authors=文昕颢1, 刘伟1, 岳晓冬2, 陈宇飞1, *, authorsList=文昕颢, 刘伟, 岳晓冬, 陈宇飞, authorCompany=null, correspAuthors=陈宇飞, authorNote=

文昕颢,男,硕士研究生,主要研究方向为医学图像分析。E-mail:

刘伟,男,博士研究生,主要研究方向为多视角深度学习、不确定性估计和医学图像分析。E-mail:

岳晓冬,男,教授,主要研究方向为机器学习的理论与应用。E-mail:

陈宇飞,通信作者,女,副教授,主要研究方向为医学图像分析、多模态学习和不确定性估计。E-mail:

, correspAuthorsNote=
陈宇飞
, copyrightStatement=null, copyrightOwner=null, extLink=null, articleAbsUrl=null, sourceXml=WoA+8dWygb30R/yhP3K+jg==, magXml=uAheIoax3U6HAUgNYhTNnA==, pdfUrl=null, pdf=C5Akq3sIsz2ZCzd+i0Bu1w==, pdfFileSize=2887541, pdfExtLink=null, richHtmlUrl=null, mobilePdfUrl=null, reviewReport=null, pdfFirstPage=null, abstractGraph=Zorx+XI8Gc0GYDfvxvEOWA==, abstractGraphContent=null, abstractVideo=null, citation=null, cebUrl=null, magXmlContent=H6mogZAJC2aEJe+YSM3PGA==, mapNumber=null, fund=null)}, authors=[Author(id=1322949813939086124, tenantId=1146029695717560320, journalId=1249024232475115590, articleId=1322949806875878170, orderNo=0, firstName=null, middleName=null, lastName=null, nameCn=null, orcid=null, stid=null, country=null, authorPic=null, dead=0, email=wenxinhao@tongji.edu.cn, emailSecond=null, emailThird=null, correspondingAuthor=0, authorType=1, ext={EN=AuthorExt(id=1322949814018777902, tenantId=1146029695717560320, journalId=1249024232475115590, articleId=1322949806875878170, authorId=1322949813939086124, language=EN, stringName=Xinhao Wen, firstName=Xinhao, middleName=null, lastName=Wen, prefix=null, suffix=null, authorComment=null, nameInitials=null, affiliation=null, department=null, xref=1, address=1 School of Computer Science and Technology,Tongji University,Shanghai 201804,China, bio=null, bioImg=null, bioContent=null, aboutCorrespAuthor=null), CN=AuthorExt(id=1322949814081692463, tenantId=1146029695717560320, journalId=1249024232475115590, articleId=1322949806875878170, authorId=1322949813939086124, language=CN, stringName=文昕颢, firstName=null, middleName=null, lastName=null, prefix=null, suffix=null, authorComment=null, nameInitials=null, affiliation=null, department=null, xref=1, address=1 同济大学计算机科学与技术学院,上海 201804, bio={"content":"

文昕颢,男,硕士研究生,主要研究方向为医学图像分析。E-mail:

"}, bioImg=null, bioContent=

文昕颢,男,硕士研究生,主要研究方向为医学图像分析。E-mail:

, aboutCorrespAuthor=null)}, companyList=[AuthorCompany(id=1322949813762925349, tenantId=1146029695717560320, journalId=1249024232475115590, articleId=1322949806875878170, xref=1, ext=[AuthorCompanyExt(id=1322949813771313958, tenantId=1146029695717560320, journalId=1249024232475115590, articleId=1322949806875878170, companyId=1322949813762925349, language=EN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=1 School of Computer Science and Technology,Tongji University,Shanghai 201804,China), AuthorCompanyExt(id=1322949813783896871, tenantId=1146029695717560320, journalId=1249024232475115590, articleId=1322949806875878170, companyId=1322949813762925349, language=CN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=1 同济大学计算机科学与技术学院,上海 201804)])]), Author(id=1322949814144607025, tenantId=1146029695717560320, journalId=1249024232475115590, articleId=1322949806875878170, orderNo=1, firstName=null, middleName=null, lastName=null, nameCn=null, orcid=null, stid=null, country=null, authorPic=null, dead=0, email=ldachuan@outlook.com, emailSecond=null, emailThird=null, correspondingAuthor=0, authorType=1, ext={EN=AuthorExt(id=1322949814241076019, tenantId=1146029695717560320, journalId=1249024232475115590, articleId=1322949806875878170, authorId=1322949814144607025, language=EN, stringName=Wei Liu, firstName=Wei, middleName=null, lastName=Liu, prefix=null, suffix=null, authorComment=null, nameInitials=null, affiliation=null, department=null, xref=1, address=1 School of Computer Science and Technology,Tongji University,Shanghai 201804,China, bio=null, bioImg=null, bioContent=null, aboutCorrespAuthor=null), CN=AuthorExt(id=1322949814320767796, tenantId=1146029695717560320, journalId=1249024232475115590, articleId=1322949806875878170, authorId=1322949814144607025, language=CN, stringName=刘伟, firstName=null, middleName=null, lastName=null, prefix=null, suffix=null, authorComment=null, nameInitials=null, affiliation=null, department=null, xref=1, address=1 同济大学计算机科学与技术学院,上海 201804, bio={"content":"

刘伟,男,博士研究生,主要研究方向为多视角深度学习、不确定性估计和医学图像分析。E-mail:

"}, bioImg=null, bioContent=

刘伟,男,博士研究生,主要研究方向为多视角深度学习、不确定性估计和医学图像分析。E-mail:

, aboutCorrespAuthor=null)}, companyList=[AuthorCompany(id=1322949813762925349, tenantId=1146029695717560320, journalId=1249024232475115590, articleId=1322949806875878170, xref=1, ext=[AuthorCompanyExt(id=1322949813771313958, tenantId=1146029695717560320, journalId=1249024232475115590, articleId=1322949806875878170, companyId=1322949813762925349, language=EN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=1 School of Computer Science and Technology,Tongji University,Shanghai 201804,China), AuthorCompanyExt(id=1322949813783896871, tenantId=1146029695717560320, journalId=1249024232475115590, articleId=1322949806875878170, companyId=1322949813762925349, language=CN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=1 同济大学计算机科学与技术学院,上海 201804)])]), Author(id=1322949814400459574, tenantId=1146029695717560320, journalId=1249024232475115590, articleId=1322949806875878170, orderNo=2, firstName=null, middleName=null, lastName=null, nameCn=null, orcid=null, stid=null, country=null, authorPic=null, dead=0, email=yswantfly@shu.edu.cn, emailSecond=null, emailThird=null, correspondingAuthor=0, authorType=1, ext={EN=AuthorExt(id=1322949814484345656, tenantId=1146029695717560320, journalId=1249024232475115590, articleId=1322949806875878170, authorId=1322949814400459574, language=EN, stringName=Xiaodong Yue, firstName=Xiaodong, middleName=null, lastName=Yue, prefix=null, suffix=null, authorComment=null, nameInitials=null, affiliation=null, department=null, xref=2, address=2 Artificial Intelligence Institute,Shanghai University,Shanghai 200444,China, bio=null, bioImg=null, bioContent=null, aboutCorrespAuthor=null), CN=AuthorExt(id=1322949814559843129, tenantId=1146029695717560320, journalId=1249024232475115590, articleId=1322949806875878170, authorId=1322949814400459574, language=CN, stringName=岳晓冬, firstName=null, middleName=null, lastName=null, prefix=null, suffix=null, authorComment=null, nameInitials=null, affiliation=null, department=null, xref=2, address=2 上海大学人工智能研究院,上海 200444, bio={"content":"

岳晓冬,男,教授,主要研究方向为机器学习的理论与应用。E-mail:

"}, bioImg=null, bioContent=

岳晓冬,男,教授,主要研究方向为机器学习的理论与应用。E-mail:

, aboutCorrespAuthor=null)}, companyList=[AuthorCompany(id=1322949813859394344, tenantId=1146029695717560320, journalId=1249024232475115590, articleId=1322949806875878170, xref=2, ext=[AuthorCompanyExt(id=1322949813867782953, tenantId=1146029695717560320, journalId=1249024232475115590, articleId=1322949806875878170, companyId=1322949813859394344, language=EN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=2 Artificial Intelligence Institute,Shanghai University,Shanghai 200444,China), AuthorCompanyExt(id=1322949813876171562, tenantId=1146029695717560320, journalId=1249024232475115590, articleId=1322949806875878170, companyId=1322949813859394344, language=CN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=2 上海大学人工智能研究院,上海 200444)])]), Author(id=1322949814639534907, tenantId=1146029695717560320, journalId=1249024232475115590, articleId=1322949806875878170, orderNo=3, firstName=null, middleName=null, lastName=null, nameCn=null, orcid=null, stid=null, country=null, authorPic=null, dead=0, email=yufeichen@tongji.edu.cn, emailSecond=null, emailThird=null, correspondingAuthor=1, authorType=1, ext={EN=AuthorExt(id=1322949814740198205, tenantId=1146029695717560320, journalId=1249024232475115590, articleId=1322949806875878170, authorId=1322949814639534907, language=EN, stringName=Yufei Chen, firstName=Yufei, middleName=null, lastName=Chen, prefix=null, suffix=null, authorComment=null, nameInitials=null, affiliation=null, department=null, xref=1, *, address=1 School of Computer Science and Technology,Tongji University,Shanghai 201804,China, bio=null, bioImg=null, bioContent=null, aboutCorrespAuthor=null), CN=AuthorExt(id=1322949814811501374, tenantId=1146029695717560320, journalId=1249024232475115590, articleId=1322949806875878170, authorId=1322949814639534907, language=CN, stringName=陈宇飞, firstName=null, middleName=null, lastName=null, prefix=null, suffix=null, authorComment=null, nameInitials=null, affiliation=null, department=null, xref=1, *, address=1 同济大学计算机科学与技术学院,上海 201804, bio={"content":"

陈宇飞,通信作者,女,副教授,主要研究方向为医学图像分析、多模态学习和不确定性估计。E-mail:

"}, bioImg=null, bioContent=

陈宇飞,通信作者,女,副教授,主要研究方向为医学图像分析、多模态学习和不确定性估计。E-mail:

, aboutCorrespAuthor=null)}, companyList=[AuthorCompany(id=1322949813762925349, tenantId=1146029695717560320, journalId=1249024232475115590, articleId=1322949806875878170, xref=1, ext=[AuthorCompanyExt(id=1322949813771313958, tenantId=1146029695717560320, journalId=1249024232475115590, articleId=1322949806875878170, companyId=1322949813762925349, language=EN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=1 School of Computer Science and Technology,Tongji University,Shanghai 201804,China), AuthorCompanyExt(id=1322949813783896871, tenantId=1146029695717560320, journalId=1249024232475115590, articleId=1322949806875878170, companyId=1322949813762925349, language=CN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=1 同济大学计算机科学与技术学院,上海 201804)])])], keywords=[Keyword(id=1322949814912164671, tenantId=1146029695717560320, journalId=1249024232475115590, articleId=1322949806875878170, language=EN, orderNo=1, keyword=evidential deep learning(EDL)), Keyword(id=1322949814979273536, tenantId=1146029695717560320, journalId=1249024232475115590, articleId=1322949806875878170, language=EN, orderNo=2, keyword=medical image analysis), Keyword(id=1322949815050576705, tenantId=1146029695717560320, journalId=1249024232475115590, articleId=1322949806875878170, language=EN, orderNo=3, keyword=uncertainty quantification), Keyword(id=1322949815142851394, tenantId=1146029695717560320, journalId=1249024232475115590, articleId=1322949806875878170, language=EN, orderNo=4, keyword=subjective logic), Keyword(id=1322949815205765955, tenantId=1146029695717560320, journalId=1249024232475115590, articleId=1322949806875878170, language=EN, orderNo=5, keyword=evidence theory), Keyword(id=1322949815281263428, tenantId=1146029695717560320, journalId=1249024232475115590, articleId=1322949806875878170, language=CN, orderNo=1, keyword=证据深度学习(EDL)), Keyword(id=1322949815373538117, tenantId=1146029695717560320, journalId=1249024232475115590, articleId=1322949806875878170, language=CN, orderNo=2, keyword=医学影像分析), Keyword(id=1322949815444841286, tenantId=1146029695717560320, journalId=1249024232475115590, articleId=1322949806875878170, language=CN, orderNo=3, keyword=不确定性量化), Keyword(id=1322949815516144455, tenantId=1146029695717560320, journalId=1249024232475115590, articleId=1322949806875878170, language=CN, orderNo=4, keyword=主观逻辑), Keyword(id=1322949815583253320, tenantId=1146029695717560320, journalId=1249024232475115590, articleId=1322949806875878170, language=CN, orderNo=5, keyword=证据理论)], refs=[Reference(id=1322949816434697039, tenantId=1146029695717560320, journalId=1249024232475115590, articleId=1322949806875878170, doi=null, pmid=null, pmcid=null, year=2020, volume=null, issue=null, pageStart=#1251, pageEnd=null, url=null, language=null, rfNumber=null, rfOrder=0, authorNames=Amini A, Schwarting W, Soleimany A, Rus D, journalName=null, refType=null, unstructuredReference=Amini A,Schwarting W,Soleimany A and Rus D. 2020. Deep evidential regression//Proceedings of the 34th International Conference on Neural Information Processing Systems. Vancouver,Canada: Curran Associates Inc.: #1251, articleTitle=Deep evidential regression, refAbstract=null), Reference(id=1322949816510194512, tenantId=1146029695717560320, journalId=1249024232475115590, articleId=1322949806875878170, doi=null, pmid=null, pmcid=null, year=2022, volume=null, issue=null, pageStart=675, pageEnd=685, url=null, language=null, rfNumber=null, rfOrder=1, authorNames=Balaram S, Nguyen C M, Kassim A, Krishnaswamy P, journalName=null, refType=null, unstructuredReference=Balaram S,Nguyen C M,Kassim A and Krishnaswamy P. 2022. Consistency-based semi-supervised evidential active learning for diagnostic radiograph classification//Proceedings of the 25th International Conference on Medical Image Computing and Computer-Assisted Intervention. Singapore,Singapore: Springer: 675-685 [DOI: 10.1007/978-3-031-16431-6_64], articleTitle=Consistency-based semi-supervised evidential active learning for diagnostic radiograph classification, refAbstract=null), Reference(id=1322949816602469201, tenantId=1146029695717560320, journalId=1249024232475115590, articleId=1322949806875878170, doi=null, pmid=null, pmcid=null, year=2015, volume=null, issue=null, pageStart=1613, pageEnd=1622, url=null, language=null, rfNumber=null, rfOrder=2, authorNames=Blundell C, Cornebise J, Kavukcuoglu K, Wierstra D, journalName=null, refType=null, unstructuredReference=Blundell C,Cornebise J,Kavukcuoglu K and Wierstra D. 2015. Weight uncertainty in neural network//Proceedings of the 32nd International Conference on Machine Learning. Lille,France: JMLR: 1613-1622, articleTitle=Weight uncertainty in neural network, refAbstract=null), Reference(id=1322949816682160978, tenantId=1146029695717560320, journalId=1249024232475115590, articleId=1322949806875878170, doi=null, pmid=null, pmcid=null, year=2020, volume=8, issue=11, pageStart=null, pageEnd=713, url=null, language=null, rfNumber=null, rfOrder=3, authorNames=Cai L, Gao J Y, Zhao D, journalName=Annals of Translational Medicine, refType=null, unstructuredReference=Cai L,Gao J Y and Zhao D. 2020. A review of the application of deep learning in medical image classification and segmentation. Annals of Translational Medicine,8(11): #713 [DOI: 10.21037/atm.2020.02.44], articleTitle=A review of the application of deep learning in medical image classification and segmentation, refAbstract=null), Reference(id=1322949816753464147, tenantId=1146029695717560320, journalId=1249024232475115590, articleId=1322949806875878170, doi=null, pmid=null, pmcid=null, year=2025a, volume=null, issue=null, pageStart=222, pageEnd=233, url=null, language=null, rfNumber=null, rfOrder=4, authorNames=Chen D, Zhang Q, Kaplan L, Jøsang A, Jeong D, Chen F, journalName=Ethical AI for healthcare systems: uncertainty-aware,fair federated learning, refType=null, unstructuredReference=Chen D,Zhang Q,Kaplan L,Jøsang A,Jeong D,Chen F,et al. 2025a. Ethical AI for healthcare systems: uncertainty-aware,fair federated learning//Proceedings of 2025 IEEE/ACM Conference on Connected Health: Applications,Systems and Engineering Technologies (CHASE). New York,USA: IEEE: 222-233 [DOI: 10.1145/3721201.3721367], articleTitle=null, refAbstract=null), Reference(id=1322949816824767316, tenantId=1146029695717560320, journalId=1249024232475115590, articleId=1322949806875878170, doi=null, pmid=null, pmcid=null, year=2025b, volume=null, issue=null, pageStart=15767, pageEnd=15775, url=null, language=null, rfNumber=null, rfOrder=5, authorNames=Chen H S, Xu C, Guan Z Y, Zhao W, Liu J L, journalName=null, refType=null, unstructuredReference=Chen H S,Xu C,Guan Z Y,Zhao W and Liu J L. 2025b. Biased incomplete multi-view learning//Proceedings of the 39th AAAI Conference on Artificial Intelligence. Philadelphia,USA: AAAI: 15767-15775 [DOI: 10.1609/aaai.v39i15.33731], articleTitle=Biased incomplete multi-view learning, refAbstract=null), Reference(id=1322949816904459093, tenantId=1146029695717560320, journalId=1249024232475115590, articleId=1322949806875878170, doi=null, pmid=null, pmcid=null, year=2024a, volume=null, issue=null, pageStart=11439, pageEnd=11449, url=null, language=null, rfNumber=null, rfOrder=6, authorNames=Chen J Y, Ma B T, Cui H F, Xia Y, journalName=null, refType=null, unstructuredReference=Chen J Y,Ma B T,Cui H F and Xia Y. 2024a. Think twice before selection: federated evidential active learning for medical image analysis with domain shifts//Proceedings of 2024 IEEE/CVF Conference on Computer Vision and Pattern Recognition. Seattle,USA: IEEE: 11439-11449 [DOI: 10.1109/cvpr52733.2024.01087], articleTitle=Think twice before selection: federated evidential active learning for medical image analysis with domain shifts, refAbstract=null), Reference(id=1322949816971567958, tenantId=1146029695717560320, journalId=1249024232475115590, articleId=1322949806875878170, doi=null, pmid=null, pmcid=null, year=2023a, volume=45, issue=12, pageStart=15896, pageEnd=15911, url=null, language=null, rfNumber=null, rfOrder=7, authorNames=Chen M Y, Gao J Y, Xu C S, journalName=IEEE Transactions on Pattern Analysis and Machine Intelligence, refType=null, unstructuredReference=Chen M Y,Gao J Y and Xu C S. 2023a. Uncertainty-aware dual-evidential learning for weakly-supervised temporal action localization. IEEE Transactions on Pattern Analysis and Machine Intelligence,45(12): 15896-15911 [DOI: 10.1109/TPAMI.2023.3308571], articleTitle=Uncertainty-aware dual-evidential learning for weakly-supervised temporal action localization, refAbstract=null), Reference(id=1322949817038676823, tenantId=1146029695717560320, journalId=1249024232475115590, articleId=1322949806875878170, doi=null, pmid=null, pmcid=null, year=2024b, volume=null, issue=null, pageStart=3768, pageEnd=3794, url=null, language=null, rfNumber=null, rfOrder=8, authorNames=Chen M Y, Gao J Y, Xu C S, journalName=null, refType=null, unstructuredReference=Chen M Y,Gao J Y and Xu C S. 2024b. R-EDL: relaxing nonessential settings of evidential deep learning//Proceedings of the 12th International Conference on Learning Representations. Vienna,Austria: ICLR: 3768-3794, articleTitle=R-EDL: relaxing nonessential settings of evidential deep learning, refAbstract=null), Reference(id=1322949817118368600, tenantId=1146029695717560320, journalId=1249024232475115590, articleId=1322949806875878170, doi=null, pmid=null, pmcid=null, year=2023, volume=32, issue=5, pageStart=580, pageEnd=585, url=null, language=null, rfNumber=null, rfOrder=9, authorNames=Chen W, Sun K K, Li Q X, Xie K, Ni X Y, journalName=Chinese Journal of Radiological Health, refType=null, unstructuredReference=Chen W,Sun K K,Li Q X,Xie K and Ni X Y. 2023. Research progress of multimodal medical image fusion methods. Chinese Journal of Radiological Health,32(5): 580-585, articleTitle=Research progress of multimodal medical image fusion methods, refAbstract=null), Reference(id=1322949817185477465, tenantId=1146029695717560320, journalId=1249024232475115590, articleId=1322949806875878170, doi=null, pmid=null, pmcid=null, year=2023, volume=32, issue=5, pageStart=580, pageEnd=585, url=null, language=null, rfNumber=null, rfOrder=10, authorNames=陈伟, 孙康康, 李奇轩, 谢凯, 倪昕晔, journalName=中国辐射卫生, refType=null, unstructuredReference=陈伟,孙康康,李奇轩,谢凯,倪昕晔. 2023. 多模态医学图像融合方法的研究进展. 中国辐射卫生,32(5): 580-585 [DOI: 10.13491/j.issn.1004-714X.2023.05.020], articleTitle=多模态医学图像融合方法的研究进展, refAbstract=null), Reference(id=1322949817252586330, tenantId=1146029695717560320, journalId=1249024232475115590, articleId=1322949806875878170, doi=null, pmid=null, pmcid=null, year=2023b, volume=null, issue=null, pageStart=1, pageEnd=5, url=null, language=null, rfNumber=null, rfOrder=11, authorNames=Chen Y Y, Yang Z Y, Shen C Y, Wang Z W, Qin Y, Zhang Y, journalName=Proceedings of 2023 IEEE 20th International Symposium on Biomedical Imaging (ISBI), refType=null, unstructuredReference=Chen Y Y,Yang Z Y,Shen C Y,Wang Z W,Qin Y and Zhang Y. 2023b. EVIL: evidential inference learning for trustworthy semi-supervised medical image segmentation//Proceedings of 2023 IEEE 20th International Symposium on Biomedical Imaging (ISBI). Cartagena,USA: IEEE: 1-5 [DOI: 10.1109/isbi53787.2023.10230589], articleTitle=EVIL: evidential inference learning for trustworthy semi-supervised medical image segmentation, refAbstract=null), Reference(id=1322949817328083803, tenantId=1146029695717560320, journalId=1249024232475115590, articleId=1322949806875878170, doi=null, pmid=null, pmcid=null, year=2024c, volume=170, issue=null, pageStart=null, pageEnd=108004, url=null, language=null, rfNumber=null, rfOrder=12, authorNames=Chen Y Y, Yang Z Y, Shen C Y, Wang Z W, Zhang Z Z, Qin Y, journalName=Computers in Biology and Medicine, refType=null, unstructuredReference=Chen Y Y,Yang Z Y,Shen C Y,Wang Z W,Zhang Z Z,Qin Y,et al. 2024c. Evidence-based uncertainty-aware semi-supervised medical image segmentation. Computers in Biology and Medicine,170: #108004 [DOI: 10.1016/j.compbiomed.2024.108004], articleTitle=Evidence-based uncertainty-aware semi-supervised medical image segmentation, refAbstract=null), Reference(id=1322949817399386972, tenantId=1146029695717560320, journalId=1249024232475115590, articleId=1322949806875878170, doi=null, pmid=null, pmcid=null, year=2023, volume=null, issue=null, pageStart=1, pageEnd=5, url=null, language=null, rfNumber=null, rfOrder=13, authorNames=Dawood T, Chan E, Razavi R, King A P, Puyol-Antón E, journalName=Proceedings of the 20th IEEE International Symposium on Biomedical Imaging (ISBI), refType=null, unstructuredReference=Dawood T,Chan E,Razavi R,King A P and Puyol-Antón E. 2023. Addressing deep learning model calibration using evidential neural networks and uncertainty-aware training//Proceedings of the 20th IEEE International Symposium on Biomedical Imaging (ISBI). Cartagena,USA: IEEE: 1-5 [DOI: 10.1109/isbi53787.2023.10230515], articleTitle=Addressing deep learning model calibration using evidential neural networks and uncertainty-aware training, refAbstract=null), Reference(id=1322949817516827485, tenantId=1146029695717560320, journalId=1249024232475115590, articleId=1322949806875878170, doi=null, pmid=null, pmcid=null, year=2023, volume=null, issue=null, pageStart=7596, pageEnd=7616, url=null, language=null, rfNumber=null, rfOrder=14, authorNames=Deng D R, Chen G Y, Yu Y, Liu F R, Heng P A, journalName=null, refType=null, unstructuredReference=Deng D R,Chen G Y,Yu Y,Liu F R and Heng P A. 2023. Uncertainty estimation by fisher information-based evidential deep learning//Proceedings of the 40th International Conference on Machine Learning. Honolulu,USA: PMLR: 7596-7616, articleTitle=Uncertainty estimation by fisher information-based evidential deep learning, refAbstract=null), Reference(id=1322949817604907870, tenantId=1146029695717560320, journalId=1249024232475115590, articleId=1322949806875878170, doi=null, pmid=null, pmcid=null, year=2009, volume=31, issue=2, pageStart=105, pageEnd=112, url=null, language=null, rfNumber=null, rfOrder=15, authorNames=Der Kiureghian A and Ditlevsen O, journalName=Structural Safety, refType=null, unstructuredReference=Der Kiureghian A and Ditlevsen O. 2009. Aleatory or epistemic? Does it matter? Structural Safety,31(2): 105-112 [DOI: 10.1016/j.strusafe.2008.06.020], articleTitle=Aleatory or epistemic? Does it matter?, refAbstract=null), Reference(id=1322949817701376863, tenantId=1146029695717560320, journalId=1249024232475115590, articleId=1322949806875878170, doi=null, pmid=null, pmcid=null, year=2024, volume=165, issue=null, pageStart=null, pageEnd=112050, url=null, language=null, rfNumber=null, rfOrder=16, authorNames=Ding H W, Huang N N, Cui X H, journalName=Applied Soft Computing, refType=null, unstructuredReference=Ding H W,Huang N N and Cui X H. 2024. Leveraging GANs data augmentation for imbalanced medical image classification. Applied Soft Computing,165: #112050 [DOI: 10.1016/j.asoc.2024.112050], articleTitle=Leveraging GANs data augmentation for imbalanced medical image classification, refAbstract=null), Reference(id=1322949817944646496, tenantId=1146029695717560320, journalId=1249024232475115590, articleId=1322949806875878170, doi=null, pmid=null, pmcid=null, year=2024, volume=66, issue=11, pageStart=6881, pageEnd=6932, url=null, language=null, rfNumber=null, rfOrder=17, authorNames=El-Din D M, Hassanein A E, Hassanien E E, journalName=Knowledge and Information Systems, refType=null, unstructuredReference=El-Din D M,Hassanein A E and Hassanien E E. 2024. An adaptive and late multifusion framework in contextual representation based on evidential deep learning and Dempster-Shafer theory. Knowledge and Information Systems,66(11): 6881-6932 [DOI: 10.1007/s10115-024-02150-2], articleTitle=An adaptive and late multifusion framework in contextual representation based on evidential deep learning and Dempster-Shafer theory, refAbstract=null), Reference(id=1322949818036921185, tenantId=1146029695717560320, journalId=1249024232475115590, articleId=1322949806875878170, doi=null, pmid=null, pmcid=null, year=2023, volume=49, issue=4, pageStart=805, pageEnd=819, url=null, language=null, rfNumber=null, rfOrder=18, authorNames=Fang C W, Li X, Li Z Y, Jiao L C, Zhang D W, journalName=Acta Automatica Sinica, refType=null, unstructuredReference=Fang C W,Li X,Li Z Y,Jiao L C and Zhang D W. 2023. Interactive dual-model learning for semi-supervised medical image segmentation. Acta Automatica Sinica,49(4): 805-819, articleTitle=Interactive dual-model learning for semi-supervised medical image segmentation, refAbstract=null), Reference(id=1322949818112418658, tenantId=1146029695717560320, journalId=1249024232475115590, articleId=1322949806875878170, doi=null, pmid=null, pmcid=null, year=2023, volume=49, issue=4, pageStart=805, pageEnd=819, url=null, language=null, rfNumber=null, rfOrder=19, authorNames=方超伟, 李雪, 李钟毓, 焦李成, 张鼎文, journalName=自动化学报, refType=null, unstructuredReference=方超伟,李雪,李钟毓,焦李成,张鼎文. 2023. 基于双模型交互学习的半监督医学图像分割. 自动化学报,49(4): 805-819 [DOI: 10.16383/j.aas.c210667], articleTitle=基于双模型交互学习的半监督医学图像分割, refAbstract=null), Reference(id=1322949818187916131, tenantId=1146029695717560320, journalId=1249024232475115590, articleId=1322949806875878170, doi=null, pmid=null, pmcid=null, year=2024, volume=46, issue=5, pageStart=3784, pageEnd=3795, url=null, language=null, rfNumber=null, rfOrder=20, authorNames=Fidon L, Aertsen M, Kofler F, Bink A, David A L, Deprest T, journalName=IEEE Transactions on Pattern Analysis and Machine Intelligence, refType=null, unstructuredReference=Fidon L,Aertsen M,Kofler F,Bink A,David A L,Deprest T,et al. 2024. A dempster-Shafer approach to trustworthy AI with application to fetal brain MRI segmentation. IEEE Transactions on Pattern Analysis and Machine Intelligence,46(5): 3784-3795 [DOI: 10.1109/tpami.2023.3346330], articleTitle=A dempster-Shafer approach to trustworthy AI with application to fetal brain MRI segmentation, refAbstract=null), Reference(id=1322949818296968036, tenantId=1146029695717560320, journalId=1249024232475115590, articleId=1322949806875878170, doi=null, pmid=null, pmcid=null, year=2023, volume=null, issue=null, pageStart=305, pageEnd=315, url=null, language=null, rfNumber=null, rfOrder=21, authorNames=Fu W, Chen Y F, Liu W, Yue X D, Ma C, journalName=null, refType=null, unstructuredReference=Fu W,Chen Y F,Liu W,Yue X D and Ma C. 2023. Evidence reconciled neural network for out-of-distribution detection in medical images//Proceedings of the 26th International Conference on Medical Image Computing and Computer-Assisted Intervention. Vancouver,Canada: Springer: 305-315 [DOI: 10.1007/978-3-031-43898-1_30], articleTitle=Evidence reconciled neural network for out-of-distribution detection in medical images, refAbstract=null), Reference(id=1322949818401825637, tenantId=1146029695717560320, journalId=1249024232475115590, articleId=1322949806875878170, doi=null, pmid=null, pmcid=null, year=2016, volume=null, issue=null, pageStart=1050, pageEnd=1059, url=null, language=null, rfNumber=null, rfOrder=22, authorNames=Gal Y, Ghahramani Z, journalName=null, refType=null, unstructuredReference=Gal Y and Ghahramani Z. 2016. Dropout as a Bayesian approximation: representing model uncertainty in deep learning//Proceedings of the 33rd International Conference on Machine Learning. New York,USA: PMLR: 1050-1059, articleTitle=Dropout as a Bayesian approximation: representing model uncertainty in deep learning, refAbstract=null), Reference(id=1322949818468934502, tenantId=1146029695717560320, journalId=1249024232475115590, articleId=1322949806875878170, doi=null, pmid=null, pmcid=null, year=2023, volume=null, issue=null, pageStart=67, pageEnd=72, url=null, language=null, rfNumber=null, rfOrder=23, authorNames=Gan Y C, Dai Z C, Wu L C, Liu W Y, Chen L D, journalName=null, refType=null, unstructuredReference=Gan Y C,Dai Z C,Wu L C,Liu W Y and Chen L D. 2023. Deep reinforcement learning and dempster-Shafer theory: a unified approach to imbalanced classification//Proceedings of the 3rd International Conference on Computer Science,Electronic Information Engineering and Intelligent Control Technology (CEI). Wuhan,China: IEEE: 67-72 [DOI: 10.1109/cei60616.2023.10528127], articleTitle=Deep reinforcement learning and dempster-Shafer theory: a unified approach to imbalanced classification, refAbstract=null), Reference(id=1322949818557014887, tenantId=1146029695717560320, journalId=1249024232475115590, articleId=1322949806875878170, doi=null, pmid=null, pmcid=null, year=2024, volume=null, issue=null, pageStart=308, pageEnd=317, url=null, language=null, rfNumber=null, rfOrder=24, authorNames=Gao Y B, Gao Z Y, Gao X, Liu Y Y, Wang B M, Zhuang X H, journalName=null, refType=null, unstructuredReference=Gao Y B,Gao Z Y,Gao X,Liu Y Y,Wang B M and Zhuang X H. 2024. Evidential concept embedding models: towards reliable concept explanations for skin disease diagnosis//Proceedings of the 27th International Conference on Medical Image Computing and Computer-Assisted Intervention. Marrakesh,Morocco: Springer: 308-317 [DOI: 10.1007/978-3-031-72117-5_29], articleTitle=Evidential concept embedding models: towards reliable concept explanations for skin disease diagnosis, refAbstract=null), Reference(id=1322949818624123752, tenantId=1146029695717560320, journalId=1249024232475115590, articleId=1322949806875878170, doi=null, pmid=null, pmcid=null, year=2019, volume=null, issue=null, pageStart=676, pageEnd=684, url=null, language=null, rfNumber=null, rfOrder=25, authorNames=Ghesu F C, Georgescu B, Gibson E, Guendel S, Kalra M K, Singh R, journalName=null, refType=null, unstructuredReference=Ghesu F C,Georgescu B,Gibson E,Guendel S,Kalra M K,Singh R,et al. 2019. Quantifying and leveraging classification uncertainty for chest radiograph assessment//Proceedings of the 22nd International Conference on Medical Image Computing and Computer-Assisted Intervention. Shenzhen,China: Springer: 676-684 [DOI: 10.1007/978-3-030-32226-7_75], articleTitle=Quantifying and leveraging classification uncertainty for chest radiograph assessment, refAbstract=null), Reference(id=1322949818716398441, tenantId=1146029695717560320, journalId=1249024232475115590, articleId=1322949806875878170, doi=null, pmid=null, pmcid=null, year=2021, volume=null, issue=null, pageStart=36, pageEnd=42, url=null, language=null, rfNumber=null, rfOrder=26, authorNames=Ghoshal B, Swift S, Tucker A, journalName=null, refType=null, unstructuredReference=Ghoshal B,Swift S and Tucker A. 2021. Bayesian deep active learning for medical image analysis//Proceedings of the 19th International Conference on Artificial Intelligence in Medicine. [s.l.]: Springer: 36-42 [DOI: 10.1007/978-3-030-77211-6_4], articleTitle=Bayesian deep active learning for medical image analysis, refAbstract=null), Reference(id=1322949818800284522, tenantId=1146029695717560320, journalId=1249024232475115590, articleId=1322949806875878170, doi=null, pmid=null, pmcid=null, year=2023, volume=null, issue=null, pageStart=257, pageEnd=269, url=null, language=null, rfNumber=null, rfOrder=27, authorNames=Grina F, Elouedi Z, Lefevre E, journalName=null, refType=null, unstructuredReference=Grina F,Elouedi Z and Lefevre E. 2023. Evidential generative adversarial networks for handling imbalanced learning//Proceedings of the 17th European Conference on Symbolic and Quantitative Approaches with Uncertainty. Arras,France: Springer: 257-269 [DOI: 10.1007/978-3-031-45608-4_20], articleTitle=Evidential generative adversarial networks for handling imbalanced learning, refAbstract=null), Reference(id=1322949818871587691, tenantId=1146029695717560320, journalId=1249024232475115590, articleId=1322949806875878170, doi=null, pmid=null, pmcid=null, year=2024, volume=101, issue=null, pageStart=null, pageEnd=101987, url=null, language=null, rfNumber=null, rfOrder=28, authorNames=Guo Z, Wan Z L, Zhang Q S, Zhao X J, Zhang Q, Kaplan L M, journalName=Information Fusion, refType=null, unstructuredReference=Guo Z,Wan Z L,Zhang Q S,Zhao X J,Zhang Q,Kaplan L M,et al. 2024. A survey on uncertainty reasoning and quantification in belief theory and its application to deep learning. Information Fusion,101: #101987 [DOI: 10.1016/j.inffus.2023.101987], articleTitle=A survey on uncertainty reasoning and quantification in belief theory and its application to deep learning, refAbstract=null), Reference(id=1322949818947085164, tenantId=1146029695717560320, journalId=1249024232475115590, articleId=1322949806875878170, doi=null, pmid=null, pmcid=null, year=2024, volume=245, issue=null, pageStart=null, pageEnd=123052, url=null, language=null, rfNumber=null, rfOrder=29, authorNames=Han K, Sheng V S, Song Y Q, Liu Y, Qiu C J, Ma S Q, journalName=Expert Systems with Applications, refType=null, unstructuredReference=Han K,Sheng V S,Song Y Q,Liu Y,Qiu C J,Ma S Q,et al. 2024. Deep semi-supervised learning for medical image segmentation: a review. Expert Systems with Applications,245: #123052 [DOI: 10.1016/j.eswa.2023.123052], articleTitle=Deep semi-supervised learning for medical image segmentation: a review, refAbstract=null), Reference(id=1322949819030971245, tenantId=1146029695717560320, journalId=1249024232475115590, articleId=1322949806875878170, doi=null, pmid=null, pmcid=null, year=2024, volume=null, issue=null, pageStart=2010, pageEnd=2017, url=null, language=null, rfNumber=null, rfOrder=30, authorNames=He Y P, Bi Y L, Li L J, Pun C M, Jiao W P, Jin Z, journalName=null, refType=null, unstructuredReference=He Y P,Bi Y L,Li L J,Pun C M,Jiao W P and Jin Z. 2024. Mutual evidential deep learning for semi-supervised medical image segmentation//Proceedings of 2024 IEEE International Conference on Bioinformatics and Biomedicine (BIBM). Lisbon,Portugal: IEEE: 2010-2017 [DOI: 10.1109/BIBM62325.2024.10822008], articleTitle=Mutual evidential deep learning for semi-supervised medical image segmentation, refAbstract=null), Reference(id=1322949819119051630, tenantId=1146029695717560320, journalId=1249024232475115590, articleId=1322949806875878170, doi=null, pmid=null, pmcid=null, year=2025a, volume=166, issue=null, pageStart=null, pageEnd=111639, url=null, language=null, rfNumber=null, rfOrder=31, authorNames=He Y P, Li L J, Zhan T X, Pun C M, Jiao W P, Jin Z, journalName=Pattern Recognition, refType=null, unstructuredReference=He Y P,Li L J,Zhan T X,Pun C M,Jiao W P and Jin Z. 2025a. Co-evidential fusion with information volume for semi-supervised medical image segmentation. Pattern Recognition,166: #111639 [DOI: 10.1016/j.patcog.2025.111639], articleTitle=Co-evidential fusion with information volume for semi-supervised medical image segmentation, refAbstract=null), Reference(id=1322949819190354799, tenantId=1146029695717560320, journalId=1249024232475115590, articleId=1322949806875878170, doi=null, pmid=null, pmcid=null, year=2025b, volume=null, issue=null, pageStart=908, pageEnd=919, url=null, language=null, rfNumber=null, rfOrder=32, authorNames=He Y P, Li L J, Zhan T X, Pun C M, Jiao W P, Jin Z, journalName=null, refType=null, unstructuredReference=He Y P,Li L J,Zhan T X,Pun C M,Jiao W P and Jin Z. 2025b. Evidential prototype learning for semi-supervised medical image segmentation//Proceedings of the 31st ACM SIGKDD Conference on Knowledge Discovery and Data Mining V.2. Toronto,Canada: Association for Computing Machinery: 908-919 [DOI: 10.1145/3711896.3736944], articleTitle=Evidential prototype learning for semi-supervised medical image segmentation, refAbstract=null), Reference(id=1322949819270046576, tenantId=1146029695717560320, journalId=1249024232475115590, articleId=1322949806875878170, doi=null, pmid=null, pmcid=null, year=2020, volume=null, issue=null, pageStart=865, pageEnd=870, url=null, language=null, rfNumber=null, rfOrder=33, authorNames=Hemmer P, Kühl N, Schöffer J, journalName=null, refType=null, unstructuredReference=Hemmer P,Kühl N and Schöffer J. 2020. DEAL: deep evidential active learning for image classification//Proceedings of the 19th IEEE International Conference on Machine Learning and Applications (ICMLA). Miami,USA: IEEE: 865-870 [DOI: 10.1109/icmla51294.2020.00141], articleTitle=DEAL: deep evidential active learning for image classification, refAbstract=null), Reference(id=1322949819358126961, tenantId=1146029695717560320, journalId=1249024232475115590, articleId=1322949806875878170, doi=null, pmid=null, pmcid=null, year=2024, volume=113, issue=9, pageStart=6453, pageEnd=6474, url=null, language=null, rfNumber=null, rfOrder=34, authorNames=Hoarau A, Lemaire V, Le Gall Y, Dubois J C, Martin A, journalName=Machine Learning, refType=null, unstructuredReference=Hoarau A,Lemaire V,Le Gall Y,Dubois J C and Martin A. 2024. Evidential uncertainty sampling strategies for active learning. Machine Learning,113(9): 6453-6474 [DOI: 10.1007/s10994-024-06567-2], articleTitle=Evidential uncertainty sampling strategies for active learning, refAbstract=null), Reference(id=1322949819437818738, tenantId=1146029695717560320, journalId=1249024232475115590, articleId=1322949806875878170, doi=null, pmid=null, pmcid=null, year=2020, volume=2020, issue=1, pageStart=null, pageEnd=8279342, url=null, language=null, rfNumber=null, rfOrder=35, authorNames=Huang B, Yang F, Yin M X, Mo X Y, Zhong C, journalName=Computational and Mathematical Methods in Medicine, refType=null, unstructuredReference=Huang B,Yang F,Yin M X,Mo X Y and Zhong C. 2020. A review of multimodal medical image fusion techniques. Computational and Mathematical Methods in Medicine,2020(1): #8279342 [DOI: 10.1155/2020/8279342], articleTitle=A review of multimodal medical image fusion techniques, refAbstract=null), Reference(id=1322949819513316211, tenantId=1146029695717560320, journalId=1249024232475115590, articleId=1322949806875878170, doi=null, pmid=null, pmcid=null, year=2021a, volume=null, issue=null, pageStart=30, pageEnd=39, url=null, language=null, rfNumber=null, rfOrder=36, authorNames=Huang L, Denøeux T, Tonnelet D, Decazes P, Ruan S, journalName=Proceedings of the 12th International Workshop on Machine Learning in Medical Imaging, refType=null, unstructuredReference=Huang L,Denøeux T,Tonnelet D,Decazes P and Ruan S. 2021a. Deep PET/CT fusion with Dempster-Shafer theory for lymphoma segmentation//Proceedings of the 12th International Workshop on Machine Learning in Medical Imaging. Strasbourg,France: Springer: 30-39 [DOI: 10.1007/978-3-030-87589-3_4], articleTitle=Deep PET/CT fusion with Dempster-Shafer theory for lymphoma segmentation, refAbstract=null), Reference(id=1322949819593007988, tenantId=1146029695717560320, journalId=1249024232475115590, articleId=1322949806875878170, doi=null, pmid=null, pmcid=null, year=2025, volume=113, issue=null, pageStart=null, pageEnd=102648, url=null, language=null, rfNumber=null, rfOrder=37, authorNames=Huang L, Ruan S, Decazes P, Denœux T, journalName=Information Fusion, refType=null, unstructuredReference=Huang L,Ruan S,Decazes P and Denœux T. 2025. Deep evidential fusion with uncertainty quantification and reliability learning for multimodal medical image segmentation. Information Fusion,113: #102648 [DOI: 10.1016/j.inffus.2024.102648], articleTitle=Deep evidential fusion with uncertainty quantification and reliability learning for multimodal medical image segmentation, refAbstract=null), Reference(id=1322949819664311157, tenantId=1146029695717560320, journalId=1249024232475115590, articleId=1322949806875878170, doi=null, pmid=null, pmcid=null, year=2021b, volume=null, issue=null, pageStart=#3, pageEnd=null, url=null, language=null, rfNumber=null, rfOrder=38, authorNames=Huang L, Ruan S, Denœux T, journalName=null, refType=null, unstructuredReference=Huang L,Ruan S and Denœux T. 2021b. Covid-19 classification with deep neural network and belief functions//Proceedings of the 5th International Conference on Biological Information and Biomedical Engineering. Hangzhou,China: Association for Computing Machinery: #3 [DOI: 10.1145/3469678.3469719], articleTitle=Covid-19 classification with deep neural network and belief functions, refAbstract=null), Reference(id=1322949819727225718, tenantId=1146029695717560320, journalId=1249024232475115590, articleId=1322949806875878170, doi=null, pmid=null, pmcid=null, year=2023, volume=535, issue=null, pageStart=40, pageEnd=52, url=null, language=null, rfNumber=null, rfOrder=39, authorNames=Huang L, Ruan S, Denœux T, journalName=Neurocomputing, refType=null, unstructuredReference=Huang L,Ruan S and Denœux T. 2023. Semi-supervised multiple evidence fusion for brain tumor segmentation. Neurocomputing,535: 40-52 [DOI: 10.1016/j.neucom.2023.02.047], articleTitle=Semi-supervised multiple evidence fusion for brain tumor segmentation, refAbstract=null), Reference(id=1322949819790140279, tenantId=1146029695717560320, journalId=1249024232475115590, articleId=1322949806875878170, doi=null, pmid=null, pmcid=null, year=2024, volume=null, issue=null, pageStart=3319, pageEnd=3322, url=null, language=null, rfNumber=null, rfOrder=40, authorNames=Huang Q G, Chen Y F, Liu Y Q, Yue X D, Ma C, journalName=null, refType=null, unstructuredReference=Huang Q G,Chen Y F,Liu Y Q,Yue X D and Ma C. 2024. Trusted multi-view pancreatic tumor subtypes diagnosis through augmented opinion//Proceedings of 2024 IEEE International Conference on Bioinformatics and Biomedicine (BIBM). Lisbon,Portugal: IEEE: 3319-3322 [DOI: 10.1109/bibm62325.2024.10822127], articleTitle=Trusted multi-view pancreatic tumor subtypes diagnosis through augmented opinion, refAbstract=null), Reference(id=1322949819857249144, tenantId=1146029695717560320, journalId=1249024232475115590, articleId=1322949806875878170, doi=null, pmid=null, pmcid=null, year=2024, volume=169, issue=null, pageStart=null, pageEnd=107840, url=null, language=null, rfNumber=null, rfOrder=41, authorNames=Jiao R S, Zhang Y C, Ding L, Xue B S, Zhang J C, Cai R, journalName=Computers in Biology and Medicine, refType=null, unstructuredReference=Jiao R S,Zhang Y C,Ding L,Xue B S,Zhang J C,Cai R,et al. 2024. Learning with limited annotations: a survey on deep semi-supervised learning for medical image segmentation. Computers in Biology and Medicine,169: #107840 [DOI: 10.1016/j.compbiomed.2023.107840], articleTitle=Learning with limited annotations: a survey on deep semi-supervised learning for medical image segmentation, refAbstract=null), Reference(id=1322949819915969401, tenantId=1146029695717560320, journalId=1249024232475115590, articleId=1322949806875878170, doi=null, pmid=null, pmcid=null, year=2016, volume=null, issue=null, pageStart=null, pageEnd=null, url=null, language=null, rfNumber=null, rfOrder=42, authorNames=Jøsang A, journalName=null, refType=null, unstructuredReference=Jøsang A. 2016. Subjective Logic. Cham,Germany: Springer [DOI: 10.1007/978-3-319-42337-1], articleTitle=Subjective Logic. Cham,Germany: Springer, refAbstract=null), Reference(id=1322949819991466874, tenantId=1146029695717560320, journalId=1249024232475115590, articleId=1322949806875878170, doi=null, pmid=null, pmcid=null, year=2018, volume=null, issue=null, pageStart=1998, pageEnd=2005, url=null, language=null, rfNumber=null, rfOrder=43, authorNames=Jøsang A, Cho J H, Chen F, journalName=null, refType=null, unstructuredReference=Jøsang A,Cho J H and Chen F. 2018. Uncertainty characteristics of subjective opinions//Proceedings of the 21st International Conference on Information Fusion (FUSION). Cambridge,UK: IEEE: 1998-2005 [DOI: 10.23919/icif.2018.8455454], articleTitle=Uncertainty characteristics of subjective opinions, refAbstract=null), Reference(id=1322949820058575739, tenantId=1146029695717560320, journalId=1249024232475115590, articleId=1322949806875878170, doi=null, pmid=null, pmcid=null, year=2023, volume=4, issue=2, pageStart=383, pageEnd=397, url=null, language=null, rfNumber=null, rfOrder=44, authorNames=Karimi D, Gholipour A, journalName=IEEE Transactions on Artificial Intelligence, refType=null, unstructuredReference=Karimi D and Gholipour A. 2023. Improving calibration and out-of-distribution detection in deep models for medical image segmentation. IEEE Transactions on Artificial Intelligence,4(2): 383-397 [DOI: 10.1109/TAI.2022.3159510], articleTitle=Improving calibration and out-of-distribution detection in deep models for medical image segmentation, refAbstract=null), Reference(id=1322949820125684604, tenantId=1146029695717560320, journalId=1249024232475115590, articleId=1322949806875878170, doi=null, pmid=null, pmcid=null, year=2017, volume=null, issue=null, pageStart=5580, pageEnd=5590, url=null, language=null, rfNumber=null, rfOrder=45, authorNames=Kendall A, Gal Y, journalName=null, refType=null, unstructuredReference=Kendall A and Gal Y. 2017. What uncertainties do we need in Bayesian deep learning for computer vision?//Proceedings of the 31st International Conference on Neural Information Processing Systems. Long Beach,USA: Curran Associates Inc.: 5580-5590, articleTitle=What uncertainties do we need in Bayesian deep learning for computer vision?, refAbstract=null), Reference(id=1322949820196987773, tenantId=1146029695717560320, journalId=1249024232475115590, articleId=1322949806875878170, doi=null, pmid=null, pmcid=null, year=2022, volume=null, issue=null, pageStart=2022, pageEnd=05, url=null, language=null, rfNumber=null, rfOrder=46, authorNames=Khosravi B, Faghani S, Ashraf-Ganjouei A, journalName=medRxiv, refType=null, unstructuredReference=Khosravi B,Faghani S and Ashraf-Ganjouei A. 2022. Uncertainty quantification in COVID-19 detection using evidential deep learning. medRxiv: 2022-05 [DOI: 10.1101/2022.05.29.22275732], articleTitle=Uncertainty quantification in COVID-19 detection using evidential deep learning, refAbstract=null), Reference(id=1322949820259902334, tenantId=1146029695717560320, journalId=1249024232475115590, articleId=1322949806875878170, doi=null, pmid=null, pmcid=null, year=2024, volume=312, issue=3, pageStart=null, pageEnd=null, url=null, language=null, rfNumber=null, rfOrder=47, authorNames=Koetzier L R, Wu J, Mastrodicasa D, Lutz A, Chung M, Koszek W A, journalName=Radiology, refType=null, unstructuredReference=Koetzier L R,Wu J,Mastrodicasa D,Lutz A,Chung M,Koszek W A,et al. 2024. Generating synthetic data for medical imaging. Radiology,312(3): #e232471 [DOI: 10.1148/radiol.232471], articleTitle=Generating synthetic data for medical imaging, refAbstract=null), Reference(id=1322949820356371327, tenantId=1146029695717560320, journalId=1249024232475115590, articleId=1322949806875878170, doi=null, pmid=null, pmcid=null, year=2017, volume=null, issue=null, pageStart=6402, pageEnd=6413, url=null, language=null, rfNumber=null, rfOrder=48, authorNames=Lakshminarayanan B, Pritzel A, Blundell C, journalName=null, refType=null, unstructuredReference=Lakshminarayanan B,Pritzel A and Blundell C. 2017. Simple and scalable predictive uncertainty estimation using deep ensembles//Proceedings of the 31st International Conference on Neural Information Processing Systems. Long Beach,USA: Curran Associates Inc.: 6402-6413, articleTitle=Simple and scalable predictive uncertainty estimation using deep ensembles, refAbstract=null), Reference(id=1322949820440257408, tenantId=1146029695717560320, journalId=1249024232475115590, articleId=1322949806875878170, doi=null, pmid=null, pmcid=null, year=2023, volume=27, issue=3, pageStart=1431, pageEnd=1442, url=null, language=null, rfNumber=null, rfOrder=49, authorNames=Lei T, Sun R, Du X G, Fu H Z, Zhang C Q, Nandi A K, journalName=IEEE Journal of Biomedical and Health Informatics, refType=null, unstructuredReference=Lei T,Sun R,Du X G,Fu H Z,Zhang C Q and Nandi A K. 2023. SGU-Net: shape-guided ultralight network for abdominal image segmentation. IEEE Journal of Biomedical and Health Informatics,27(3): 1431-1442 [DOI: 10.1109/JBHI.2023.3238183], articleTitle=SGU-Net: shape-guided ultralight network for abdominal image segmentation, refAbstract=null), Reference(id=1322949820507366273, tenantId=1146029695717560320, journalId=1249024232475115590, articleId=1322949806875878170, doi=null, pmid=null, pmcid=null, year=2025, volume=30, issue=5, pageStart=1510, pageEnd=1527, url=null, language=null, rfNumber=null, rfOrder=50, authorNames=Li F C, Gao S S, Liu Z, Zhang C M, Zhou Y F, journalName=Journal of Image and Graphics, refType=null, unstructuredReference=Li F C,Gao S S,Liu Z,Zhang C M and Zhou Y F. 2025. Multimodal medical image fusion with progressive feature extraction and frequency domain information complementation. Journal of Image and Graphics,30(5): 1510-1527, articleTitle=Multimodal medical image fusion with progressive feature extraction and frequency domain information complementation, refAbstract=null), Reference(id=1322949820566086530, tenantId=1146029695717560320, journalId=1249024232475115590, articleId=1322949806875878170, doi=null, pmid=null, pmcid=null, year=2025, volume=30, issue=5, pageStart=1510, pageEnd=1527, url=null, language=null, rfNumber=null, rfOrder=51, authorNames=李夫辰, 高珊珊, 刘峥, 张彩明, 周元峰, journalName=中国图象图形学报, refType=null, unstructuredReference=李夫辰,高珊珊,刘峥,张彩明,周元峰. 2025. 渐进特征提取和频域信息补充的多模态医学图像融合. 中国图象图形学报,30(5): 1510-1527 [DOI: 10.11834/jig.240509], articleTitle=渐进特征提取和频域信息补充的多模态医学图像融合, refAbstract=null), Reference(id=1322949820629001091, tenantId=1146029695717560320, journalId=1249024232475115590, articleId=1322949806875878170, doi=null, pmid=null, pmcid=null, year=2023, volume=35, issue=30, pageStart=22071, pageEnd=22085, url=null, language=null, rfNumber=null, rfOrder=52, authorNames=Li H, Nan Y, Del Ser J, Yang G, journalName=Neural Computing and Applications, refType=null, unstructuredReference=Li H,Nan Y,Del Ser J and Yang G. 2023. Region-based evidential deep learning to quantify uncertainty and improve robustness of brain tumor segmentation. Neural Computing and Applications,35(30): 22071-22085 [DOI: 10.1007/s00521-022-08016-4], articleTitle=Region-based evidential deep learning to quantify uncertainty and improve robustness of brain tumor segmentation, refAbstract=null), Reference(id=1322949820691915652, tenantId=1146029695717560320, journalId=1249024232475115590, articleId=1322949806875878170, doi=null, pmid=null, pmcid=null, year=2022, volume=9, issue=1, pageStart=null, pageEnd=52, url=null, language=null, rfNumber=null, rfOrder=53, authorNames=Li P H, Kong X Y, He Y Z, Liu Y, Peng X, Li Z H, journalName=Military Medical Research, refType=null, unstructuredReference=Li P H,Kong X Y,He Y Z,Liu Y,Peng X,Li Z H,et al. 2022. Recent developments in application of single-cell RNA sequencing in the tumour immune microenvironment and cancer therapy. Military Medical Research,9(1): #52 [DOI: 10.1186/s40779-022-00414-y], articleTitle=Recent developments in application of single-cell RNA sequencing in the tumour immune microenvironment and cancer therapy, refAbstract=null), Reference(id=1322949820763218821, tenantId=1146029695717560320, journalId=1249024232475115590, articleId=1322949806875878170, doi=null, pmid=null, pmcid=null, year=2025a, volume=null, issue=null, pageStart=57147, pageEnd=57173, url=null, language=null, rfNumber=null, rfOrder=54, authorNames=Liang X Y, Fu P H, Qian Y H, Guo Q, Liu G Q, journalName=null, refType=null, unstructuredReference=Liang X Y,Fu P H,Qian Y H,Guo Q and Liu G Q. 2025a. Trusted multi-view classification via evolutionary multi-view fusion//Proceedings of the 13th International Conference on Learning Representations. Singapore,Singapore: ICLR: 57147-57173, articleTitle=Trusted multi-view classification via evolutionary multi-view fusion, refAbstract=null), Reference(id=1322949820842910598, tenantId=1146029695717560320, journalId=1249024232475115590, articleId=1322949806875878170, doi=null, pmid=null, pmcid=null, year=2025b, volume=null, issue=null, pageStart=37409, pageEnd=37426, url=null, language=null, rfNumber=null, rfOrder=55, authorNames=Liang X Y, Wang S J, Qian Y H, Guo Q, Du L, Jiang B B, journalName=null, refType=null, unstructuredReference=Liang X Y,Wang S J,Qian Y H,Guo Q,Du L,Jiang B B,et al. 2025b. Trusted multi-view classification with expert knowledge constraints//Proceedings of the 42nd International Conference on Machine Learning. Vancouver,Canada: PMLR: 37409-37426, articleTitle=Trusted multi-view classification with expert knowledge constraints, refAbstract=null), Reference(id=1322949820905825159, tenantId=1146029695717560320, journalId=1249024232475115590, articleId=1322949806875878170, doi=null, pmid=null, pmcid=null, year=2024, volume=null, issue=null, pageStart=7278, pageEnd=7287, url=null, language=null, rfNumber=null, rfOrder=56, authorNames=Liu W, Chen Y F, Yue X D, journalName=null, refType=null, unstructuredReference=Liu W,Chen Y F and Yue X D. 2024. Building trust in decision with conformalized multi-view deep classification//Proceedings of the 32nd ACM International Conference on Multimedia. Melbourne,Australia: Association for Computing Machinery: 7278-7287 [DOI: 10.1145/3664647.3681297], articleTitle=Building trust in decision with conformalized multi-view deep classification, refAbstract=null), Reference(id=1322949820968739720, tenantId=1146029695717560320, journalId=1249024232475115590, articleId=1322949806875878170, doi=null, pmid=null, pmcid=null, year=2025a, volume=null, issue=null, pageStart=15508, pageEnd=15517, url=null, language=null, rfNumber=null, rfOrder=57, authorNames=Liu W, Chen Y F, Yue X D, journalName=null, refType=null, unstructuredReference=Liu W,Chen Y F and Yue X D. 2025a. Enhancing testing-time robustness for trusted multi-view classification in the wild//Proceedings of 2025 Computer Vision and Pattern Recognition Conference. Nashville,USA: IEEE: 15508-15517 [DOI: 10.1109/cvpr52734.2025.01445], articleTitle=Enhancing testing-time robustness for trusted multi-view classification in the wild, refAbstract=null), Reference(id=1322949821031654281, tenantId=1146029695717560320, journalId=1249024232475115590, articleId=1322949806875878170, doi=null, pmid=null, pmcid=null, year=2025b, volume=null, issue=null, pageStart=18969, pageEnd=18977, url=null, language=null, rfNumber=null, rfOrder=58, authorNames=Liu W, Chen Y F, Yue X D, journalName=null, refType=null, unstructuredReference=Liu W,Chen Y F and Yue X D. 2025b. Enhancing multi-view classification reliability with adaptive rejection//Proceedings of the 39th AAAI Conference on Artificial Intelligence. Philadelphia,USA: AAAI: 18969-18977 [DOI: 10.1609/aaai.v39i18.34088], articleTitle=Enhancing multi-view classification reliability with adaptive rejection, refAbstract=null), Reference(id=1322949821086180234, tenantId=1146029695717560320, journalId=1249024232475115590, articleId=1322949806875878170, doi=null, pmid=null, pmcid=null, year=2023a, volume=null, issue=null, pageStart=8870, pageEnd=8878, url=null, language=null, rfNumber=null, rfOrder=59, authorNames=Liu W, Chen Y F, Yue X D, Zhang C Q, Xie S R, journalName=null, refType=null, unstructuredReference=Liu W,Chen Y F,Yue X D,Zhang C Q and Xie S R. 2023a. Safe multi-view deep classification//Proceedings of the 37th AAAI Conference on Artificial Intelligence. Washington USA: AAAI: 8870-8878 [DOI: 10.1609/aaai.v37i7.26066], articleTitle=Safe multi-view deep classification, refAbstract=null), Reference(id=1322949821149094795, tenantId=1146029695717560320, journalId=1249024232475115590, articleId=1322949806875878170, doi=null, pmid=null, pmcid=null, year=2025c, volume=36, issue=2, pageStart=2290, pageEnd=2303, url=null, language=null, rfNumber=null, rfOrder=60, authorNames=Liu W, Chen Y F, Yue X D, Zhang C Q, Xie S R, journalName=IEEE Transactions on Circuits and Systems for Video Technology, refType=null, unstructuredReference=Liu W,Chen Y F,Yue X D,Zhang C Q and Xie S R. 2025c. Enhancing reliability in medical image classification of imperfect views. IEEE Transactions on Circuits and Systems for Video Technology,36(2): 2290-2303 [DOI: 10.1109/TCSVT.2025.3610762], articleTitle=Enhancing reliability in medical image classification of imperfect views, refAbstract=null), Reference(id=1322949821212009356, tenantId=1146029695717560320, journalId=1249024232475115590, articleId=1322949806875878170, doi=null, pmid=null, pmcid=null, year=2022, volume=null, issue=null, pageStart=7585, pageEnd=7593, url=null, language=null, rfNumber=null, rfOrder=61, authorNames=Liu W, Yue X D, Chen Y F, Denoeux T, journalName=null, refType=null, unstructuredReference=Liu W,Yue X D,Chen Y F and Denoeux T. 2022. Trusted multi-view deep learning with opinion aggregation//Proceedings of the 36th AAAI Conference on Artificial Intelligence. [s.l.]: AAAI: 7585-7593 [DOI: 10.1609/aaai.v36i7.20724], articleTitle=Trusted multi-view deep learning with opinion aggregation, refAbstract=null), Reference(id=1322949821287506829, tenantId=1146029695717560320, journalId=1249024232475115590, articleId=1322949806875878170, doi=null, pmid=null, pmcid=null, year=2023b, volume=42, issue=2, pageStart=557, pageEnd=567, url=null, language=null, rfNumber=null, rfOrder=62, authorNames=Liu Y B, Li H N, Luo T, Zhang C Q, Xiao Z T, Wei Y, journalName=IEEE Transactions on Medical Imaging, refType=null, unstructuredReference=Liu Y B,Li H N,Luo T,Zhang C Q,Xiao Z T,Wei Y,et al. 2023b. Structural attention graph neural network for diagnosis and prediction of COVID-19 severity. IEEE Transactions on Medical Imaging,42(2): 557-567 [DOI: 10.1109/TMI.2022.3226575], articleTitle=Structural attention graph neural network for diagnosis and prediction of COVID-19 severity, refAbstract=null), Reference(id=1322949821371392910, tenantId=1146029695717560320, journalId=1249024232475115590, articleId=1322949806875878170, doi=null, pmid=null, pmcid=null, year=2025d, volume=102, issue=null, pageStart=null, pageEnd=103507, url=null, language=null, rfNumber=null, rfOrder=63, authorNames=Liu Y Y, Gao Z Y, Shi N N, Wu F P, Shi Y X, Chen Q C, journalName=Medical Image Analysis, refType=null, unstructuredReference=Liu Y Y,Gao Z Y,Shi N N,Wu F P,Shi Y X,Chen Q C,et al. 2025d. MERIT: multi-view evidential learning for reliable and interpretable liver fibrosis staging. Medical Image Analysis,102: #103507 [DOI: 10.1016/j.media.2025.103507], articleTitle=MERIT: multi-view evidential learning for reliable and interpretable liver fibrosis staging, refAbstract=null), Reference(id=1322949821425918863, tenantId=1146029695717560320, journalId=1249024232475115590, articleId=1322949806875878170, doi=null, pmid=null, pmcid=null, year=2024, volume=69, issue=6, pageStart=null, pageEnd=065012, url=null, language=null, rfNumber=null, rfOrder=64, authorNames=Long J W, Ren Y, Yang C X, Ren P C, Zeng Z Q, journalName=Physics in Medicine and Biology, refType=null, unstructuredReference=Long J W,Ren Y,Yang C X,Ren P C and Zeng Z Q. 2024. MDT: semi-supervised medical image segmentation with mixup-decoupling training. Physics in Medicine and Biology,69(6): #065012 [DOI: 10.1088/1361-6560/ad2715], articleTitle=MDT: semi-supervised medical image segmentation with mixup-decoupling training, refAbstract=null), Reference(id=1322949821493027728, tenantId=1146029695717560320, journalId=1249024232475115590, articleId=1322949806875878170, doi=null, pmid=null, pmcid=null, year=2023, volume=13, issue=1, pageStart=null, pageEnd=14904, url=null, language=null, rfNumber=null, rfOrder=65, authorNames=Nagahama A, journalName=Scientific Reports, refType=null, unstructuredReference=Nagahama A. 2023. Learning and predicting the unknown class using evidential deep learning. Scientific Reports,13(1): #14904 [DOI: 10.1038/s41598-023-40649-w], articleTitle=Learning and predicting the unknown class using evidential deep learning, refAbstract=null), Reference(id=1322949821547553681, tenantId=1146029695717560320, journalId=1249024232475115590, articleId=1322949806875878170, doi=null, pmid=null, pmcid=null, year=2025, volume=159, issue=null, pageStart=null, pageEnd=111225, url=null, language=null, rfNumber=null, rfOrder=66, authorNames=Noor M H M, Ige A O, journalName=Engineering Applications of Artificial Intelligence, refType=null, unstructuredReference=Noor M H M and Ige A O. 2025. A survey on state-of-the-art deep learning applications and challenges. Engineering Applications of Artificial Intelligence,159: #111225 [DOI: 10.1016/j.engappai.2025.111225], articleTitle=A survey on state-of-the-art deep learning applications and challenges, refAbstract=null), Reference(id=1322949821610468242, tenantId=1146029695717560320, journalId=1249024232475115590, articleId=1322949806875878170, doi=null, pmid=null, pmcid=null, year=2024, volume=57, issue=11, pageStart=null, pageEnd=304, url=null, language=null, rfNumber=null, rfOrder=67, authorNames=Pacal I, Kunduracioglu I, Alma M H, Deveci M, Kadry S, Nedoma J, journalName=Artificial Intelligence Review, refType=null, unstructuredReference=Pacal I,Kunduracioglu I,Alma M H,Deveci M,Kadry S,Nedoma J,et al. 2024. A systematic review of deep learning techniques for plant diseases. Artificial Intelligence Review,57(11): #304 [DOI: 10.1007/s10462-024-10944-7], articleTitle=A systematic review of deep learning techniques for plant diseases, refAbstract=null), Reference(id=1322949821681771411, tenantId=1146029695717560320, journalId=1249024232475115590, articleId=1322949806875878170, doi=null, pmid=null, pmcid=null, year=2025, volume=30, issue=6, pageStart=1985, pageEnd=2000, url=null, language=null, rfNumber=null, rfOrder=68, authorNames=Pan Y S, Ma H J, Xia Y, Zhang Y N, journalName=Journal of Image and Graphics, refType=null, unstructuredReference=Pan Y S,Ma H J,Xia Y and Zhang Y N. 2025. Application of content generation in medical images. Journal of Image and Graphics,30(6): 1985-2000, articleTitle=Application of content generation in medical images, refAbstract=null), Reference(id=1322949821753074580, tenantId=1146029695717560320, journalId=1249024232475115590, articleId=1322949806875878170, doi=null, pmid=null, pmcid=null, year=2025, volume=30, issue=6, pageStart=1985, pageEnd=2000, url=null, language=null, rfNumber=null, rfOrder=69, authorNames=潘永生, 马豪杰, 夏勇, 张艳宁, journalName=中国图象图形学报, refType=null, unstructuredReference=潘永生,马豪杰,夏勇,张艳宁. 2025. 医学影像中的生成技术. 中国图象图形学报,30(6): 1985-2000 [DOI: 10.11834/jig.240459], articleTitle=医学影像中的生成技术, refAbstract=null), Reference(id=1322949821828572053, tenantId=1146029695717560320, journalId=1249024232475115590, articleId=1322949806875878170, doi=null, pmid=null, pmcid=null, year=2025, volume=null, issue=null, pageStart=1, pageEnd=5, url=null, language=null, rfNumber=null, rfOrder=70, authorNames=Pang S J, Zhao Y Z, Zhao R W, Feng R, journalName=Proceedings of the 22nd IEEE International Symposium on Biomedical Imaging (ISBI), refType=null, unstructuredReference=Pang S J,Zhao Y Z,Zhao R W and Feng R. 2025. Towards multi-label open set diagnosis for medical images//Proceedings of the 22nd IEEE International Symposium on Biomedical Imaging (ISBI). Houston,USA: IEEE: 1-5 [DOI: 10.1109/isbi60581.2025.10980926], articleTitle=Towards multi-label open set diagnosis for medical images, refAbstract=null), Reference(id=1322949821895680918, tenantId=1146029695717560320, journalId=1249024232475115590, articleId=1322949806875878170, doi=null, pmid=null, pmcid=null, year=2024, volume=46, issue=8, pageStart=5288, pageEnd=5305, url=null, language=null, rfNumber=null, rfOrder=71, authorNames=Pei J B, Men A, Liu Y, Zhuang X H, Chen Q C, journalName=IEEE Transactions on Pattern Analysis and Machine Intelligence, refType=null, unstructuredReference=Pei J B,Men A,Liu Y,Zhuang X H and Chen Q C. 2024. Evidential multi-source-free unsupervised domain adaptation. IEEE Transactions on Pattern Analysis and Machine Intelligence,46(8): 5288-5305 [DOI: 10.1109/tpami.2024.3361978], articleTitle=Evidential multi-source-free unsupervised domain adaptation, refAbstract=null), Reference(id=1322949821954401175, tenantId=1146029695717560320, journalId=1249024232475115590, articleId=1322949806875878170, doi=null, pmid=null, pmcid=null, year=2023, volume=16, issue=1, pageStart=null, pageEnd=44, url=null, language=null, rfNumber=null, rfOrder=72, authorNames=Pei X D, Zuo K, Li Y, Pang Z B, journalName=International Journal of Computational Intelligence Systems, refType=null, unstructuredReference=Pei X D,Zuo K,Li Y and Pang Z B. 2023. A review of the application of multi-modal deep learning in medicine: bibliometrics and future directions. International Journal of Computational Intelligence Systems,16(1): #44 [DOI: 10.1007/s44196-023-00225-6], articleTitle=A review of the application of multi-modal deep learning in medicine: bibliometrics and future directions, refAbstract=null), Reference(id=1322949822017315736, tenantId=1146029695717560320, journalId=1249024232475115590, articleId=1322949806875878170, doi=null, pmid=null, pmcid=null, year=2024, volume=43, issue=6, pageStart=2050, pageEnd=2060, url=null, language=null, rfNumber=null, rfOrder=73, authorNames=Penso C, Frenkel L, Goldberger J, journalName=IEEE Transactions on Medical Imaging, refType=null, unstructuredReference=Penso C,Frenkel L and Goldberger J. 2024. Confidence calibration of a medical imaging classification system that is robust to label noise. IEEE Transactions on Medical Imaging,43(6): 2050-2060 [DOI: 10.1109/TMI.2024.3353762], articleTitle=Confidence calibration of a medical imaging classification system that is robust to label noise, refAbstract=null), Reference(id=1322949822092813209, tenantId=1146029695717560320, journalId=1249024232475115590, articleId=1322949806875878170, doi=null, pmid=null, pmcid=null, year=2025, volume=324, issue=null, pageStart=null, pageEnd=113838, url=null, language=null, rfNumber=null, rfOrder=74, authorNames=Qi Y X, Lin L, Zhang B, Zhang J Y, Wang J J, journalName=Knowledge-Based Systems, refType=null, unstructuredReference=Qi Y X,Lin L,Zhang B,Zhang J Y and Wang J J. 2025. Multi-modal evidential fusion network for trustworthy PET/CT tumor segmentation. Knowledge-Based Systems,324: #113838 [DOI: 10.1016/j.knosys.2025.113838], articleTitle=Multi-modal evidential fusion network for trustworthy PET/CT tumor segmentation, refAbstract=null), Reference(id=1322949822172504986, tenantId=1146029695717560320, journalId=1249024232475115590, articleId=1322949806875878170, doi=null, pmid=null, pmcid=null, year=2017, volume=28, issue=4, pageStart=917, pageEnd=933, url=null, language=null, rfNumber=null, rfOrder=75, authorNames=Qiu Z C, Miller D J, Kesidis G, journalName=IEEE Transactions on Neural Networks and Learning Systems, refType=null, unstructuredReference=Qiu Z C,Miller D J and Kesidis G. 2017. A maximum entropy framework for semisupervised and active learning with unknown and label-scarce classes. IEEE Transactions on Neural Networks and Learning Systems,28(4): 917-933 [DOI: 10.1109/TNNLS.2016.2514401], articleTitle=A maximum entropy framework for semisupervised and active learning with unknown and label-scarce classes, refAbstract=null), Reference(id=1322949822260585371, tenantId=1146029695717560320, journalId=1249024232475115590, articleId=1322949806875878170, doi=null, pmid=null, pmcid=null, year=2024, volume=null, issue=null, pageStart=#2687, pageEnd=null, url=null, language=null, rfNumber=null, rfOrder=76, authorNames=Qu J E, Chen Y F, Yue X D, Fu W, Huang Q G, journalName=null, refType=null, unstructuredReference=Qu J E,Chen Y F,Yue X D,Fu W and Huang Q G. 2024. Hyper-opinion evidential deep learning for out-of-distribution detection//Proceedings of the 38th International Conference on Neural Information Processing Systems. Vancouver,Canada: Curran Associates Inc.: #2687, articleTitle=Hyper-opinion evidential deep learning for out-of-distribution detection, refAbstract=null), Reference(id=1322949822336082844, tenantId=1146029695717560320, journalId=1249024232475115590, articleId=1322949806875878170, doi=null, pmid=null, pmcid=null, year=2024, volume=47, issue=null, pageStart=null, pageEnd=101504, url=null, language=null, rfNumber=null, rfOrder=77, authorNames=Rayed M E, Islam S M S, Niha S I, Jim J R, Kabir M M, Mridha M F, journalName=Informatics in Medicine Unlocked, refType=null, unstructuredReference=Rayed M E,Islam S M S,Niha S I,Jim J R,Kabir M M and Mridha M F. 2024. Deep learning for medical image segmentation: state-of-the-art advancements and challenges. Informatics in Medicine Unlocked,47: #101504 [DOI: 10.1016/j.imu.2024.101504], articleTitle=Deep learning for medical image segmentation: state-of-the-art advancements and challenges, refAbstract=null), Reference(id=1322949822415774621, tenantId=1146029695717560320, journalId=1249024232475115590, articleId=1322949806875878170, doi=null, pmid=null, pmcid=null, year=2023, volume=null, issue=null, pageStart=35, pageEnd=45, url=null, language=null, rfNumber=null, rfOrder=78, authorNames=Ren K, Zou K, Liu X J, Chen Y D, Yuan X D, Shen X J, journalName=null, refType=null, unstructuredReference=Ren K,Zou K,Liu X J,Chen Y D,Yuan X D,Shen X J,et al. 2023. Uncertainty-informed mutual learning for joint medical image classification and segmentation//Proceedings of the 26th International Conference on Medical Image Computing and Computer-Assisted Intervention. Vancouver,Canada: Springer: 35-45 [DOI: 10.1007/978-3-031-43901-8_4], articleTitle=Uncertainty-informed mutual learning for joint medical image classification and segmentation, refAbstract=null), Reference(id=1322949822491272094, tenantId=1146029695717560320, journalId=1249024232475115590, articleId=1322949806875878170, doi=null, pmid=null, pmcid=null, year=2022, volume=75, issue=null, pageStart=null, pageEnd=102274, url=null, language=null, rfNumber=null, rfOrder=79, authorNames=Roy A G, Ren J, Azizi S, Loh A, Natarajan V, Mustafa B, journalName=Medical Image Analysis, refType=null, unstructuredReference=Roy A G,Ren J,Azizi S,Loh A,Natarajan V,Mustafa B,et al. 2022. Does your dermatology classifier know what it doesn’t know? Detecting the long-tail of unseen conditions. Medical Image Analysis,75: #102274 [DOI: 10.1016/j.media.2021.102274], articleTitle=Does your dermatology classifier know what it doesn’t know? Detecting the long-tail of unseen conditions, refAbstract=null), Reference(id=1322949822587741087, tenantId=1146029695717560320, journalId=1249024232475115590, articleId=1322949806875878170, doi=null, pmid=null, pmcid=null, year=2018, volume=null, issue=null, pageStart=3183, pageEnd=3193, url=null, language=null, rfNumber=null, rfOrder=80, authorNames=Sensoy M, Kaplan L M, Kandemir M, journalName=null, refType=null, unstructuredReference=Sensoy M,Kaplan L M and Kandemir M. 2018. Evidential deep learning to quantify classification uncertainty//Proceedings of the 32nd International Conference on Neural Information Processing Systems. Montréal,Canada: Curran Associates Inc.: 3183-3193, articleTitle=Evidential deep learning to quantify classification uncertainty, refAbstract=null), Reference(id=1322949822659044256, tenantId=1146029695717560320, journalId=1249024232475115590, articleId=1322949806875878170, doi=null, pmid=null, pmcid=null, year=1976, volume=null, issue=null, pageStart=null, pageEnd=null, url=null, language=null, rfNumber=null, rfOrder=81, authorNames=Shafer G, journalName=null, refType=null, unstructuredReference=Shafer G. 1976. A Mathematical Theory of Evidence. Princeton,USA: Princeton University Press, articleTitle=A Mathematical Theory of Evidence. Princeton,USA: Princeton University Press, refAbstract=null), Reference(id=1322949822759707553, tenantId=1146029695717560320, journalId=1249024232475115590, articleId=1322949806875878170, doi=null, pmid=null, pmcid=null, year=2023, volume=10, issue=7, pageStart=3897, pageEnd=3904, url=null, language=null, rfNumber=null, rfOrder=82, authorNames=Sharifani K, Amini M, journalName=World Information Technology and Engineering Journal, refType=null, unstructuredReference=Sharifani K and Amini M. 2023. Machine learning and deep learning: a review of methods and applications. World Information Technology and Engineering Journal,10(7): 3897-3904, articleTitle=Machine learning and deep learning: a review of methods and applications, refAbstract=null), Reference(id=1322949822835205026, tenantId=1146029695717560320, journalId=1249024232475115590, articleId=1322949806875878170, doi=null, pmid=null, pmcid=null, year=2024, volume=97, issue=null, pageStart=null, pageEnd=103294, url=null, language=null, rfNumber=null, rfOrder=83, authorNames=Shi J B, Li C, Gong T L, Fu H Z, journalName=Medical Image Analysis, refType=null, unstructuredReference=Shi J B,Li C,Gong T L and Fu H Z. 2024. E2-MIL: an explainable and evidential multiple instance learning framework for whole slide image classification. Medical Image Analysis,97: #103294 [DOI: 10.1016/j.media.2024.103294], articleTitle=E2-MIL: an explainable and evidential multiple instance learning framework for whole slide image classification, refAbstract=null), Reference(id=1322949822910702499, tenantId=1146029695717560320, journalId=1249024232475115590, articleId=1322949806875878170, doi=null, pmid=null, pmcid=null, year=2022, volume=11, issue=1, pageStart=19, pageEnd=38, url=null, language=null, rfNumber=null, rfOrder=84, authorNames=Suganyadevi S, Seethalakshmi V, Balasamy K, journalName=International Journal of Multimedia Information Retrieval, refType=null, unstructuredReference=Suganyadevi S,Seethalakshmi V and Balasamy K. 2022. A review on deep learning in medical image analysis. International Journal of Multimedia Information Retrieval,11(1): 19-38 [DOI: 10.1007/s13735-021-00218-1], articleTitle=A review on deep learning in medical image analysis, refAbstract=null), Reference(id=1322949822973617060, tenantId=1146029695717560320, journalId=1249024232475115590, articleId=1322949806875878170, doi=null, pmid=null, pmcid=null, year=2026, volume=45, issue=1, pageStart=6, pageEnd=27, url=null, language=null, rfNumber=null, rfOrder=85, authorNames=Sun S C, Chen Y F, Yue X D, Ma C, Zhuang X H, journalName=IEEE Transactions on Medical Imaging, refType=null, unstructuredReference=Sun S C,Chen Y F,Yue X D,Ma C and Zhuang X H. 2026. REDNet: reliable evidential discounting network for multi-modality medical image segmentation. IEEE Transactions on Medical Imaging,45(1): 6-27 [DOI: 10.1109/tmi.2025.3591124], articleTitle=REDNet: reliable evidential discounting network for multi-modality medical image segmentation, refAbstract=null), Reference(id=1322949823040725925, tenantId=1146029695717560320, journalId=1249024232475115590, articleId=1322949806875878170, doi=null, pmid=null, pmcid=null, year=2024, volume=48, issue=1, pageStart=null, pageEnd=84, url=null, language=null, rfNumber=null, rfOrder=86, authorNames=Takahashi S, Sakaguchi Y, Kouno N, Takasawa K, Ishizu K, Akagi Y, journalName=Journal of Medical Systems, refType=null, unstructuredReference=Takahashi S,Sakaguchi Y,Kouno N,Takasawa K,Ishizu K,Akagi Y,et al. 2024. Comparison of vision transformers and convolutional neural networks in medical image analysis: a systematic review. Journal of Medical Systems,48(1): #84 [DOI: 10.1007/s10916-024-02105-8], articleTitle=Comparison of vision transformers and convolutional neural networks in medical image analysis: a systematic review, refAbstract=null), Reference(id=1322949823099446182, tenantId=1146029695717560320, journalId=1249024232475115590, articleId=1322949806875878170, doi=null, pmid=null, pmcid=null, year=2024a, volume=null, issue=null, pageStart=91, pageEnd=105, url=null, language=null, rfNumber=null, rfOrder=87, authorNames=Tan H S, Wang K C, Mcbeth R, journalName=null, refType=null, unstructuredReference=Tan H S,Wang K C and Mcbeth R. 2024a. Uncertainty-error correlations in evidential deep learning models for biomedical segmentation//Proceedings of the 29th International Conference on Technologies and Applications of Artificial Intelligence. Hsinchu,China: Springer: 91-105 [DOI: 10.1007/978-981-96-4589-3_7], articleTitle=Uncertainty-error correlations in evidential deep learning models for biomedical segmentation, refAbstract=null), Reference(id=1322949823174943655, tenantId=1146029695717560320, journalId=1249024232475115590, articleId=1322949806875878170, doi=null, pmid=null, pmcid=null, year=2024b, volume=182, issue=null, pageStart=null, pageEnd=109172, url=null, language=null, rfNumber=null, rfOrder=88, authorNames=Tan H S, Wang K C, McBeth R, journalName=Computers in Biology and Medicine, refType=null, unstructuredReference=Tan H S,Wang K C and McBeth R. 2024b. Deep evidential learning for radiotherapy dose prediction. Computers in Biology and Medicine,182: #109172 [DOI: 10.1016/j.compbiomed.2024.109172], articleTitle=Deep evidential learning for radiotherapy dose prediction, refAbstract=null), Reference(id=1322949823258829736, tenantId=1146029695717560320, journalId=1249024232475115590, articleId=1322949806875878170, doi=null, pmid=null, pmcid=null, year=2020, volume=null, issue=null, pageStart=4, pageEnd=354, url=null, language=null, rfNumber=null, rfOrder=89, authorNames=Ulmer D, Meijerink L, Cin􀅣 G, journalName=ML, refType=null, unstructuredReference=Ulmer D,Meijerink L,Cin􀅣 G. 2020. Trust issues: uncertainty estimation does not enable reliable OOD detection on medical tabular data//Proceedings of the Machine Learning for Health Workshop,ML4H@NeurIPS 2020. [s.l.]: PMLR: 341-354, articleTitle=Trust issues: uncertainty estimation does not enable reliable OOD detection on medical tabular data, refAbstract=null), Reference(id=1322949823388853161, tenantId=1146029695717560320, journalId=1249024232475115590, articleId=1322949806875878170, doi=null, pmid=null, pmcid=null, year=2025, volume=16, issue=1, pageStart=null, pageEnd=5528, url=null, language=null, rfNumber=null, rfOrder=90, authorNames=Wang M, Lin T, Lin A D, Yu K, Peng Y Y, Wang L Y, journalName=Nature Communications, refType=null, unstructuredReference=Wang M,Lin T,Lin A D,Yu K,Peng Y Y,Wang L Y,et al. 2025. Enhancing diagnostic accuracy in rare and common fundus diseases with a knowledge-rich vision-language model. Nature Communications,16(1): #5528 [DOI: 10.1038/s41467-025-60577-9], articleTitle=Enhancing diagnostic accuracy in rare and common fundus diseases with a knowledge-rich vision-language model, refAbstract=null), Reference(id=1322949823460156330, tenantId=1146029695717560320, journalId=1249024232475115590, articleId=1322949806875878170, doi=null, pmid=null, pmcid=null, year=2023, volume=14, issue=1, pageStart=null, pageEnd=6757, url=null, language=null, rfNumber=null, rfOrder=91, authorNames=Wang M, Lin T, Wang L Y, Lin A D, Zou K, Xu X X, journalName=Nature Communications, refType=null, unstructuredReference=Wang M,Lin T,Wang L Y,Lin A D,Zou K,Xu X X,et al. 2023. Uncertainty-inspired open set learning for retinal anomaly identification. Nature Communications,14(1): #6757 [DOI: 10.1038/s41467-023-42444-7], articleTitle=Uncertainty-inspired open set learning for retinal anomaly identification, refAbstract=null), Reference(id=1322949823531459499, tenantId=1146029695717560320, journalId=1249024232475115590, articleId=1322949806875878170, doi=null, pmid=null, pmcid=null, year=2024, volume=43, issue=11, pageStart=3990, pageEnd=4003, url=null, language=null, rfNumber=null, rfOrder=92, authorNames=Wu Q, Chen Y F, Liu W, Yue X D, Zhuang X H, journalName=IEEE Transactions on Medical Imaging, refType=null, unstructuredReference=Wu Q,Chen Y F,Liu W,Yue X D and Zhuang X H. 2024. Deep closing: enhancing topological connectivity in medical tubular segmentation. IEEE Transactions on Medical Imaging,43(11): 3990-4003 [DOI: 10.1109/TMI.2024.3405982], articleTitle=Deep closing: enhancing topological connectivity in medical tubular segmentation, refAbstract=null), Reference(id=1322949823619539884, tenantId=1146029695717560320, journalId=1249024232475115590, articleId=1322949806875878170, doi=null, pmid=null, pmcid=null, year=2024, volume=28, issue=11, pageStart=6417, pageEnd=6428, url=null, language=null, rfNumber=null, rfOrder=93, authorNames=Xia T, Dang T, Han J, Qendro L, Mascolo C, journalName=IEEE Journal of Biomedical and Health Informatics, refType=null, unstructuredReference=Xia T,Dang T,Han J,Qendro L and Mascolo C. 2024. Uncertainty-aware health diagnostics via class-balanced evidential deep learning. IEEE Journal of Biomedical and Health Informatics,28(11): 6417-6428 [DOI: 10.1109/jbhi.2024.3360002], articleTitle=Uncertainty-aware health diagnostics via class-balanced evidential deep learning, refAbstract=null), Reference(id=1322949823686648749, tenantId=1146029695717560320, journalId=1249024232475115590, articleId=1322949806875878170, doi=null, pmid=null, pmcid=null, year=2022, volume=null, issue=null, pageStart=1, pageEnd=2022, url=null, language=null, rfNumber=null, rfOrder=94, authorNames=Xia T, Han J, Qendro L, Dang T, Mascolo C, journalName=Proceedings of Workshop on Trustworthy and Socially Responsible Machine Learning, refType=null, unstructuredReference=Xia T,Han J,Qendro L,Dang T and Mascolo C. 2022. Hybrid-EDL: improving evidential deep learning for uncertainty quantification on imbalanced data//Proceedings of Workshop on Trustworthy and Socially Responsible Machine Learning,NeurIPS 2022. OrleansNew,USA: Curran Associates Inc.: 1-16, articleTitle=Hybrid-EDL: improving evidential deep learning for uncertainty quantification on imbalanced data, refAbstract=null), Reference(id=1322949823757951918, tenantId=1146029695717560320, journalId=1249024232475115590, articleId=1322949806875878170, doi=null, pmid=null, pmcid=null, year=2024, volume=14, issue=5, pageStart=1023, pageEnd=1035, url=null, language=null, rfNumber=null, rfOrder=95, authorNames=Xiao Z Y, Sun H, Liu F, journalName=Biomedical Engineering Letters, refType=null, unstructuredReference=Xiao Z Y,Sun H and Liu F. 2024. Semi-supervised CT image segmentation via contrastive learning based on entropy constraints. Biomedical Engineering Letters,14(5): 1023-1035 [DOI: 10.1007/s13534-024-00387-y], articleTitle=Semi-supervised CT image segmentation via contrastive learning based on entropy constraints, refAbstract=null), Reference(id=1322949823833449391, tenantId=1146029695717560320, journalId=1249024232475115590, articleId=1322949806875878170, doi=null, pmid=null, pmcid=null, year=2022a, volume=82, issue=null, pageStart=null, pageEnd=102643, url=null, language=null, rfNumber=null, rfOrder=96, authorNames=Xu L, Wu H, He C M, Wang J, Zhang C Q, Nie F P, journalName=Medical Image Analysis, refType=null, unstructuredReference=Xu L,Wu H,He C M,Wang J,Zhang C Q,Nie F P,et al. 2022a. Multi-modal sequence learning for Alzheimer’s disease progression prediction with incomplete variable-length longitudinal data. Medical Image Analysis,82: #102643 [DOI: 10.1016/j.media.2022.102643], articleTitle=Multi-modal sequence learning for Alzheimer’s disease progression prediction with incomplete variable-length longitudinal data, refAbstract=null), Reference(id=1322949823904752560, tenantId=1146029695717560320, journalId=1249024232475115590, articleId=1322949806875878170, doi=null, pmid=null, pmcid=null, year=2022b, volume=150, issue=null, pageStart=188, pageEnd=198, url=null, language=null, rfNumber=null, rfOrder=97, authorNames=Xu S X, Chen Y F, Ma C, Yue X D, journalName=International Journal of Approximate Reasoning, refType=null, unstructuredReference=Xu S X,Chen Y F,Ma C and Yue X D. 2022b. Deep evidential fusion network for medical image classification. International Journal of Approximate Reasoning,150: 188-198 [DOI: 10.1016/j.ijar.2022.08.013], articleTitle=Deep evidential fusion network for medical image classification, refAbstract=null), Reference(id=1322949823971861425, tenantId=1146029695717560320, journalId=1249024232475115590, articleId=1322949806875878170, doi=null, pmid=null, pmcid=null, year=2023, volume=null, issue=null, pageStart=10657, pageEnd=10665, url=null, language=null, rfNumber=null, rfOrder=98, authorNames=Xu Z K, Yue X D, Lv Y, Liu W, Li Z H, journalName=null, refType=null, unstructuredReference=Xu Z K,Yue X D,Lv Y,Liu W and Li Z H. 2023. Trusted fine-grained image classification through hierarchical evidence fusion//Proceedings of the 37th AAAI Conference on Artificial Intelligence. Washington,USA: AAAI: 10657-10665 [DOI: 10.1609/aaai.v37i9.26265], articleTitle=Trusted fine-grained image classification through hierarchical evidence fusion, refAbstract=null), Reference(id=1322949824038970291, tenantId=1146029695717560320, journalId=1249024232475115590, articleId=1322949806875878170, doi=null, pmid=null, pmcid=null, year=2025, volume=37, issue=11, pageStart=7541, pageEnd=7560, url=null, language=null, rfNumber=null, rfOrder=99, authorNames=Yan M, Lin Y, Peng X, Zeng Z, journalName=Neural Computing and Applications, refType=null, unstructuredReference=Yan M,Lin Y,Peng X and Zeng Z. 2025. mixDA: mixup domain adaptation for glaucoma detection on fundus images. Neural Computing and Applications,37(11): 7541-7560 [DOI: 10.1007/s00521-023-08572-3], articleTitle=mixDA: mixup domain adaptation for glaucoma detection on fundus images, refAbstract=null), Reference(id=1322949824118662068, tenantId=1146029695717560320, journalId=1249024232475115590, articleId=1322949806875878170, doi=null, pmid=null, pmcid=null, year=2024, volume=null, issue=null, pageStart=#714, pageEnd=null, url=null, language=null, rfNumber=null, rfOrder=100, authorNames=Yao H Y, Han Z B, Fu H Z, Peng X, Hu Q H, Zhang C Q, journalName=Out-of-distribution detection with diversification (provably)//Proceedings of the 38th International Conference on Neural Information Processing Systems, refType=null, unstructuredReference=Yao H Y,Han Z B,Fu H Z,Peng X,Hu Q H and Zhang C Q. 2024. Out-of-distribution detection with diversification (provably)//Proceedings of the 38th International Conference on Neural Information Processing Systems. Vancouver,Canada: Curran Associates Inc.: #714, articleTitle=null, refAbstract=null), Reference(id=1322949824194159541, tenantId=1146029695717560320, journalId=1249024232475115590, articleId=1322949806875878170, doi=null, pmid=null, pmcid=null, year=2022, volume=95, issue=null, pageStart=null, pageEnd=102026, url=null, language=null, rfNumber=null, rfOrder=101, authorNames=Yeung M, Sala E, Schönlieb C B, Rundo L, journalName=Computerized Medical Imaging and Graphics, refType=null, unstructuredReference=Yeung M,Sala E,Schönlieb C B and Rundo L. 2022. Unified focal loss: generalising dice and cross entropy-based losses to handle class imbalanced medical image segmentation. Computerized Medical Imaging and Graphics,95: #102026 [DOI: 10.1016/j.compmedimag.2021.102026], articleTitle=Unified focal loss: generalising dice and cross entropy-based losses to handle class imbalanced medical image segmentation, refAbstract=null), Reference(id=1322949824265462710, tenantId=1146029695717560320, journalId=1249024232475115590, articleId=1322949806875878170, doi=null, pmid=null, pmcid=null, year=2021, volume=null, issue=null, pageStart=102, pageEnd=111, url=null, language=null, rfNumber=null, rfOrder=102, authorNames=Zhang O, Delbrouck J B, Rubin D L, journalName=Proceedings of the 3rd International Workshop on Uncertainty for Safe Utilization of Machine Learning in Medical Imaging,and Perinatal Imaging,Placental and Preterm Image Analysis, refType=null, unstructuredReference=Zhang O,Delbrouck J B and Rubin D L. 2021. Out of distribution detection for medical images//Proceedings of the 3rd International Workshop on Uncertainty for Safe Utilization of Machine Learning in Medical Imaging,and Perinatal Imaging,Placental and Preterm Image Analysis. Strasbourg,France: Springer: 102-111 [DOI: 10.1007/978-3-030-87735-4_10], articleTitle=Out of distribution detection for medical images, refAbstract=null), Reference(id=1322949824328377271, tenantId=1146029695717560320, journalId=1249024232475115590, articleId=1322949806875878170, doi=null, pmid=null, pmcid=null, year=2024a, volume=176, issue=null, pageStart=null, pageEnd=108609, url=null, language=null, rfNumber=null, rfOrder=103, authorNames=Zhang Z X, Zhou H, Shi X R, Ran R, Tian C N, Zhou F, journalName=Computers in Biology and Medicine, refType=null, unstructuredReference=Zhang Z X,Zhou H,Shi X R,Ran R,Tian C N and Zhou F. 2024a. Quality-driven deep cross-supervised learning network for semi-supervised medical image segmentation. Computers in Biology and Medicine,176: #108609 [DOI: 10.1016/j.compbiomed.2024.108609], articleTitle=Quality-driven deep cross-supervised learning network for semi-supervised medical image segmentation, refAbstract=null), Reference(id=1322949824416457656, tenantId=1146029695717560320, journalId=1249024232475115590, articleId=1322949806875878170, doi=null, pmid=null, pmcid=null, year=2024b, volume=73, issue=null, pageStart=null, pageEnd=2535713, url=null, language=null, rfNumber=null, rfOrder=104, authorNames=Zhang Z X, Zhou H, Shi X R, Ran R, Tian C N, Zhou F, journalName=IEEE Transactions on Instrumentation and Measurement, refType=null, unstructuredReference=Zhang Z X,Zhou H,Shi X R,Ran R,Tian C N and Zhou F. 2024b. An evidential-enhanced tri-branch consistency learning method for semi-supervised medical image segmentation. IEEE Transactions on Instrumentation and Measurement,73: #2535713 [DOI: 10.1109/tim.2024.3488143], articleTitle=An evidential-enhanced tri-branch consistency learning method for semi-supervised medical image segmentation, refAbstract=null), Reference(id=1322949824491955129, tenantId=1146029695717560320, journalId=1249024232475115590, articleId=1322949806875878170, doi=null, pmid=null, pmcid=null, year=2024, volume=null, issue=null, pageStart=4028, pageEnd=4031, url=null, language=null, rfNumber=null, rfOrder=105, authorNames=Zhao R W, Xu Y Q, Zhang Y, Feng R, journalName=null, refType=null, unstructuredReference=Zhao R W,Xu Y Q,Zhang Y and Feng R. 2024. Evidential open set recognition for imbalanced medical images via multi-level data augmentation//Proceedings of 2024 IEEE International Conference on Bioinformatics and Biomedicine (BIBM). Lisbon,Portugal: IEEE: 4028-4031 [DOI: 10.1109/bibm62325.2024.10822558], articleTitle=Evidential open set recognition for imbalanced medical images via multi-level data augmentation, refAbstract=null), Reference(id=1322949824571646906, tenantId=1146029695717560320, journalId=1249024232475115590, articleId=1322949806875878170, doi=null, pmid=null, pmcid=null, year=2025, volume=null, issue=null, pageStart=null, pageEnd=null, url=https://arxiv.org/pdf/2505.20691.pdf, language=null, rfNumber=null, rfOrder=106, authorNames=Zhao S K, Zhang X A, Pan L P, Xu X B, Pelusi D, journalName=null, refType=null, unstructuredReference=Zhao S K,Zhang X A,Pan L P,Xu X B and Pelusi D. 2025. Evidential deep active learning for semi-supervised classification [EB/OL]. [2025-09-15]., articleTitle=Evidential deep active learning for semi-supervised classification, refAbstract=null), Reference(id=1322949824638755771, tenantId=1146029695717560320, journalId=1249024232475115590, articleId=1322949806875878170, doi=null, pmid=null, pmcid=null, year=2025, volume=30, issue=2, pageStart=0575, pageEnd=0588, url=null, language=null, rfNumber=null, rfOrder=107, authorNames=Zhao X M, Shi P L, Wang D D, Fu Y Y, Zhang S Q, Fang J X, journalName=Journal of Image and Graphics, refType=null, unstructuredReference=Zhao X M,Shi P L,Wang D D,Fu Y Y,Zhang S Q and Fang J X. 2025. Distribution-aware mean teacher networks for semi-supervised medical image segmentation. Journal of Image and Graphics,30(2): 0575-0588, articleTitle=Distribution-aware mean teacher networks for semi-supervised medical image segmentation, refAbstract=null), Reference(id=1322949824697476028, tenantId=1146029695717560320, journalId=1249024232475115590, articleId=1322949806875878170, doi=null, pmid=null, pmcid=null, year=2025, volume=30, issue=2, pageStart=575, pageEnd=588, url=null, language=null, rfNumber=null, rfOrder=108, authorNames=赵小明, 石培炼, 王丹丹, 付有瑶, 张石清, 方江雄, journalName=中国图象图形学报, refType=null, unstructuredReference=赵小明,石培炼,王丹丹,付有瑶,张石清,方江雄. 2025. 分布感知均值教师网络的半监督医学影像分割. 中国图象图形学报,30(2): 575-588 [DOI: 10.11834/jig.240208], articleTitle=分布感知均值教师网络的半监督医学影像分割, refAbstract=null), Reference(id=1322949824768779197, tenantId=1146029695717560320, journalId=1249024232475115590, articleId=1322949806875878170, doi=null, pmid=null, pmcid=null, year=2020, volume=25, issue=4, pageStart=629, pageEnd=654, url=null, language=null, rfNumber=null, rfOrder=109, authorNames=Zhao Y Q, Rao Y, Dong S P, Zhang J Y, journalName=Journal of image and graphics, refType=null, unstructuredReference=Zhao Y Q,Rao Y,Dong S P and Zhang J Y. 2020. Survey on deep learning object detection. Journal of image and graphics,25(4): 629-654, articleTitle=Survey on deep learning object detection, refAbstract=null), Reference(id=1322949824840082366, tenantId=1146029695717560320, journalId=1249024232475115590, articleId=1322949806875878170, doi=null, pmid=null, pmcid=null, year=2020, volume=25, issue=4, pageStart=629, pageEnd=654, url=null, language=null, rfNumber=null, rfOrder=110, authorNames=赵永强, 饶元, 董世鹏, 张君毅, journalName=中国图象图形学报, refType=null, unstructuredReference=赵永强,饶元,董世鹏,张君毅. 2020. 深度学习目标检测方法综述. 中国图象图形学报,25(4): 629-654 [DOI: 10.11834/jig.190307], articleTitle=深度学习目标检测方法综述, refAbstract=null), Reference(id=1322949824911385535, tenantId=1146029695717560320, journalId=1249024232475115590, articleId=1322949806875878170, doi=null, pmid=null, pmcid=null, year=2025, volume=null, issue=null, pageStart=1, pageEnd=5, url=null, language=null, rfNumber=null, rfOrder=111, authorNames=Zhu H F, Qu Y, Ma Y T, Pei Y C, journalName=null, refType=null, unstructuredReference=Zhu H F,Qu Y,Ma Y T and Pei Y C. 2025. Evidential deep learning with reweighted margin adjustment for uncertainty-driven cervical OCT image diagnosis//Proceedings of 2025 IEEE International Conference on Acoustics,Speech and Signal Processing (ICASSP). Hyderabad,India: IEEE: 1-5 [DOI: 10.1109/icassp49660.2025.10890637], articleTitle=Evidential deep learning with reweighted margin adjustment for uncertainty-driven cervical OCT image diagnosis, refAbstract=null), Reference(id=1322949824995271616, tenantId=1146029695717560320, journalId=1249024232475115590, articleId=1322949806875878170, doi=null, pmid=null, pmcid=null, year=2023, volume=70, issue=1, pageStart=193, pageEnd=204, url=null, language=null, rfNumber=null, rfOrder=112, authorNames=Zhu T Y, Li K Z, Herrero P, Georgiou P, journalName=IEEE Transactions on Biomedical Engineering, refType=null, unstructuredReference=Zhu T Y,Li K Z,Herrero P and Georgiou P. 2023. Personalized blood glucose prediction for type 1 diabetes using evidential deep learning and meta-learning. IEEE Transactions on Biomedical Engineering,70(1): 193-204 [DOI: 10.1109/tbme.2022.3187703], articleTitle=Personalized blood glucose prediction for type 1 diabetes using evidential deep learning and meta-learning, refAbstract=null), Reference(id=1322949825074963393, tenantId=1146029695717560320, journalId=1249024232475115590, articleId=1322949806875878170, doi=null, pmid=null, pmcid=null, year=2024, volume=96, issue=null, pageStart=null, pageEnd=103214, url=null, language=null, rfNumber=null, rfOrder=113, authorNames=Zou K, Lin T, Han Z B, Wang M, Yuan X D, Chen H Y, journalName=Medical Image Analysis, refType=null, unstructuredReference=Zou K,Lin T,Han Z B,Wang M,Yuan X D,Chen H Y,et al. 2024. Confidence-aware multi-modality learning for eye disease screening. Medical Image Analysis,96: #103214 [DOI: 10.1016/j.media.2024.103214], articleTitle=Confidence-aware multi-modality learning for eye disease screening, refAbstract=null), Reference(id=1322949825150460866, tenantId=1146029695717560320, journalId=1249024232475115590, articleId=1322949806875878170, doi=null, pmid=null, pmcid=null, year=2022, volume=null, issue=null, pageStart=503, pageEnd=513, url=null, language=null, rfNumber=null, rfOrder=114, authorNames=Zou K, Yuan X D, Shen X J, Wang M, Fu H Z, journalName=null, refType=null, unstructuredReference=Zou K,Yuan X D,Shen X J,Wang M and Fu H Z. 2022. TBraTS: trusted brain tumor segmentation//Proceedings of the 25th International Conference on Medical Image Computing and Computer-Assisted Intervention. Singapore,Singapore: Springer: 503-513 [DOI: 10.1007/978-3-031-16452-1_48], articleTitle=TBraTS: trusted brain tumor segmentation, refAbstract=null)], funds=[Fund(id=1322949816262730573, tenantId=1146029695717560320, journalId=1249024232475115590, articleId=1322949806875878170, awardId=null, language=EN, fundingSource=National Natural Science Foundation of China (62472315,62476165), fundOrder=null, country=null), Fund(id=1322949816334033742, tenantId=1146029695717560320, journalId=1249024232475115590, articleId=1322949806875878170, awardId=null, language=CN, fundingSource=国家自然科学基金项目(62472315,62476165), fundOrder=null, country=null)], companyList=[AuthorCompany(id=1322949813762925349, tenantId=1146029695717560320, journalId=1249024232475115590, articleId=1322949806875878170, xref=1, ext=[AuthorCompanyExt(id=1322949813771313958, tenantId=1146029695717560320, journalId=1249024232475115590, articleId=1322949806875878170, companyId=1322949813762925349, language=EN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=1 School of Computer Science and Technology,Tongji University,Shanghai 201804,China), AuthorCompanyExt(id=1322949813783896871, tenantId=1146029695717560320, journalId=1249024232475115590, articleId=1322949806875878170, companyId=1322949813762925349, language=CN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=1 同济大学计算机科学与技术学院,上海 201804)]), AuthorCompany(id=1322949813859394344, tenantId=1146029695717560320, journalId=1249024232475115590, articleId=1322949806875878170, xref=2, ext=[AuthorCompanyExt(id=1322949813867782953, tenantId=1146029695717560320, journalId=1249024232475115590, articleId=1322949806875878170, companyId=1322949813859394344, language=EN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=2 Artificial Intelligence Institute,Shanghai University,Shanghai 200444,China), AuthorCompanyExt(id=1322949813876171562, tenantId=1146029695717560320, journalId=1249024232475115590, articleId=1322949806875878170, companyId=1322949813859394344, language=CN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=2 上海大学人工智能研究院,上海 200444)])], figs=[ArticleFig(id=1322949815717471049, tenantId=1146029695717560320, journalId=1249024232475115590, articleId=1322949806875878170, language=EN, label=Fig.1, caption=Three situations of uncertainty ((a) high certainty;(b) high vacuity; (c) high dissonance), figureFileSmall=jWNxw+cfQkqD8KAG8l2XSQ==, figureFileBig=Zorx+XI8Gc0GYDfvxvEOWA==, tableContent=null), ArticleFig(id=1322949815797162826, tenantId=1146029695717560320, journalId=1249024232475115590, articleId=1322949806875878170, language=CN, label=图1, caption=不确定性的3种情况, figureFileSmall=jWNxw+cfQkqD8KAG8l2XSQ==, figureFileBig=Zorx+XI8Gc0GYDfvxvEOWA==, tableContent=null), ArticleFig(id=1322949815998489419, tenantId=1146029695717560320, journalId=1249024232475115590, articleId=1322949806875878170, language=EN, label=Fig.2, caption=Overview of evidence deep learning methods under complex medical data, figureFileSmall=lUI2lsp/PFtT8SJ4YwfYuw==, figureFileBig=kTKl6s8sqHVMU9OLqG+dYA==, tableContent=null), ArticleFig(id=1322949816078181196, tenantId=1146029695717560320, journalId=1249024232475115590, articleId=1322949806875878170, language=CN, label=图2, caption=复杂医学数据下的证据深度学习方法总览, figureFileSmall=lUI2lsp/PFtT8SJ4YwfYuw==, figureFileBig=kTKl6s8sqHVMU9OLqG+dYA==, tableContent=null)], attaches=null, journal=Journal(id=1249023527618129992, delFlag=0, nameCn=中国图象图形学报, nameEn=Journal of Image and Graphics, nameHistory1=null, nameHistory2=null, issn=1006-8961, eissn=null, cn=11-3758, coden=CODEN ZTTXFZ, periodic=0, language=CN, oaType=null, ccby=null, superviseOffice=null, ownerOffice=null, pubOffice=null, editorOffice=null, officeType=null, aims=null, clcCode=null, officeProv=null, officeCity=null, officeAddr=null, officeZip=null, officeEmail=null, officePhone=null, editDirector=null, officeDirector=null, officeDirectorPhone=null, officeStaffNum=null, officeEmpNum=null, coverPicUrl=uirXtX858YS3zEpFXZttJA==, journalPrice=null, startedYear=null, abbrevIsoEn=Journal of Image and Graphics, journalRemark=null, publicationField=null, createdTime=1775720014721, updatedTime=1788951488513, createdBy=18614031015, updatedBy=13041195026, firstLetterCn=Z, firstLetterEn=Z, subjectCode=Electronic Information, subjectName=null, subjectCodeEn=Electronic Information, subjectNameEn=null, picCn=uirXtX858YS3zEpFXZttJA==, picEn=bud7qaxfvWHeFsbyBTAiKQ==, jcr=null, cjcr=null, exts=[JournalExt(id=1304520351199814396, language=CN, name=中国图象图形学报, nameHistory1=null, nameHistory2=null, managedBy=, sponsoredBy=, publishedBy=, editorOffice=, officeProv=null, officeCity=null, officeAddr=, officeZip=, editDirector=, officeDirector=null, officePhone=null, coverPicUrl=null, journalRemark=, submitArticleUrl=null, websiteUrl=, createdTime=1788951488542, updatedTime=1788951488542, createdBy=13041195026, updatedBy=13041195026, submissionGuidelinesUrl=, submissionAuthorUrl=https://journal.ids.fzyun.cn/auth/realms/journal/protocol/openid-connect/auth?client_id=journal-cjig-author&redirect_uri=https%3A%2F%2Fcjig.portal.founderss.cn%2Foauth%2Fcallback&response_type=code&scope=phone+openid+email+profile&state=e6369def-2842-41d8, submissionEditorUrl=https://journal.ids.fzyun.cn/auth/realms/journal/protocol/openid-connect/auth?client_id=journal-cjig-editor&redirect_uri=https%3A%2F%2Fcjigeditor.portal.founderss.cn%2Foauth%2Fcallback&response_type=code&scope=phone+openid+email+profile&state=9ccec05b-6bd, submissionReviewUrl=https://journal.ids.fzyun.cn/auth/realms/journal/protocol/openid-connect/auth?client_id=journal-cjig-author&redirect_uri=https%3A%2F%2Fcjig.portal.founderss.cn%2Foauth%2Fcallback&response_type=code&scope=phone+openid+email+profile&state=1e8a31c8-5434-4f78, submissionCeEditorUrl=, submissionAeEditorUrl=, option={"copyright":""}), JournalExt(id=1304520351250146045, language=EN, name=Journal of Image and Graphics, nameHistory1=null, nameHistory2=null, managedBy=, sponsoredBy=, publishedBy=, editorOffice=, officeProv=null, officeCity=null, officeAddr=, officeZip=, editDirector=, officeDirector=null, officePhone=null, coverPicUrl=null, journalRemark=, submitArticleUrl=null, websiteUrl=, createdTime=1788951488554, updatedTime=1788951488554, createdBy=13041195026, updatedBy=13041195026, submissionGuidelinesUrl=, submissionAuthorUrl=https://journal.ids.fzyun.cn/auth/realms/journal/protocol/openid-connect/auth?client_id=journal-cjig-author&redirect_uri=https%3A%2F%2Fcjig.portal.founderss.cn%2Foauth%2Fcallback&response_type=code&scope=phone+openid+email+profile&state=e6369def-2842-41d8, submissionEditorUrl=https://journal.ids.fzyun.cn/auth/realms/journal/protocol/openid-connect/auth?client_id=journal-cjig-editor&redirect_uri=https%3A%2F%2Fcjigeditor.portal.founderss.cn%2Foauth%2Fcallback&response_type=code&scope=phone+openid+email+profile&state=9ccec05b-6bd, submissionReviewUrl=https://journal.ids.fzyun.cn/auth/realms/journal/protocol/openid-connect/auth?client_id=journal-cjig-author&redirect_uri=https%3A%2F%2Fcjig.portal.founderss.cn%2Foauth%2Fcallback&response_type=code&scope=phone+openid+email+profile&state=1e8a31c8-5434-4f78, submissionCeEditorUrl=, submissionAeEditorUrl=, option={"copyright":""})], databaseList=null, tenantJournalId=1249024232475115590, websiteList=[Website(id=1249025782459334881, webName=null, webTitle=null, webDomain=null, webCopyrigh=null, webIpcNo=null, seoTitle=null, seoKeywords=null, seoDescription=null, tenantJournalId=null, journalId=1249024232475115590, journalNameCn=null, journalNameEn=null, grayFlag=null, tenantId=1146029695717560320, platformId=null, journalGroupId=null, journalGroupNameCn=null, journalGroupNameEn=null, type=1, domain=https://castjournals.cast.org.cn/joweb/zgtxtxxb/CN, language=CN, createTime=1775720552315, createBy=18614031015, updateTime=1775720586268, updateBy=18614031015, name=中国图象图形学报-中文, tplId=1146099689490845704, title=中国图象图形学报, delFlag=0, indexPage=/home, props=[WebsiteProps(id=1249026166254928133, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1249025782459334881, code=articleTextType, value=kx, createTime=1775720643819, updateTime=1775720643819, creator=18614031015, updator=18614031015), WebsiteProps(id=1249026166221373698, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1249025782459334881, code=banner, value=null, createTime=1775720643811, updateTime=1775720643811, creator=18614031015, updator=18614031015), WebsiteProps(id=1249026166271705352, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1249025782459334881, code=grayFlag, value=0, createTime=1775720643823, updateTime=1775720643823, creator=18614031015, updator=18614031015), WebsiteProps(id=1249026166212985089, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1249025782459334881, code=logo, value=https://castjournals.cast.org.cn/joweb/zgtxtxxb/CN/file/pic?fileId=TDRjKTHfgAnvFKZaDA70wA==, createTime=1775720643809, updateTime=1775720643809, creator=18614031015, updator=18614031015), WebsiteProps(id=1249026166288482570, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1249025782459334881, code=minRunFlag, value=0, createTime=1775720643827, updateTime=1775720643827, creator=18614031015, updator=18614031015), WebsiteProps(id=1249026166246539524, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1249025782459334881, code=picServerUrl, value=https://castjournals.cast.org.cn/joweb/zgtxtxxb/CN/file/pic, createTime=1775720643817, updateTime=1775720643817, creator=18614031015, updator=18614031015), WebsiteProps(id=1249026166280093961, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1249025782459334881, code=silenceFlag, value=0, createTime=1775720643825, updateTime=1775720643825, creator=18614031015, updator=18614031015), WebsiteProps(id=1249026166233956611, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1249025782459334881, code=staticResourcePath, value=https://castjournals.cast.org.cn/joweb/cast_kjdb_cn_619/, createTime=1775720643814, updateTime=1775720643814, creator=18614031015, updator=18614031015), WebsiteProps(id=1249026166259122438, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1249025782459334881, code=themeColor, value=null, createTime=1775720643820, updateTime=1775720643820, creator=18614031015, updator=18614031015), WebsiteProps(id=1249026166267511047, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1249025782459334881, code=themeStyle, value=null, createTime=1775720643822, updateTime=1775720643822, creator=18614031015, updator=18614031015)]), Website(id=1249025782681633001, webName=null, webTitle=null, webDomain=null, webCopyrigh=null, webIpcNo=null, seoTitle=null, seoKeywords=null, seoDescription=null, tenantJournalId=null, journalId=1249024232475115590, journalNameCn=null, journalNameEn=null, grayFlag=null, tenantId=1146029695717560320, platformId=null, journalGroupId=null, journalGroupNameCn=null, journalGroupNameEn=null, type=1, domain=https://castjournals.cast.org.cn/joweb/zgtxtxxb/EN, language=EN, createTime=1775720552368, createBy=18614031015, updateTime=1775720607118, updateBy=18614031015, name=中国图象图形学报-英文, tplId=1146101810881728533, title=Journal of Image and Graphics, delFlag=0, indexPage=/home, props=[WebsiteProps(id=1249026195371786511, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1249025782681633001, code=articleTextType, value=kx, createTime=1775720650761, updateTime=1775720650761, creator=18614031015, updator=18614031015), WebsiteProps(id=1249026195355009292, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1249025782681633001, code=banner, value=null, createTime=1775720650757, updateTime=1775720650757, creator=18614031015, updator=18614031015), WebsiteProps(id=1249026195392758034, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1249025782681633001, code=grayFlag, value=0, createTime=1775720650766, updateTime=1775720650766, creator=18614031015, updator=18614031015), WebsiteProps(id=1249026195342426379, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1249025782681633001, code=logo, value=https://castjournals.cast.org.cn/joweb/zgtxtxxb/EN/file/pic?fileId=TDRjKTHfgAnvFKZaDA70wA==, createTime=1775720650754, updateTime=1775720650754, creator=18614031015, updator=18614031015), WebsiteProps(id=1249026195409535252, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1249025782681633001, code=minRunFlag, value=0, createTime=1775720650770, updateTime=1775720650770, creator=18614031015, updator=18614031015), WebsiteProps(id=1249026195367592206, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1249025782681633001, code=picServerUrl, value=https://castjournals.cast.org.cn/joweb/zgtxtxxb/EN/file/pic, createTime=1775720650760, updateTime=1775720650760, creator=18614031015, updator=18614031015), WebsiteProps(id=1249026195401146643, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1249025782681633001, code=silenceFlag, value=0, createTime=1775720650768, updateTime=1775720650768, creator=18614031015, updator=18614031015), WebsiteProps(id=1249026195359203597, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1249025782681633001, code=staticResourcePath, value=https://castjournals.cast.org.cn/joweb/cast_kjdb_en_623/, createTime=1775720650758, updateTime=1775720650758, creator=18614031015, updator=18614031015), WebsiteProps(id=1249026195380175120, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1249025782681633001, code=themeColor, value=null, createTime=1775720650763, updateTime=1775720650763, creator=18614031015, updator=18614031015), WebsiteProps(id=1249026195388563729, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1249025782681633001, code=themeStyle, value=null, createTime=1775720650765, updateTime=1775720650765, creator=18614031015, updator=18614031015)])], journalTitle=中国图象图形学报, weixinUrl=null, journalUrl=https://www.cjig.cn/, iacademicId=null, status=1, seqNo=null, journalTitleEn=Journal of Image and Graphics, journalPhotoCn=uirXtX858YS3zEpFXZttJA==, journalPhotoEn=bud7qaxfvWHeFsbyBTAiKQ==, journalFirstLetter=Z, journalRecommend=null, journalNew=null, journalCollection=null, jcrJf=null, cjcrJf=null, jcrJfStr=null, cjcrJfStr=null, submissionFirstDecision=null, sciSubjectClassification=null, casSubjectClassification=null, citeScore=null, totalCitationFrequency=null, icpCode=null, psCode=null, advertisingLicenseCode=null, copyrightInformation=null, country=null, option=, provinceCode=null, provinceName=null, collectFlag=false, interPubPlatform=, interPubPlatformUrl=null), detailUrlCn=https://castjournals.cast.org.cn/joweb/zgtxtxxb/CN/10.11834/jig.250475, detailUrlEn=https://castjournals.cast.org.cn/joweb/zgtxtxxb/EN/10.11834/jig.250475, pdfUrlCn=https://castjournals.cast.org.cn/joweb/zgtxtxxb/CN/PDF/10.11834/jig.250475, pdfUrlEn=https://castjournals.cast.org.cn/joweb/zgtxtxxb/EN/PDF/10.11834/jig.250475, aliStartDate=0, aliEndDate=0, collectionFlag=false, citedCount=null, citedUrl=null, previewStatus=0, delFlag=0, hasFullText=1, orderTime=1784131200000, fullTextJson=null, articleText=null, reference=null)
收藏切换
应用证据深度学习的可信医学影像分析方法综述
收藏切换
PDF下载
文昕颢 1 , 刘伟 1 , 岳晓冬 2 , 陈宇飞 1, *
中国图象图形学报 | 面向医学影像智能分析的基础模型 2026,31(7): 2505-2525
收起
收藏切换
中国图象图形学报 |面向医学影像智能分析的基础模型 2026 , 31 (7) : 2505 -2525
应用证据深度学习的可信医学影像分析方法综述
全屏
[Author(id=1322949813939086124, tenantId=1146029695717560320, journalId=1249024232475115590, articleId=1322949806875878170, orderNo=0, firstName=null, middleName=null, lastName=null, nameCn=null, orcid=null, stid=null, country=null, authorPic=null, dead=0, email=wenxinhao@tongji.edu.cn, emailSecond=null, emailThird=null, correspondingAuthor=0, authorType=1, ext={EN=AuthorExt(id=1322949814018777902, tenantId=1146029695717560320, journalId=1249024232475115590, articleId=1322949806875878170, authorId=1322949813939086124, language=EN, stringName=Xinhao Wen, firstName=Xinhao, middleName=null, lastName=Wen, prefix=null, suffix=null, authorComment=null, nameInitials=null, affiliation=null, department=null, xref=1, address=1 School of Computer Science and Technology,Tongji University,Shanghai 201804,China, bio=null, bioImg=null, bioContent=null, aboutCorrespAuthor=null), CN=AuthorExt(id=1322949814081692463, tenantId=1146029695717560320, journalId=1249024232475115590, articleId=1322949806875878170, authorId=1322949813939086124, language=CN, stringName=文昕颢, firstName=null, middleName=null, lastName=null, prefix=null, suffix=null, authorComment=null, nameInitials=null, affiliation=null, department=null, xref=1, address=1 同济大学计算机科学与技术学院,上海 201804, bio={"content":"

文昕颢,男,硕士研究生,主要研究方向为医学图像分析。E-mail:

"}, bioImg=null, bioContent=

文昕颢,男,硕士研究生,主要研究方向为医学图像分析。E-mail:

, aboutCorrespAuthor=null)}, companyList=[AuthorCompany(id=1322949813762925349, tenantId=1146029695717560320, journalId=1249024232475115590, articleId=1322949806875878170, xref=1, ext=[AuthorCompanyExt(id=1322949813771313958, tenantId=1146029695717560320, journalId=1249024232475115590, articleId=1322949806875878170, companyId=1322949813762925349, language=EN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=1 School of Computer Science and Technology,Tongji University,Shanghai 201804,China), AuthorCompanyExt(id=1322949813783896871, tenantId=1146029695717560320, journalId=1249024232475115590, articleId=1322949806875878170, companyId=1322949813762925349, language=CN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=1 同济大学计算机科学与技术学院,上海 201804)])]), Author(id=1322949814144607025, tenantId=1146029695717560320, journalId=1249024232475115590, articleId=1322949806875878170, orderNo=1, firstName=null, middleName=null, lastName=null, nameCn=null, orcid=null, stid=null, country=null, authorPic=null, dead=0, email=ldachuan@outlook.com, emailSecond=null, emailThird=null, correspondingAuthor=0, authorType=1, ext={EN=AuthorExt(id=1322949814241076019, tenantId=1146029695717560320, journalId=1249024232475115590, articleId=1322949806875878170, authorId=1322949814144607025, language=EN, stringName=Wei Liu, firstName=Wei, middleName=null, lastName=Liu, prefix=null, suffix=null, authorComment=null, nameInitials=null, affiliation=null, department=null, xref=1, address=1 School of Computer Science and Technology,Tongji University,Shanghai 201804,China, bio=null, bioImg=null, bioContent=null, aboutCorrespAuthor=null), CN=AuthorExt(id=1322949814320767796, tenantId=1146029695717560320, journalId=1249024232475115590, articleId=1322949806875878170, authorId=1322949814144607025, language=CN, stringName=刘伟, firstName=null, middleName=null, lastName=null, prefix=null, suffix=null, authorComment=null, nameInitials=null, affiliation=null, department=null, xref=1, address=1 同济大学计算机科学与技术学院,上海 201804, bio={"content":"

刘伟,男,博士研究生,主要研究方向为多视角深度学习、不确定性估计和医学图像分析。E-mail:

"}, bioImg=null, bioContent=

刘伟,男,博士研究生,主要研究方向为多视角深度学习、不确定性估计和医学图像分析。E-mail:

, aboutCorrespAuthor=null)}, companyList=[AuthorCompany(id=1322949813762925349, tenantId=1146029695717560320, journalId=1249024232475115590, articleId=1322949806875878170, xref=1, ext=[AuthorCompanyExt(id=1322949813771313958, tenantId=1146029695717560320, journalId=1249024232475115590, articleId=1322949806875878170, companyId=1322949813762925349, language=EN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=1 School of Computer Science and Technology,Tongji University,Shanghai 201804,China), AuthorCompanyExt(id=1322949813783896871, tenantId=1146029695717560320, journalId=1249024232475115590, articleId=1322949806875878170, companyId=1322949813762925349, language=CN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=1 同济大学计算机科学与技术学院,上海 201804)])]), Author(id=1322949814400459574, tenantId=1146029695717560320, journalId=1249024232475115590, articleId=1322949806875878170, orderNo=2, firstName=null, middleName=null, lastName=null, nameCn=null, orcid=null, stid=null, country=null, authorPic=null, dead=0, email=yswantfly@shu.edu.cn, emailSecond=null, emailThird=null, correspondingAuthor=0, authorType=1, ext={EN=AuthorExt(id=1322949814484345656, tenantId=1146029695717560320, journalId=1249024232475115590, articleId=1322949806875878170, authorId=1322949814400459574, language=EN, stringName=Xiaodong Yue, firstName=Xiaodong, middleName=null, lastName=Yue, prefix=null, suffix=null, authorComment=null, nameInitials=null, affiliation=null, department=null, xref=2, address=2 Artificial Intelligence Institute,Shanghai University,Shanghai 200444,China, bio=null, bioImg=null, bioContent=null, aboutCorrespAuthor=null), CN=AuthorExt(id=1322949814559843129, tenantId=1146029695717560320, journalId=1249024232475115590, articleId=1322949806875878170, authorId=1322949814400459574, language=CN, stringName=岳晓冬, firstName=null, middleName=null, lastName=null, prefix=null, suffix=null, authorComment=null, nameInitials=null, affiliation=null, department=null, xref=2, address=2 上海大学人工智能研究院,上海 200444, bio={"content":"

岳晓冬,男,教授,主要研究方向为机器学习的理论与应用。E-mail:

"}, bioImg=null, bioContent=

岳晓冬,男,教授,主要研究方向为机器学习的理论与应用。E-mail:

, aboutCorrespAuthor=null)}, companyList=[AuthorCompany(id=1322949813859394344, tenantId=1146029695717560320, journalId=1249024232475115590, articleId=1322949806875878170, xref=2, ext=[AuthorCompanyExt(id=1322949813867782953, tenantId=1146029695717560320, journalId=1249024232475115590, articleId=1322949806875878170, companyId=1322949813859394344, language=EN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=2 Artificial Intelligence Institute,Shanghai University,Shanghai 200444,China), AuthorCompanyExt(id=1322949813876171562, tenantId=1146029695717560320, journalId=1249024232475115590, articleId=1322949806875878170, companyId=1322949813859394344, language=CN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=2 上海大学人工智能研究院,上海 200444)])]), Author(id=1322949814639534907, tenantId=1146029695717560320, journalId=1249024232475115590, articleId=1322949806875878170, orderNo=3, firstName=null, middleName=null, lastName=null, nameCn=null, orcid=null, stid=null, country=null, authorPic=null, dead=0, email=yufeichen@tongji.edu.cn, emailSecond=null, emailThird=null, correspondingAuthor=1, authorType=1, ext={EN=AuthorExt(id=1322949814740198205, tenantId=1146029695717560320, journalId=1249024232475115590, articleId=1322949806875878170, authorId=1322949814639534907, language=EN, stringName=Yufei Chen, firstName=Yufei, middleName=null, lastName=Chen, prefix=null, suffix=null, authorComment=null, nameInitials=null, affiliation=null, department=null, xref=1, *, address=1 School of Computer Science and Technology,Tongji University,Shanghai 201804,China, bio=null, bioImg=null, bioContent=null, aboutCorrespAuthor=null), CN=AuthorExt(id=1322949814811501374, tenantId=1146029695717560320, journalId=1249024232475115590, articleId=1322949806875878170, authorId=1322949814639534907, language=CN, stringName=陈宇飞, firstName=null, middleName=null, lastName=null, prefix=null, suffix=null, authorComment=null, nameInitials=null, affiliation=null, department=null, xref=1, *, address=1 同济大学计算机科学与技术学院,上海 201804, bio={"content":"

陈宇飞,通信作者,女,副教授,主要研究方向为医学图像分析、多模态学习和不确定性估计。E-mail:

"}, bioImg=null, bioContent=

陈宇飞,通信作者,女,副教授,主要研究方向为医学图像分析、多模态学习和不确定性估计。E-mail:

, aboutCorrespAuthor=null)}, companyList=[AuthorCompany(id=1322949813762925349, tenantId=1146029695717560320, journalId=1249024232475115590, articleId=1322949806875878170, xref=1, ext=[AuthorCompanyExt(id=1322949813771313958, tenantId=1146029695717560320, journalId=1249024232475115590, articleId=1322949806875878170, companyId=1322949813762925349, language=EN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=1 School of Computer Science and Technology,Tongji University,Shanghai 201804,China), AuthorCompanyExt(id=1322949813783896871, tenantId=1146029695717560320, journalId=1249024232475115590, articleId=1322949806875878170, companyId=1322949813762925349, language=CN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=1 同济大学计算机科学与技术学院,上海 201804)])])]
文昕颢1 , 刘伟1 , 岳晓冬2 , 陈宇飞1, *
作者信息
  • 1 同济大学计算机科学与技术学院,上海 201804
  • 2 上海大学人工智能研究院,上海 200444
通讯作者:
作者简介:

文昕颢,男,硕士研究生,主要研究方向为医学图像分析。E-mail:

刘伟,男,博士研究生,主要研究方向为多视角深度学习、不确定性估计和医学图像分析。E-mail:

岳晓冬,男,教授,主要研究方向为机器学习的理论与应用。E-mail:

陈宇飞,通信作者,女,副教授,主要研究方向为医学图像分析、多模态学习和不确定性估计。E-mail:

Survey of trustworthy medical image analysis methods on the basis of evidential deep learning
Xinhao Wen1 , Wei Liu1 , Xiaodong Yue2 , Yufei Chen1, *
Affiliations
  • 1 School of Computer Science and Technology,Tongji University,Shanghai 201804,China
  • 2 Artificial Intelligence Institute,Shanghai University,Shanghai 200444,China
出版时间: 2026-07-16 doi: 10.11834/jig.250475
文章导航
收藏切换

随着人工智能技术在医疗领域的深度融合,基于深度学习的影像分析模型已在多项任务中展现出显著潜力。然而,这类模型的性能严重依赖于大规模、高质量的训练数据,而真实临床环境下的医学数据常面临三大类挑战:数据稀缺、数据异质性以及数据不均衡问题。传统深度学习模型在处理此类复杂数据时,往往表现出过度自信的错误预测,且缺乏对认知不确定性的表达能力,从而为高风险的临床决策带来潜在安全威胁。证据深度学习(evidential deep learning,EDL)作为一种新兴的不确定性量化范式,通过将狄利克雷分布引入到神经网络的输出对预测进行证据建模,能够一定程度上建模偶然不确定性和认知不确定性,这种能力为应对真实复杂医学数据提供了强有力的工具。本文从数据稀缺、数据异质性及数据不均衡这三大挑战的视角切入,系统性地综述了证据深度学习在复杂医学数据下的影像分析应用,并阐明了证据深度学习凭借其不确定性量化和内在的可解释性,在利用无标签数据、检测异常样本、解析并融合冲突证据、识别尾部数据以及仲裁噪声数据等方面展现的核心优势。最后,本文对该领域当前面临的挑战进行了总结,并对未来的研究方向进行了展望,为研究者们提供了一个清晰的路线图,以推动构建更安全、更可靠、更可信的下一代智能医学影像分析系统。本文提及的相关算法已汇总至https://github.com/oscarab/Medical-EDL。

证据深度学习(EDL)  /  医学影像分析  /  不确定性量化  /  主观逻辑  /  证据理论

The integration of deep learning into medical diagnostics has heralded a new era of data-driven healthcare,demonstrating profound potential in tasks ranging from radiological image interpretation and pathological slide analysis to genomic sequencing and clinical outcome prediction. These advanced computational models have achieved,and in some cases even surpassed,human-level performance on various well-defined tasks,promising to enhance diagnostic accuracy,streamline clinical workflows,and facilitate personalized medicine. However,the remarkable success of these models is fundamentally predicated on an implicit assumption: the availability of large-scale,high-quality,comprehensively annotated datasets. This assumption is in stark contrast with the reality of clinical data,which are intrinsically complex and challenging. Real-world medical data are typically characterized by scarcity,heterogeneity,and imbalance. The deployment of conventional deep learning models,which often behave as black boxes in such complex data environments,poses critical risks. These models are notoriously prone to producing overconfident yet erroneous predictions and lack a mechanism to express their uncertainty or to reliably signal when they are operating outside their domain of expertise. In the high-stake context of clinical decision-making,where a single incorrect prediction can have dire effects on patient safety,this lack of trustworthy uncertainty awareness is a critical barrier to the responsible adoption of artificial intelligence (AI). This review is motivated by the urgent need to address the challenge of real-world complex data in medical AI. We systematically structure the landscape of data complexities into three principal categories,which form the organizational framework of our analysis. The first one is data scarcity,which is common in rare diseases,underrepresented populations,and emerging pathologies; it leads to overfitting and poor generalization. The second one is data heterogeneity,which includes label noise,feature noise,multimodal fusion conflicts,and out-of-distribution (OOD) samples. The third one is data imbalance,where the dominance of normal over abnormal cases biases models toward the majority class,undermining their sensitivity for rare pathologies. A new paradigm—one that goes beyond mere prediction to incorporate principled uncertainty quantification—is required to address these challenges. This review centers on evidential deep learning (EDL),an emergent and powerful framework designed for this purpose. Unlike traditional models that output a single point-estimate probability via a softmax function,EDL reframes classification by placing a high-order probability distribution over the categorical distribution of classes. The core contribution of this review is that it provides a systematic,in-depth analysis of how the EDL framework and its associated uncertainty metrics can be specifically utilized to mitigate the problems arising from real-world complex medical data. We meticulously map EDL-based solutions to our three-pronged taxonomy of data challenges. For data scarcity,EDL models naturally express high epistemic uncertainty when making predictions for classes with insufficient training evidence,thereby providing a quantitative and actionable signal that the model’s output should be treated with caution. For data heterogeneity,EDL offers robust solutions. In cases of data noise and multimodal conflicts,the conflicting or low-quality evidence is reflected in the parameters of the learned Dirichlet distribution,leading to high overall uncertainty and preventing an overconfident but incorrect conclusion. EDL is particularly effective for OOD detection; an OOD sample,by definition,lacks supporting evidence in the training data,causing a well-trained EDL model to produce a prediction with high epistemic uncertainty and a“vacuous”evidence vector. This situation enables the model to effectively“abstain”from making a high-risk prediction and instead refer the case to a human expert. In the context of data imbalance,although not a direct remedy,EDL provides a more nuanced view than standard outputs. It can reveal when a model is uncertain about its predictions for minority classes,even if the predicted probability is low,thus helping identify instances that require further scrutiny. This study,therefore,provides a comprehensive and structured survey of the current state of the art in applying EDL to medical image analysis under real-world complex data conditions. We synthesize the disparate literature into a coherent conceptual framework,elucidating the underlying principles,core methodologies,and practical applications of EDL in this critical domain. In conclusion,we posit that evidential deep learning represents a pivotal step toward building safe,reliable,and trustworthy AI systems for medical applications. However,the field is still nascent and faces several challenges that open up avenues for future research. These challenges include the need for robust theoretical foundations for evidence modeling,improving the computational scalability of EDL methods,and developing techniques for effective uncertainty calibration to ensure that the quantified uncertainty meaningfully corresponds to the true likelihood of error. The clinical translatability of uncertainty needs to be explored,with focus on how to present uncertainty information to clinicians in an intuitive and actionable manner. Moreover,large-scale,prospective clinical validation studies are imperative to move beyond retrospective dataset benchmarks and rigorously assess the real-world effect and safety of EDL-based diagnostic systems. This review aims to serve as a valuable resource and roadmap,inspiring further research to bridge the gap between the theoretical promise of EDL and its practical deployment in future clinical workflows. The codes link: https://github.com/oscarab/Medical-EDL.

evidential deep learning(EDL)  /  medical image analysis  /  uncertainty quantification  /  subjective logic  /  evidence theory
文昕颢, 刘伟, 岳晓冬, 陈宇飞. 应用证据深度学习的可信医学影像分析方法综述. 中国图象图形学报, 2026 , 31 (7) : 2505 -2525 . DOI: 10.11834/jig.250475
Xinhao Wen, Wei Liu, Xiaodong Yue, Yufei Chen. Survey of trustworthy medical image analysis methods on the basis of evidential deep learning[J]. Journal of Image and Graphics, 2026 , 31 (7) : 2505 -2525 . DOI: 10.11834/jig.250475
近年来,深度学习技术在众多领域展现出强大的能力,甚至在一些任务上达到或超越了人类专家的水平,例如自然图像目标检测、图像分割等(Noor和Ige,2025;Pacal等,2024;Sharifani和Amini,2023;赵永强 等,2020)。尽管深度学习模型取得了诸多令人瞩目的突破性成果,但将其应用于实际医疗实践时,其诊断准确性与泛化性能往往会出现显著的衰减(Rayed等,2024;Suganyadevi等,2022)。这种性能落差的核心根源在于理想化的实验室环境与复杂多变的真实临床环境之间存在的巨大鸿沟。医学数据与标准计算机视觉或自然语言处理等常见数据有着本质区别,其本身固有的一系列独特且复杂的特性,如数据稀缺性、数据异质性以及数据不均衡等,相互交织、相互影响,构成了当前医学人工智能从技术验证迈向大规模临床落地的核心瓶颈,也催生了新型方法论体系以满足医学实际需求。
具体而言,复杂医学数据的第1个特点是数据稀缺,其根源可追溯至医学数据获取的双重壁垒(Jiao等,2024)。客观层面,特定疾病的流行病学特征决定了其自然发生率存在显著差异,导致某些罕见病或早期病变的样本获取较为困难;主观层面,高质量医学标注依赖于深厚的专业知识,这使得标注成本极为高昂,在需要精细化像素级、区域级标注的医学影像分析任务中尤为明显。值得注意的是,医学数据的严重稀缺性在半监督学习和主动学习场景中会表现出独特的挑战:一旦标注样本数量降至某个关键水平以下,算法性能会出现急剧的、非线性的下降(Han等,2024;Long等,2024)。这种现象深刻反映了医学数据在极低标注量下的学习行为与普通数据存在根本差异。
数据异质性是复杂医学数据的第2个特点,其核心在于数据从训练到测试阶段均存在显著分布差异和内在不一致性。首先,训练数据中可能包含噪声,这源于成像伪影和设备噪声带来的像素级扰动,也源于诊断标注过程固有的主观性差异,更棘手的是,这类噪声使得模型难以有效滤除干扰并捕获关键特征(Penso等,2024)。此外,当整合多模态数据如计算机断层扫描(computed tomography,CT)、磁共振成像(magnetic resonance imaging,MRI)和正电子发射断层扫描(positron emission tomography,PET),或多视角信息进行综合决策时,不同来源或视角间可能蕴含相互补充或矛盾的信息(Pei等,2023),这种多模态冲突不仅加剧了信息的不确定性,也形成了一种特殊的、训练数据内部规律不一致的挑战,要求模型具备处理不一致性并权衡信源可靠性的能力。不仅如此,模型常常会在测试数据中遇到在训练阶段从未接触过的样本。这些分布外样本可根据其偏离程度进一步区分:近分布外样本(与已知疾病模式高度相似,如已知病种的细微亚型变异或非典型表现)和远分布外样本(与已知疾病模式显著不同,典型的例子包括全新的疾病类型,如新型病原体感染或极其罕见的并发症形态)。无论近分布外还是远分布外样本,其根本特征都是显著偏离了模型的原始训练数据分布,对模型的认知边界构成了直接挑战(Karimi和Gholipour,2023)。
数据不均衡是复杂医学数据的第3类特点,其核心在于数据中存在多层次的结构性不平衡,这种不平衡同时贯穿于疾病类别与患者群体之间,并直接引发模型预测的公平性与伦理问题。在疾病类别上,不同疾病的自然发生率差异显著,常见病(如肺癌或高血压慢性病)往往拥有大量样本,而罕见病(如罕见的胰腺神经内分泌肿瘤)样本往往极其有限,训练数据中不同类别可能相差数个数量级,形成严重的类别不平衡。进一步地,不同模态间的维度也存在显著不均衡,例如指标数据的特征维度通常低于影像等高维模态,这种差异加剧了多模态融合的难度。此外,数据在群体维度上也存在代表性不均,例如年龄、性别、种族、地域来源等群体在特定疾病数据集中比例失调,可能源于现实人群结构、医疗可及性或数据采集策略的限制。深度学习模型在面对这种类别与组间双重不平衡的数据时,会倾向于过度拟合主导类别与主流属性群体(Ding等,2024),导致对少数类别和代表不足群体的判别性能显著下降:不仅难以准确识别罕见疾病,也无法在不同人口学群体中保持一致的预测效能,从而加剧健康公平性问题,并带来潜在的伦理风险。
数据稀缺、异质性与不均衡这三大挑战往往并非孤立存在,而是在复杂医学场景中同时出现。例如数据不均衡问题常与多模态融合的挑战相互交织,突出表现为常见病与罕见病在模态数量上的显著差异(Chen等,2025b)。面对这种多维交织的复杂特性,传统以医生个人经验为核心的决策模式在效率、一致性以及对罕见或非典型病例的处理能力上逐渐显现出局限性。随着计算能力的提升和大规模医学数据集的积累,数据驱动的决策模式,特别是基于深度学习等人工智能技术的方法,展现出巨大潜力。这类方法分析海量的医学数据自动学习复杂模式,在提升效率和特定任务性能上展现出优势(Cai等,2020;Li等,2022;Liu等,2023b;Lei等,2023)。然而,医学数据中固有的数据稀缺性、数据异质性和数据不均衡三大特性,严重制约了其应用。深度学习模型“黑箱”特性导致的解释性缺失及难以处理多源数据冲突,也削弱了决策的可信度与临床接受度(Takahashi等,2024)。在此背景下,医学决策范式正逐步由经验驱动、数据驱动,向证据驱动演进(Khosravi等,2022;Xu等,2022b;Ghesu等,2019)。证据驱动模式以量化可信的医学证据为基础,结合多源数据深度融合和不确定性量化,为复杂医学情境下的高可信决策提供了理论支撑和实践工具。
具体而言,证据驱动的医学决策范式以证据深度学习(evidential deep learning,EDL)(Sensoy等,2018)为核心技术。EDL中的证据是指证据神经网络从输入数据中收集的支持分类决策的信息强度,进而,EDL利用狄利克雷分布对模型输出进行参数化,为模型预测提供了显式的不确定性量化能力。这种量化的不确定性,正是统一应对医学数据三大核心挑战——数据稀缺、数据异质性与数据不均衡——的关键所在。例如,模型可利用不确定性指导半监督或主动学习以应对数据稀缺;识别分布外样本或仲裁多模态数据冲突来处理数据异质性;以及对样本进行重加权或指导样本合成来缓解数据不均衡的影响。在此基础上,EDL产生的证据(即其狄利克雷参数可直接转换为“基本信度分配”)还能与经典的D-S(Dempster-Shafer)证据理论(Shafer,1976)无缝结合。这为融合不同来源(如影像与病历)的医学证据,处理多源信息冲突,提供了数学上更为严谨的推理框架。总而言之,EDL不仅通过其证据强度反映了知识不足(源于稀缺)与识别困难(源于异质),其量化的不确定性也为解决数据不均衡问题提供了新的思路,这种统一的解决能力正是医疗场景所亟需的。
本综述系统梳理了证据深度学习方法在真实医学复杂数据场景下的应用进展。第1节介绍各种不确定性来源,并阐明主观逻辑与证据理论的理论基础;第2、3、4节依据医学数据特点阐述代表性工作;最后讨论当前局限与发展趋势,为未来研究提供方法论指导。
当前研究普遍将不确定性来源划分为两类:认知不确定性(epistemic uncertainty,EU)与偶然不确定性(aleatoric uncertainty,AU)(Der Kiureghian和Ditlevsen,2009;Kendall和Gal,2017;Guo等,2024)。
认知不确定性亦称模型不确定性,其核心在于反映了模型自身认知状态的局限,源于对潜在数据生成机制或特定输入相关知识储备的不足,通常由训练数据的有限性、不完整性或代表性偏差所引发。当模型面对与训练数据分布显著偏离的样本,即分布外(out-of-distribution,OOD)样本进行预测时,往往会表现出高认知不确定性,这清晰地表明模型缺乏足够的知识或信息对此输入做出高置信度的可靠推断。从本质上讲,认知不确定性是可减少的,其缓解策略主要包括持续扩充高质量且具代表性的训练数据、注入领域先验知识、优化模型架构以增强其表达能力等。在度量方法上,蒙特卡罗方法(Monte-Carlo dropout,MC dropout)(Gal和Ghahramani,2016)通过在测试阶段对带有Dropout层的模型进行多次前向传播采样,并计算预测结果的方差(或熵)来近似模型的不确定性。深度集成(Lakshminarayanan等,2017)则通过训练多个独立初始化的模型并通过计算预测均值或方差汇总其预测,能更有效地捕捉模型假设空间的不确定性,通常比单一贝叶斯近似方法效果更优。此外,贝叶斯神经网络(Bayesian neural network,BNN)(Blundell等,2015)通过为网络权重引入概率分布,理论上能提供更完备的不确定性估计。现有的工作认为证据深度学习可以通过一次前向计算直接建模认知不确定性(Chen等,2023a)。有效降低认知不确定性意味着拓展了模型的知识边界和对数据分布的覆盖范围,是提升模型在未知或边缘场景下鲁棒性的关键。
偶然不确定性亦称数据不确定性,是内生于数据自身固有的、不可约减的随机性,其产生与模型的知识状态无关。其来源多样,包括类间视觉相似性导致的判别模糊、传感器误差等观测噪声,以及标签噪声。这种不确定性的核心属性在于,它限制了模型预测精度的理论上限,通常无法通过增加同源数据或优化模型来消除。但值得注意的是,通过融合多源互补信息(例如,结合影像与病理报告),可以有效澄清单一数据源中的固有模糊性,是降低偶然不确定性的有效途径。因此,除采用多源融合策略外,常规应对思路并非试图消除不确定性,而是转向对数据固有随机性的直接建模。在度量上,回归任务通常让模型同时预测目标值及其方差,通过学习一个输入相关的噪声参数量化偶然不确定性(Amini等,2020);分类任务则通过输出的概率分布熵或置信度反映数据固有的模糊性。此外,置信度校准技术是确保这些概率度量可靠性的关键。准确量化偶然不确定性,是理解和传达模型预测可靠性边界的基础。
作为贝叶斯概率理论的进阶框架,主观逻辑(subjective logic,SL)(Jøsang,2016)在传统概率逻辑基础上进行了关键性扩展。经典概率逻辑将命题真值定义为连续概率值(取值区间 [0,1]),概率总和恒定为1,无法在不确定的场景下显式地回答“我不知道”。主观逻辑则进一步显式引入了不确定性质量,对于在定义域
𝕏
分类随机变量
X
,主观意见可形式化为有序三元组
τ=b,u,a
,其中
b
为
X
上的信念质量分布,
u
为不确定性质量,
a
为先验概率(先验分布),而信念质量和不确定性质量的总和始终为1。当定义域为二元时(
𝕏=2
),该意见称为二元主观意见,多类别情形(
𝕏>2
)则称为多元主观意见。
根据主观逻辑的形式化定义,不确定性质量
u
本质可解释为向整个定义域
𝕏
分配的置信质量。根据先验概率
a
,该不确定性质量可按概率比例重分配到各具体类别中,从而导出具有可加性的投影概率:
Pi=bi+aiu
。其中
i
为类别索引且满足概率总和为1,该投影概率构建了主观逻辑与经典概率的衔接桥梁。基于上述定义,多元主观意见与狄利克雷概率密度函数存在严格的双射对应关系。形式化表述如下,设
p
为定义域
𝕏
上的概率分布,参数为
α=α1,α2,⋅⋅⋅,αK
的狄利克雷分布为
Dirp,α
。若给定一个先验分布
a
,则存在主观意见
τ
和狄利克雷分布的双射映射
F:τ→Dirp,α
,参数转换关系满足:
αi=biWu+αiW
,其中,
W
为先验权重,这种关系通过概率密度函数对二阶不确定性进行建模。正因为如此,利用主观逻辑建立的证据深度学习可以很好地估计不确定度,从而有了表达“我不知道”的能力。主观逻辑中主要关注捕获两种类型的不确定性:空缺不确定性(vacuity)和冲突不确定性(dissonance)(Jøsang等,2018),它们分别对应认知不确定性和偶然不确定性,空缺不确定性的定义为
Vacα=KS
式中,
α
是狄利克雷分布参数,
K
是类别总数,
S
为狄利克雷参数的总和。当模型无法获得足够证据时,较高的空缺度表示决策时缺乏证据,这表明样本可能是训练分布外。
冲突不确定性的定义为
Dissα=∑k=1K bk∑j≠k (bjBal(bj, bk))∑j≠k bj
式中,
bk
表示类别
k
的信念质量,
Bal(⋅)
为相对信念质量平衡函数。冲突不确定度是基于分配给不同类别的信念之间的冲突程度进行衡量的。
图1展示了一个三分类问题中不同证据量下的狄利克雷分布,红、绿和蓝色代表3个类别,其他的是训练分布外的样本,这里分别展示了不确定性度量的3种情况。高确定性对应于某类证据充足,模型可做出可靠决策;高空缺不确定性源于所有类别的证据均匮乏,这常对应于模型因信息不足而对分布外样本的无知;高冲突不确定性则产生于类别间证据的强烈冲突,这反映了模型对类间模糊样本的困惑。
D-S证据理论最初由Dempster在统计推断领域奠基,后经Shafer发展为表征认知不确定性的系统框架。该理论通过突破传统概率论的可加性,即允许互斥事件的概率总和小于1,为不确定性建模开辟了新路径。证据理论与主观逻辑的核心共性在于二者均能通过向定义域全域分配置信质量显式表征概率不确定性,但本质差异在于:主观逻辑建立了先验概率
a
作为先验分布的调节器,强制高不确定性样本的证据分布回撤至预设先验;而经典证据理论框架并不包含此类先验概率机制。
证据深度学习将主观逻辑应用到深度神经网络中,使神经网络可以在输出样本信念质量的同时给出不确定性质量(Sensoy等,2018),其已经有一系列证据神经网络的变种如I-EDL(Fisher information-based EDL)(Deng等,2023)、R-EDL(relaxed-EDL)(Chen等,2024b)、HEDL(hyper-opinion EDL)(Qu等,2024)等,并已经广泛应用于各个领域(Liu等,2023a,2025a,b;Liang等,2025a,b)。在包含
K
类的分类任务中,对于给定的样本
x
,证据神经网络可以给出对应每个类别的证据量
e=e1,⋯,eK
,每个证据
ei
表示支持“样本属于类别
i
”的程度,相对于传统的深度神经网络,证据神经网络会使用Softplus激活函数作为softmax的替代以产生非负的证据
e=Softplusfx
。
基于上述定义,样本
x
的信念质量和不确定性质量可以由证据
e
计算得出,具体为
bi=eiS, u=WS
式中,
ei
为类别
i
的证据量。根据建立的主观意见和狄利克雷分布之间的双射,狄利克雷分布参数可用证据
e
表示为
αi=ei+aiW
,其中先验分布
a
可设为在定义域
X
上的均匀分布,即
ai=1/K
,同时先验权重
W=K
。
在证据深度学习中,损失函数的设计也发生了改变,由于类别概率
p
是服从狄利克雷分布的随机变量,因此传统的交叉熵损失函数会被改变为
ℒce-edl=1D∑x,y∈D ∑i∈𝕏 yiψS-ψαi
式中,
D
为数据集,
y
为标签值,
ψ⋅
为双伽玛函数。同样地,均方差损失函数也会被重构为
ℒm-edl=1D∑x,y∈D ∑i∈𝕏 yi-αiS2+αiS-αiαi2αi+1
最后,证据深度学习的工作中通常还会使用一个正则化项减少错误证据的产生,这个正则化损失是利用KL(Kullback-Leibler)散度约束错误类别的证据量降低到0来实现的。
医学影像智能诊断的临床落地始终面临一个核心矛盾:高质量标注数据的稀缺性与深度学习模型对大规模标注的依赖之间的冲突。由于医学图像采集成本高昂、专家标注耗时长以及罕见疾病本身的发生率极低(例如,在CT或MRI影像上对病灶进行三维分割),研究者必须在“数据稀缺”的约束下寻求解决方案(Han等,2024;Gao等,2024;Wu等,2024;Zhu等,2023;赵小明 等,2025;方超伟 等,2023)。在传统应对数据稀缺性的方案中,基于softmax概率的置信度评估是常见思路。例如,主动学习通过熵值最高或概率差值最大等启发式策略筛选样本以供标注(Qiu等,2017);半监督学习则可依据熵值高低筛选伪标签(Xiao等,2024)。而另一类重要方法则基于贝叶斯理论,通过估计后验分布来量化模型不确定性,为样本选择或伪标签加权提供了基础(Ghoshal等,2021)。
在这一背景下,证据深度学习为多种低资源学习范式提供了一个统一的理论基础与实现途径。近年来,大量基于证据理论的前沿工作正是在此框架下展开,并取得了显著进展。例如,研究人员已成功将其应用于半监督或无监督学习(Huang等,2021b;Chen等,2023b)以及主动学习(Zhao等,2025)。这些工作的共同核心在于,它们都将原本依赖启发式策略的环节(如伪标签筛选、样本选择)转化为对模型认知不确定性的量化与优化问题。通过显式量化因数据稀缺导致的认知不确定性,证据深度学习为低资源学习建立了统一的可靠性建模基础,成为缓解数据稀缺挑战的重要工具。
如图2所示,本节将系统梳理这些前沿工作,剖析证据深度学习如何提升低资源下的医学诊断性能、降低标注成本并增强临床可信度。
为了缓解标注瓶颈,半监督学习(semi-supervised learning,SSL)在医学影像领域获得了广泛关注。其核心思想是:在仅有少量标注样本的前提下,充分利用大量未标注样本的潜在结构信息提升模型泛化能力。传统半监督学习通常基于一致性正则化、伪标签或图神经网络,但在医学场景中,伪标签噪声的不确定性往往导致性能急剧退化。证据深度学习的引入为半监督学习提供了一种新的不确定性视角,使得网络能够显式地量化伪标签的可靠性,并通过证据融合机制动态地分配信任权重,从而在不增加额外标注的前提下实现鲁棒训练。
SEFNet(semi-supervised multiple evidence fusion network)(Huang等,2023)尝试将证据深度学习用于脑肿瘤半监督分割任务中,采用图像增强策略构造两个证据源:概率深度网络与证据神经网络,并利用证据理论规则融合两个不同来源的证据。在脑肿瘤分割任务上仅用部分标注即获得接近全监督的分割性能,展示了证据融合在脑肿瘤边界模糊场景中的鲁棒性,启发了后续研究对证据深度学习的应用。然而,MEDL(mutual evidential deep learning)(He等,2024)发现两个证据源的直接融合策略会因低质量伪标签的存在而放大偏差。为此,提出让不同架构的网络
𝒩1、𝒩2
互为证据源,利用改进的类别感知证据融合策略对伪标签进行置信加权,得到融合后的伪标签后使用每个体素的不确定性计算伪标签的可靠性。通过基于渐近费舍信息的证据学习策略,优先关注高置信度样本,随后逐步引入低置信度样本,从而避免了早期训练阶段因伪标签噪声导致的性能退化。该方法在CT(computed tomography)和MRI(magnetic resonance imaging)数据集上进行实验,在胰腺分割、脑肿瘤分割等需要医生精细标注的任务上实现了更高的分割精度。
受此启发,后续研究开始探索如何利用不同网络的融合以得到更可靠的伪标签。QDC-Net(quality-driven deep cross-supervised learning network)(Zhang等,2024a)在双分支互监督框架中引入实时质量估计模块,在模型层面对比证据网络与传统网络的性能并选择更优者作为指导,在样本层面,对比两个网络生成的伪标签一致性,并筛选质量更高的伪标签来指导后续训练,有效抑制了伪标签噪声。其在胰腺CT数据集上的实验再次证明,证据网络与常规网络互补可显著降低半监督训练对标注量的依赖。同样地,ETC-Net(evidential tri-branch consistency network)(Zhang等,2024b)则设计保守、进步与融合3条证据分支,通过证据理论融合实现伪标签自校正。然而,融合伪标签的方法缺少考虑将不确定性度量应用于一致性正则约束或伪标签过滤中。为此,EVIL(evidence-based uncertainty-aware semi-supervised medical image segmentation)(Chen等,2024c)将主观逻辑和证据理论引入到一致性正则化训练,通过双网络扰动预测的一致性约束实现理论保证的单前向不确定性量化,具体是通过一致性约束的损失函数从无标签的CT或MRI影像中学习到有效信息,利用了不确定度降低伪标签中的噪声,具体为
ℒcon=ℒeviF1x,Y2+ℒceF1x, ℳ⊙Y1
式中,
F1,F2
分别为证据神经网络和普通的神经网络,
ℒevi
为式(4)或式(5)所示的证据深度学习中的常用损失函数,
ℳ=𝕀(u<T)
为基于不确定性度量
u
的阈值函数,
𝕀
为指示函数,
Y1,Y2
表示分别从两个不同网络中获得的伪标签。
传统的蒙特卡洛采样或深度集成的实现需要多次前向推理,带来不可接受的时间成本,而EVIL仅需一次前向即可输出可信伪标签。CEF-IV(co-evidential fusion with information volume)(He等,2025a)引入了pignistic概率转化与信息体积度量,对多源体素级不确定性进行联合建模,通过证据理论融合获得更精确的逐体素置信度。在心脏分割、胰腺分割等任务上进行了实验,这些任务的标签都需要依靠医生的专业知识,该方法在伪标签可靠性评估上优于传统熵阈值法,为半监督语义分割提供了新的不确定性度量视角。同时还有研究考虑将原型学习与证据深度学习相结合,以进一步缓解数据稀缺的问题,EPL(evidential prototype learning)(He等,2025b)将标注数据与未标注数据的特征原型统一在广义证据框架内,通过体素级双重不确定掩膜实现自我纠错,双重不确定性评估由下式计算得到,具体为
um=-∑i=0K ux,y,zbCix,y,zlogbCix,y,z2Ci-1
式中,
m
为体素
x,y,z
的最终预测结果,
ux,y,z
为主观逻辑中式(1)所计算的空缺不确定性,
bCix,y,z
为类别
Ci
的信念质量,
K
为类别总数。在胰腺分割和左心房分割等任务上,EPL在不同标注比例下均取得最佳表现,验证了原型级证据对稀疏标注的有效利用。
上述工作共同揭示了一个核心观点:证据深度学习不仅提供了伪标签不确定性的可量化指标,更通过证据冲突与证据融合两个机制,使得网络能够在训练过程中自发地学会忽略低质量未标注样本,从而在数据稀缺条件下实现性能的稳定提升。
当标注预算极度受限时,主动学习(active learning,AL)通过迭代地选择信息量最大的样本以请求专家标注,能够以最小的标注代价换取最大的性能增益。传统主动学习依赖于不确定性采样或多样性采样,但在医学影像中,类别不平衡、稀有病变以及跨域差异常常导致不确定性估计失准,进而难以选择到信息量最大的样本。证据深度学习为主动学习提供了更精细的不确定性度量工具,将预测不确定性拆分为认知不确定性与偶然不确定性,可以进一步将高不确定性的样本分为位于训练分布外的样本和位于决策边界附近的样本,从而显著提升采样性能。
CSEAL(consistency-based semi-supervised evidential active learning)(Balaram等,2022)为了尽可能地利用稀缺的医学数据,尝试将基于证据深度学习的主动学习与一致性半监督学习结合,利用主观逻辑估计多标签分类下的偶然不确定性,而概率
p
此时服从
Beta
分布,即式(1)的狄利克雷分布的二元特殊形式,使用了每一类的熵的期望作为偶然不确定性的评估。具体为
AU=1ln2∑γ∈α, β γEψE+1-ψγ+1
式中,
α
和
β
分别为
Beta
分布的参数,同时满足
E=α+β
。该方法基于这个偶然不确定性来筛选无标签样本进行标注,在胸部X光多标签任务中使稀有异常类别的标注效率提升超30%。与此同时,Tan等人(2024a)验证了使用证据深度学习建模的式(1)所示的空缺不确定性在前列腺分割任务上的有效性。还有研究将传统的最大概率差值方法与证据深度学习结合,DEAL(deep evidential active learning)(Hemmer等,2020)采用狄利克雷分布替代softmax,直接以狄利克雷分布下最大和最小分类概率的差值刻画不确定性,在肺炎X光检测数据上,与其他传统不确定性评方法相比,该方法显著降低了样本的标准需求。但是,仅使用单一维度不确定性选择无标签样本的性能有限。因此,Hoarau等人(2024)基于证据理论将不确定性进一步细分为1.1节所述的认知不确定性和偶然不确定性,前者表达的是信息缺乏的情况,后者表达的是信息冲突的情况,在训练时就以二者之和作为样本的选择标准。在心脏疾病和帕金森疾病诊断的图像分类实验上,该策略显著优于传统的熵采样方法。FEAL(federated evidential active learning)(Chen等,2024a)进一步将主动学习拓展至联邦学习场景,借助狄利克雷分布同时建模偶然不确定性与认知不确定性,并提出多样性松弛策略以缓解跨域数据冗余.在5个多中心数据集上,包含皮肤病分类、眼底视盘分割等任务,即使存在域漂移,FEAL仍能提高全局模型性能,验证了证据不确定性在联邦标注预算分配中的实用价值。这些工作共同表明,证据理论不仅能在单站点数据匮乏时有效引导样本选择,更能在跨机构协作中校准不同域的不确定度,从而实现标注资源的全局最优配置。
总之,基于证据深度学习的主动学习方法利用了认知不确定性和偶然不确定性来驱动样本选择。通过优先标注模型最不确定的样本,能以最低的标注成本高效提升医学辅助决策系统的性能与泛化能力。
在真实临床环境中,医学影像诊断系统面临的一个挑战是如何在异质性数据下实现可靠的证据推理与决策支持。如图2所示,这类医学数据的表现形式多样,但其本质共性可归结为医学数据在来源、质量与分布上普遍存在的不确定性,具体体现在训练与测试两个阶段中:在训练阶段,数据本身常受到噪声干扰,如采集过程中的设备伪影、标注误差等非真实医学规律引入的扰动;同时,多模态数据融合中存在冲突与不一致,可视为训练分布内部的矛盾,例如同一病灶在超声图像与病理报告中呈现的范围差异,反映出不同模态因采集机制差异导致的异质性。在测试阶段,则可能出现训练分布未能覆盖的样本,如罕见病影像或疾病进展中的特殊表型,构成分布外泛化问题。在传统应对数据异质性的方法中,针对医学多模态信息融合,常基于中间特征设计融合策略(Huang等,2020;陈伟 等,2023;李夫辰 等,2025)。对于分布外样本检测或噪声样本的检测,则通常利用softmax概率的熵等统计量作为判别依据(Zhang等,2021)。
面对这样的挑战,基于证据深度学习与证据理论的思路正逐渐成为一种系统性的解决方案。近年来,大量研究已从不同侧面验证了其有效性,例如在噪声或伪影干扰下的鲁棒性学习(Fidon等,2024;Tan等,2024a,b;Dawood等,2023;Xu等,2023)、多模态信息的高效融合(Liu等,2022,2025c;Xu等,2022a;Huang等,2021a),以及针对分布外样本的有效检测(Yao等,2024;Roy等,2022;Ulmer等,2020)、域适应方法(Yan等,2025;Pei等,2024)等方向均取得了显著进展。这些看似独立的研究方向,其共同的理论内核在于突破了传统确定性预测的局限。它们利用EDL显式建模了认知不确定性(模型知识的局限)与偶然不确定性(数据固有的噪声或模糊),并通过增强模型可解释性来辅助决策。正是这种将不确定性量化与模型可解释性相结合的能力,为异质性数据下的可靠推理提供了统一框架,共同回应了数据普遍存在不可靠性的这一本质问题。
面对数据采集或标注过程中引入的内部噪声污染(如伪影、运动模糊),基于证据深度学习方法的核心策略是显式耦合输入数据质量评估与模型推理过程,建模数据固有信息损坏引发的偶然不确定性。此类方法在架构中集成图像质量感知模块、噪声鲁棒特征学习机制等,动态评估关键区域的信息有效性。核心能力在于当噪声显著破坏目标区域可诊断性时,模型能输出与噪声干扰程度匹配的高不确定性值,明确指示预测不可靠性源于数据质量问题本身。
TBraTS(trusted brain tumor segmentation)网络(Zou等,2022)开始在U-Net脑肿瘤分割网络中引入可信证据获取机制,通过主观逻辑将分割网络的预测视为主观意见,并用狄利克雷强度衡量分割结果的可靠性,可以额外输出空缺不确定性图,指出分割结果中不可靠的区域。实验表明,在CT中逐步加入高斯噪声以模拟CT的固有噪声后,TBraTS的Dice指标下降幅度均小于其他方法。类似地,Region-based EDL(Li等,2023)将脑肿瘤分割任务分解为“区域证据收集—狄利克雷参数化—主观逻辑决策”3步,同样是将卷积神经网络的输出视为主观逻辑中的主观意见
τ
,与TBraTS的区别在于,它直接优化Dice损失函数的贝叶斯风险。由于主观逻辑框架下概率是服从狄利克雷分布的,相比于普通的Dice损失函数提升了收集区域证据的可靠性。后续工作开始不再局限于分割任务,医学分类任务的研究也受到了关注。UML(uncertainty-informed mutual learning)(Ren等,2023)框架将证据深度学习引入分类任务与分割任务的协同学习,证据神经网络为图像级与像素级同时输出狄利克雷分布,借此用式(1)量化输入噪声与标签噪声导致的空缺不确定性,接着设计了不确定性导航器与不确定性指导器两个模块——前者用分割任务的高置信区域引导分类网络聚焦可靠特征;后者则用分类任务的高置信掩膜反向屏蔽分割中的潜在噪声区域。通过将不确定性作为桥梁,UML在任务间实现了可信特征的动态筛选与共享,从而将原本可能因噪声而互相拖累的联合学习转变为互相纠偏、共同降噪的正向循环,其方法为构建可解释、可信赖的视盘分割与青光眼诊断多任务医学影像模型开辟了新思路。同时,Shi等人(2024)发现病理切片图像分辨率较高,难以直接将证据深度学习用于处理噪声,因此提出E2-MIL(explainable and evidential multiple instance learning)网络,处理了病理全切片图像的稀疏标注环境中的噪声问题,这些噪声来自于医生的标注偏差。E2-MIL网络引入细节感知注意力模块利用互补信息,将局部网络学到的细粒度注意力迁移到全局网络,从而抑制非肿瘤噪声区域的高响应,结构感知注意力模块在实例聚类层面显式建模空间一致性,避免孤立噪声误导决策,最后基于主观逻辑理论的不确定性感知实例分类器为每个实例输出狄利克雷分布,将噪声带来的认知不确定性显式量化,为下游决策提供可信度量。
过往的工作往往仅关注单一图像模态的噪声处理,随着应用场景复杂化及多模态数据的广泛使用,噪声多视图或多模态数据下的可靠分类问题也受到越来越多的关注。CMDC(conformalized multi-view deep classification)(Liu等,2024)首先将共形预测引入噪声多视图分类中,该方法生成集合式的预测结果而非单一输出,首先其将证据深度学习中的不确定度作为一个分布外类别统一到输出空间中,具体为
𝒪iv=biv=bikvk=1K+1
式中,
𝒪
为输出空间,
bikvk≤K
表示由式(3)定义的、类别
k
在视图下的信念质量,
K
为类别总数,而空缺不确定性被视为一个新的类别,并用
biK+1v
表示。接着设计了一个共形预测中的不一致得分函数,进而根据该不一致得分函数构建预测集合。CMDC在保证真实标签以用户设定的概率包含于预测集合的前提下,尽可能缩小预测集大小,从而增强决策的可信度。该方法在含有噪声的膝盖MRI数据上进行实验,使用轴状位、冠状位和矢状位3个视图进行综合诊断,仍保持着良好的性能。而EyeMoSt+网络(Zou等,2024)针对多模态眼底图像极易出现的成像噪声的问题,构建了一条全新的证据融合管线。其核心思想是假设标签服从高斯分布,让网络先为每一模态在预训练权重上建立正态—逆伽玛先验,借此将噪声带来的偶然不确定性与数据不足导致的认知不确定性同时显式建模。具体为
AU=βmαm-1, EU=βmδmαm-1
式中,
αm,βm,δm
为正态—逆伽玛先验的参数。随后将先验转换为具有重尾特性的学生t分布,再用混合学生t分布的视角对多模态信息进行软加权融合。这种设计赋予模型对噪声异常值和缺失模态天然的鲁棒性,而置信度感知排序正则项则进一步引导网络在面对含噪单模态或低质量融合结果时,能够自动下调其置信评分,从而避免错误决策。该方法使多模态系统在真实临床噪声场景下仍能保持可解释、可信任的行为。文本数据在医学数据中也是重要的一部分,但上述工作只关注了图像模态。所以,知识增强型视觉—语言基础模型RetiZero(Wang等,2025)通过双架构融合框架:MAE(masked autoencoders)自监督学习与CLIP(contrastive language-image pre-training)框架,实现了病理结构与文本语义的协同表征。为了应对受噪声污染的眼底图像(如扫描时患者的移动造成运动模糊)或图像—文本对对齐程度不高影响训练的问题,该方法创新性地引入基于狄利克雷分布的不确定性视觉语言特征校准方法:将图像—文本相似度转化为证据,具体为
EV2L=SoftPlusVTLEL2V=SoftPlusLTV
式中,
V
为图像特征,
L
为文本特征。通过证据深度学习方法来基于证据构建对应的狄利克雷分布,并设计基于式(4)的证据交叉熵损失函数与KL散度损失函数建模图像和文本之间的相似度,从而显著提升模型对噪声数据和跨域特征的鲁棒性。RetiZero能够显著提升对罕见眼底疾病的识别能力,展现出良好的临床应用潜力,尤其适用于复杂眼底疾病的辅助诊断场景。
证据深度学习为提升医学模型在噪声数据下的鲁棒性提供了统一框架。在单视图任务中,该类方法利用模型产生的不确定性识别潜在噪声样本,从而抑制模型的过度自信预测。在多视图任务中,不确定性则用于自动降低含噪视图的贡献权重,促使模型融合更可靠的视图信息以生成确定性更高的结果。二者的核心共性在于引导模型对噪声保持内在的审慎性,这种基于不确定性的噪声识别与抑制机制,是构建鲁棒医学辅助决策系统的关键。
为解决医学数据中不同模态或不同视图数据的天然异质性及其表征同一目标可能存在的语义冲突带来的融合难题,基于证据深度学习方法的关键在于显式建模跨模态一致性与冲突关系,构建能检测模态间信息矛盾并量化其强度的自适应融合机制,超越简单的特征拼接,避免冲突信息下输出误导性的单一高置信度预测。同时实现在模态间证据存在显著冲突时,模型能有效捕捉并量化这种不一致性,输出反映融合决策置信度分歧的综合不确定性,为复杂病例诊断提供透明化的信息。
研究人员在融合医学多模态数据时通常先处理模态间的矛盾。例如MMEF(multi-modality evidence fusion)(Huang等,2025)将每个模态的卷积神经网络的输出映射为证据理论中的信念函数,再利用上下文折扣机制根据模态可靠性动态降低低质量证据的权重,其中模态的可靠系数通过端到端学习获得而无需人工设定,基于这种融合机制得到最终预测结果。这种融合方式提高了可靠模态中获得的证据的权重,进而改善了PET-CT模态融合后的淋巴瘤分割效果。但是Sun等人(2026)发现模态内的一致性程度也影响了模态融合的效果,因此REDNet(reliable evidential discounting network)(Sun等,2026)进一步引入模态内一致性评估模块,在同模态内部测量体素级一致性;跨模态差异聚合模块则在特征空间显式对齐模态差异,最终由妥协融合机制完成带妥协的基于证据理论的融合。该机制根据模态的质量进行动态调整从证据神经网络中提取的证据,以削弱低质量模态的影响。所设计的妥协系数包含了模态内一致性评估值和模态间一致性评估值,当一致性更高时将会保持模态对融合结果的影响,基于妥协系数修改信念质量和不确定度。具体为
bkd=dk⋅bk,   ud=1-∑k=1K dkbk
式中,
dk
为妥协系数,
bk
表示式(3)定义的第
k
类信念质量。这种对模态的可靠性估计,使分割系统在多期相CT中存在强度非均匀性或运动伪影时,可减少模态间冲突的影响,保持可信分割。在这个基础上,研究人员开始探索模态质量差异极大时的融合问题,MEFN(multi-modal evidential fusion network)(Qi等,2025)将流程拆为两阶段:跨模态特征学习阶段,通过域对齐策略削弱PET与CT的域差异,提取鲁棒的双模态特征;多模态可信融合阶段,利用互注意力机制与不确定性校正器动态量化各模态可靠度,并基于证据理论融合出最终的分割结果。此外,引入不确定性感知损失迫使网络聚焦低置信区域,以提升可信信息的提取能力。MEFN还可以同步输出逐体素可信区间,为临床接受或拒绝自动结果提供定量依据。
在多视图问题上,研究人员也采用了与多模态融合类似的方法。在胰腺肿瘤亚型的鉴别诊断中,由于不同亚型间形态学差异细微且易受图像噪声干扰,因此融合临床先验的可信多视图学习框架(Huang等,2024)依据胰腺肿瘤恶性风险程度将亚型分为高危肿瘤、低危肿瘤与非肿瘤超类,每个类包含具体的肿瘤亚型,该方法将此先验知识作为辅助信息引导神经网络生成增强意见,具体为
bpx=∑y∈A𝕏aA(xy)bAy
式中,
A(𝕏)
为超类,
aA⋅
为先验权重,
bA⋅
为信念质量。同时引入全局视图以补充局部特征缺失,增强模型表征能力,最后通过证据理论的融合规则集成多视图意见,实现可靠诊断。多视图证据学习框架MERIT(multi-view evidential learning for reliable and interpretable)(Liu等,2025d)将每个局部视角的预测视为主观意见
τ
,同时输出类别概率及量化不确定度,从而兼顾可解释性。它引入分布感知先验,缓解类别不平衡与分布漂移带来的偏差。主观逻辑中的融合策略用于融合多个视角的意见,具体为
bk=bkmun+bknumum+un-umun,  u=umunum+un-umun
式中,
m
和
n
表示两个不同视角,
bkm
为式(3)定义的视角
m
第
k
类的信念质量,
um
为式(1)定义的视角
m
的空缺不确定性。可解释的融合过程使得MERIT可以直观展示不同视野(如肝被膜下、实质区等)对纤维化程度判断的相对重要性,为临床决策提供透明依据。
基于证据深度学习的多模态融合方法为各模态数据分别构建模型以量化其不确定性,再使用证据理论进行融合。该范式的优势在于能显式处理模态间的冲突与不一致性,最终输出低不确定性的联合决策,从而提升医学辅助决策的可靠性与透明度。
针对模型训练中未充分覆盖的临床场景所引发的分布外风险,基于证据深度学习的方法可以提供一套从感知到适应的递进式解决方案。其核心在于构建对非训练分布特征的鲁棒感知能力,并量化其引发的认知不确定性。此类方法能够可靠地识别偏离已知分布的样本,无论偏离程度是细微还是显著,均输出可解释的高不确定性估计,从而为临床决策提供关键的安全警戒。更进一步,对于分布偏移相对较小的近分布外场景,例如不同医疗机构间的数据差异,仅实现检测已不足够。域适应技术在此基础上可主动引导模型适应目标域的数据分布,从而有效提升模型在面对分布差异时的泛化性能,实现对近分布外数据的稳健诊断。
证据深度学习中的空缺不确定性度量
u
表达了“我不知道”的程度,所以m-EDL(modified-EDL)(Nagahama,2023)通过将证据深度学习的空缺不确定性直接解释为未知类的概率,将原有的
K
个类别扩展成了
K+1
个类别,这种实现方式只需从
K+1
类中选择概率最高的类作为最终预测结果,从而绕过了阈值设定的难题。在3个跨域数据集(医学影像、交通标志和自然图像)上的实验显示,m-EDL在已知类准确率不降的前提下,对未知类的检测相对于普通证据深度学习得到了提升。然而,由于近分布外样本与分布内样本极其相似,m-EDL难以检测出近分布外样本,所以ERNN(evidence reconciled neural network)(Fu等,2023)为了应对近分布外样本难以与分布内样本区分的问题,在传统证据深度学习框架之上增加证据调和块,通过改造证据头获得的证据以约束模型的决策边界,使证据神经网络对近分布外样本更敏感,证据获取的改造具体为
e¨=e-minei, i∈1,⋯, K
式中,
e¨
和
e
分别为改造后和原始的证据。在理论上可以证明ERNN降低了经验损失,在皮肤病数据集与私有胰腺肿瘤数据集上的实验表明,其在AUC(area under curve)指标上超越其他基于证据深度学习的不确定性方法,能很好地识别出罕见的皮肤病和胰腺肿瘤亚型。传统深度学习在训练阶段假设类别集合封闭,一旦测试阶段出现未知类别,模型只能给出错误的高置信度。上述方法在利用不确定度检测分布外样本时,往往会遇到阈值难以选取的问题,这也促使了后续对于阈值设定的研究。基于不确定性驱动的开放集识别模型(uncertainty-inspired open set,UIOS)(Wang等,2023)针对的是视网膜疾病诊断的实际应用中难以有效处理测试阶段中未知类别的问题。在方法上,UIOS不仅能够输出类别概率,还可利用证据深度学习计算每个样本的空缺不确定性,从而定量表达模型对预测结果的置信程度。为了更准确地识别分布外样本还提出了一个自适应阈值策略,期望低于阈值的样本均是预测正确的,高于阈值的样本则需要人工介入,因此使用一个验证集计算这个阈值,基于真阳率TPRs和假阳率FPRs构建了一个目标函数
𝓁θ
,具体为
𝓁θ=2×TPRsθ-FPRsθ
式中,
θ
为不确定度阈值,进而通过
θ=argmaxθ𝓁θ
计算得到最终阈值。这一机制显著提升了模型对未知类别样本和分布外样本的识别能力与鲁棒性。UIOS在应对分布外类别的视网膜疾病、低质量眼底图像及非眼底图像等复杂情况时,能有效产生高不确定性评分,从而触发人工审核机制,体现出其在真实临床筛查环境中良好的适用性与可靠性。该方法为视网膜异常的自动化筛查提供了更为安全、可信的解决方案,尤其适用于开放世界场景下的辅助诊断应用。
然而,仅能识别分布外样本在实际临床应用中仍显不足,为了使模型在新数据分布下的性能衰减问题得到缓解,域适应技术被应用到医学数据中。M-MSSEU(source-free domain adaptation framework)(Pei等,2024)在无源域适应场景中处理多模态脑卒中数据,其基于伪标签方法进行无源域适应学习,并使用证据深度学习的空缺不确定性筛选可靠的伪标签加入训练,从而使得模型在不同域中保持较好的脑卒中病灶分割效果。
该类方法的共同基础是利用证据深度学习量化认知不确定性,其应用已从被动的分布外样本检测,演进至主动的域适应。这不仅包括通过高不确定性发出警示,更体现在引导模型优化、适应新数据分布上,从而系统性地提升临床决策的鲁棒性。
在医学人工智能应用中,数据公平伦理已成为至关重要的问题,其核心在于确保模型决策不因数据分布的系统性偏斜而产生对特定群体或类别的不公正结果。具体而言,这一类偏斜常表现为两种形式:类别不平衡,例如在肺癌中腺癌和鳞癌常见,而肉瘤样癌等罕见亚型样本量极少;组间不平衡,指数据在年龄、性别和种族等人口学属性上存在显著分布不均。尽管二者的表现形式不同,但其共同本质在于:部分类别或群体在训练数据中代表性不足,导致模型在这些区域产生有偏预测或过度置信的错误判断。传统方法通常基于类别频率处理数据不均衡问题。其中重加权方法通过反向类别频率等启发式策略为不同类别设置损失函数权重(Yeung等,2022),以平衡梯度更新;样本合成方法则是在特征空间中对少数类样本进行过采样,通过插值或生成式模型人工合成新样本(Koetzier等,2024),从而平衡各类别的数据分布。
近年来,研究者围绕“证据深度学习如何在数据不均衡条件下保持公平性评估”这一核心问题展开了系统探索(El-Din等,2024;Gan等,2023)。如图2所示,证据深度学习为解决此类公平性问题提供了机制上的内在优势,它通过显式建模认知不确定性,识别样本稀少区域,在设计上天然地适用于检测和缓解由数据偏斜引发的决策偏见,为构建更公平的医学人工智能系统提供了途径。
为了缓解这种因数据分布不均带来的模型偏差,一类核心策略聚焦于修正训练过程中数据分布的影响力或采样频率。此类方法主要包括两种思路: 1)损失函数权重调整,即在模型训练的损失计算中,赋予少数类样本更高的惩罚权重。当模型错误分类少数类样本时,会产生更大的梯度更新信号,从而迫使模型给予少数类更多关注,在决策边界上做出更有利于识别少数类的调整。2)过采样少数类,其核心思想是通过复制或基于插值原理在现有少数类样本周围生成“相似”的新样本,从而在输入空间层面人为地提升少数类样本的数量或密度,以期平衡训练集的数据分布,让模型有更多机会学习少数类的特征。这两种思路的共同点在于,它们都主要依赖并基于有限的原始少数类样本信息,通过改变其在学习过程中的权重或出现频率对抗数据倾斜带来的负面影响。
混合式证据深度学习Hybrid-EDL(Xia等,2022)通过组合使用数据过采样和后校准的两级策略提升证据深度学习模型的鲁棒性,训练阶段利用过采样少数类的方法实现所有类别的数量平衡,缓解少数类的证据稀缺问题;推理阶段采用温度标定与直方图分箱对狄利克雷参数进行后验校准,消除残余置信偏差。该方法将证据深度学习与经典校准技术结合,其效果在不平衡的皮肤病数据上得到了验证。但是,这种组合式的方法难以有效利用证据深度学习的优势,为了解决证据深度学习在医疗任务中忽视了类别不平衡对不确定性量化的系统性偏差,Class-balanced EDL(Xia等,2024)提出类别平衡证据深度学习框架。相比于主观逻辑中手动设置的先验分布
a
,该方法使用可学习的狄利克雷先验在训练阶段根据类别逆频率动态调整先验分布,基于式(17)进行先验分布的调整
ℒp=∑k=1K βklogβkηk
式中,
βk
为狄利克雷先验,
ηk
为类别逆频率,
K
为类别总数。该损失函数让先验不再是均匀分布,从而抑制少数类证据的过拟合倾向。同时设计了证据池化损失在批次内对不同类别证据进行自适应均衡,样本量更小的类别会得到更多关注,缓解了证据累积过程中的类别偏向。该方法在皮肤病数据集上验证了其在数据不平衡时识别近分布外样本和远分布外样本的有效性。另一个改进方向是调整决策边界的大小,证据深度学习重加权边缘调整策略(Zhu等,2025)被提出来应对不平衡的宫颈疾病分类任务,宫颈疾病中宫颈癌样本最少而轻度炎症最为常见,呈现明显的长尾分布。该方法在证据深度学习框架内为每一类别设置决策边界调整系数,具体为
m=CmaxKNi-14-1Ni-14i=1K
式中,
C
为人工设定的超参数,
Ni
表示类别
i
的样本数量,
K
表示类别数量,少数类样本将拥有更大的决策空间。同时以类别频率的倒数对证据深度学习的损失函数进行加权,因此少数类样本将拥有更大的权重,迫使网络为少数类保留更大决策裕度,从而抑制虚假高置信度。
上述方法仅关注了类别不均衡问题,而算法在现实医疗场景中的群体公平性问题同样重要。U-FARE(uncertainty-aware fair federated learning)(Chen等,2025a)针对公平性问题提出了一种面向医疗疾病预测的不确定性感知公平联邦学习框架,其在公平联邦学习中引入证据神经网络以量化不确定性,并基于全局与本地模型之间的意见冲突程度,识别出最具差异化的本地模型参与聚合,从而在优化预测性能的同时,系统性地增强对样本数较少的群体(如不同医疗机构、人口学组别)的公平性。该框架通过在群体层面提供一致的预测性能,有效应对来自不同医疗机构的数据异质性问题,从而在隐私保护的前提下,显著提升阿尔茨海默病等疾病诊断的可靠性及跨站点的泛化能力。U-FARE为在实际医疗环境中构建既公平又可信的人工智能辅助诊断系统提供了新途径。
证据深度学习处理医学数据不均衡的方法在于利用模型输出的不确定性作为动态指导信号,自动识别并为模型尚未把握的困难样本或少数类样本赋予更大的权重,有效缓解了数据不均衡的问题。
为突破现有少数类样本的信息限制并引入更丰富的表征多样性,另一类核心策略应运而生:直接合成新的、非原始存在的训练样本扩充数据集,尤其侧重于少数类样本的生成。这种合成可以通过两种主要范式实现:1)基于深度生成模型的合成(Grina等,2023;潘永生 等,2025),例如生成对抗网络(generative adversarial network,GAN)和变分自编码器(variational autoencoder,VAE)。这些模型通过学习数据底层的分布特性,特别是少数类的特定模式,能够创造性地生成语义有效、多样性丰富的全新样本,旨在覆盖原始数据流形中未被充分代表的区域。2)基于混合(Mixup)原则的插值驱动合成,该方法通过在输入或特征层面对样本进行线性或非线性的混合操作,生成位于原始样本之间的虚拟过渡样本。这类方法的共同目标都是突破原始少数类数据的稀缺性,注入新颖的、信息量更大的合成样本,为模型提供更全面、更具代表性的少数类样本,从而更有效地应对类别不平衡的挑战。
目前的研究工作集中在使用Mixup来合成新样本,ME-OSR(multi-level mixup-based evidential open set recognition)(Zhao等,2024)在图像级与特征级上对于同一类别样本应用Mixup操作合成新样本,而对不同类别样本也进行同样的Mixup操作,但如果所用的混合权重接近0.5就将新合成的样本视为分布外样本。这样的多级混合机制可以有效合成出可靠的少数类样本。随后引入正则化证据深度学习分类器,通过开放集正则项约束网络对未知样本输出均匀狄利克雷分布,实现未知类别的可靠拒识。该方法在不平衡的视网膜疾病数据上进行了验证,其中正常视网膜的样本数量远超病变样本,经过样本合成操作后有效提升了病变检测的准确率。数据不均衡问题不仅存在于单标签分类,在多标签分类中表现得更为突出,MOSD(multi-label open set diagnosis)(Pang等,2025)将类别不平衡的处理拓展至同时存在多标签共现与未知类别的眼部疾病诊断场景。多标签开放集诊断框架包含了两大模块:1)平衡多标签数据增扩模块,其将随机采样器和少数类采样器采样得到的两类样本进行混合得到新样本,两个采样器的共同使用可以使后续合成出来的样本既有多样性又兼顾了少数类,从而缓解多标签中的长尾分布。2)多标签证据深度学习模块,其为每一标签独立输出狄利克雷参数,并通过开放集证据损失约束未知类别输出均匀分布,实现未知病变的可信拒识。该研究将类别不平衡、多标签学习与开放集识别统一于证据深度学习框架,拓展了证据深度学习的应用范围。
基于证据深度学习的合成少数类样本方法在于利用不确定性指导样本生成,从而生成信息量更丰富、更具代表性的新样本,从数据层面有效提升模型对少数类的学习能力。
本综述系统性地回顾并梳理了证据深度学习在应对真实复杂医学数据、提升医学影像分析系统可靠性与安全性的研究进展。与点估计预测且时常伴随非合理高置信度的传统深度学习模型相比,证据深度学习提供了一个有理论依据的框架来量化不确定性。这种对模型自身认知边界进行显式建模与量化的能力,是推动人工智能在临床决策等高风险领域安全应用的核心要素。具体而言,针对数据稀缺问题,证据深度学习通过其对认知不确定性的量化,为识别因训练样本不足而导致的不可靠预测提供了定量指标。在应对数据异质性这一核心挑战时,证据深度学习展现出了关键优势,它不仅能通过提升总体不确定性反映输入证据的内在矛盾或质量缺陷,为进一步处理提供依据,还能通过显著升高的认知不确定性提供一个有效的分布外样本拒绝机制,这是保障模型在开放临床环境中安全运行的关键技术保障。此外,对于数据不均衡问题,其不确定性输出也为评估模型在少数类上的预测稳定性提供了重要视角。综上所述,证据深度学习的根本性贡献在于,它推动医学决策范式从数据驱动向证据驱动进行演进,通过为模型输出赋予可量化的不确定性,显著增强了模型的透明度与可靠性。
尽管证据深度学习已取得显著进展,但在其通往临床常规应用的道路上,仍存在若干挑战与机遇,需要未来研究在多个方向上进行重点突破。首要问题在于证据建模的理论完备性与公平性。当前方法大多直接沿用主观逻辑框架,将模型输出直接解释为证据,但证据向量生成的可靠性缺乏严格的理论支撑,其是否真实反映模型信心仍待商榷。更重要的是,现有研究较少讨论不确定性度量的公平性。如果训练数据存在偏差,模型不仅预测结果可能不公,其输出的不确定性本身也可能对某些群体产生系统性高估,这构成了新的伦理挑战。因此,发展能够保障证据可靠性与公平性的新理论模型是基础。
在此基础上,建模证据的损失函数的选择也亟待理论指导。目前的研究往往依靠经验从几种常用的证据建模损失函数中挑选,但这些损失函数对模型校准性和鲁棒性的影响机制并不明晰。未来需要深入探究不同损失函数的理论性质,并发展能适应特定临床任务风险偏好的动态损失函数,而非仅凭经验选择。与此紧密相关的则是狄利克雷先验分布的选择问题。当前先验分布多被手动设置为简单超参数,一般为均匀分布,但不同医学任务可能蕴含不同的先验模式,因而需要更具适应性的先验设置。虽然已有工作探索可学习的先验分布,但其理论基础薄弱,易出现过拟合,且学习到的先验是否真正捕捉了数据内在的先验模式尚缺乏严格验证。因此,探索合适的自适应狄利克雷先验分布是必要的。
证据深度学习价值的最终体现,在于其不确定性度量能否有效集成到临床决策流程中。这带来了不确定性校准与决策支持的标准化挑战。核心在于实现临床校准,即模型的不确定性不仅要满足数学上的校准,更需与临床风险等级建立明确关联。例如,一个高不确定性输出应直接对应高误诊风险或需进一步检查的临床意义。进而,必须依据临床效用理论,通过量化不同决策后果的代价,来科学设定触发医生介入的不确定性阈值。同时,如何将复杂的不确定性信息以直观方式呈现给医生,并设计流畅的人机交互流程,是技术得以落地的关键环节。最终,所有技术进步的有效性与安全性,均有赖于多中心、大规模的前瞻性临床研究的严格验证,这是连接理论研究与临床实践的必要环节。
以证据深度学习为代表的不确定性量化方法,是构建下一代鲁棒、可信的智能医疗系统的不可或缺的组成部分。它不仅是提升模型性能的技术手段,更是建立临床信任、实现有效人机协同的基石,将为精准医疗和患者护理水平的提升做出实质性贡献。
  • 国家自然科学基金项目(62472315,62476165)
参考文献 引证文献
排序方式:
Amini A,Schwarting W,Soleimany A and Rus D. 2020. Deep evidential regression//Proceedings of the 34th International Conference on Neural Information Processing Systems. Vancouver,Canada: Curran Associates Inc.: #1251
Balaram S,Nguyen C M,Kassim A and Krishnaswamy P. 2022. Consistency-based semi-supervised evidential active learning for diagnostic radiograph classification//Proceedings of the 25th International Conference on Medical Image Computing and Computer-Assisted Intervention. Singapore,Singapore: Springer: 675-685 [DOI: 10.1007/978-3-031-16431-6_64]
Blundell C,Cornebise J,Kavukcuoglu K and Wierstra D. 2015. Weight uncertainty in neural network//Proceedings of the 32nd International Conference on Machine Learning. Lille,France: JMLR: 1613-1622
Cai L,Gao J Y and Zhao D. 2020. A review of the application of deep learning in medical image classification and segmentation. Annals of Translational Medicine,8(11): #713 [DOI: 10.21037/atm.2020.02.44]
Chen D,Zhang Q,Kaplan L,Jøsang A,Jeong D,Chen F,et al. 2025a. Ethical AI for healthcare systems: uncertainty-aware,fair federated learning//Proceedings of 2025 IEEE/ACM Conference on Connected Health: Applications,Systems and Engineering Technologies (CHASE). New York,USA: IEEE: 222-233 [DOI: 10.1145/3721201.3721367]
Chen H S,Xu C,Guan Z Y,Zhao W and Liu J L. 2025b. Biased incomplete multi-view learning//Proceedings of the 39th AAAI Conference on Artificial Intelligence. Philadelphia,USA: AAAI: 15767-15775 [DOI: 10.1609/aaai.v39i15.33731]
Chen J Y,Ma B T,Cui H F and Xia Y. 2024a. Think twice before selection: federated evidential active learning for medical image analysis with domain shifts//Proceedings of 2024 IEEE/CVF Conference on Computer Vision and Pattern Recognition. Seattle,USA: IEEE: 11439-11449 [DOI: 10.1109/cvpr52733.2024.01087]
Chen M Y,Gao J Y and Xu C S. 2023a. Uncertainty-aware dual-evidential learning for weakly-supervised temporal action localization. IEEE Transactions on Pattern Analysis and Machine Intelligence,45(12): 15896-15911 [DOI: 10.1109/TPAMI.2023.3308571]
Chen M Y,Gao J Y and Xu C S. 2024b. R-EDL: relaxing nonessential settings of evidential deep learning//Proceedings of the 12th International Conference on Learning Representations. Vienna,Austria: ICLR: 3768-3794
Chen W,Sun K K,Li Q X,Xie K and Ni X Y. 2023. Research progress of multimodal medical image fusion methods. Chinese Journal of Radiological Health,32(5): 580-585
陈伟,孙康康,李奇轩,谢凯,倪昕晔. 2023. 多模态医学图像融合方法的研究进展. 中国辐射卫生,32(5): 580-585 [DOI: 10.13491/j.issn.1004-714X.2023.05.020]
Chen Y Y,Yang Z Y,Shen C Y,Wang Z W,Qin Y and Zhang Y. 2023b. EVIL: evidential inference learning for trustworthy semi-supervised medical image segmentation//Proceedings of 2023 IEEE 20th International Symposium on Biomedical Imaging (ISBI). Cartagena,USA: IEEE: 1-5 [DOI: 10.1109/isbi53787.2023.10230589]
Chen Y Y,Yang Z Y,Shen C Y,Wang Z W,Zhang Z Z,Qin Y,et al. 2024c. Evidence-based uncertainty-aware semi-supervised medical image segmentation. Computers in Biology and Medicine,170: #108004 [DOI: 10.1016/j.compbiomed.2024.108004]
Dawood T,Chan E,Razavi R,King A P and Puyol-Antón E. 2023. Addressing deep learning model calibration using evidential neural networks and uncertainty-aware training//Proceedings of the 20th IEEE International Symposium on Biomedical Imaging (ISBI). Cartagena,USA: IEEE: 1-5 [DOI: 10.1109/isbi53787.2023.10230515]
Deng D R,Chen G Y,Yu Y,Liu F R and Heng P A. 2023. Uncertainty estimation by fisher information-based evidential deep learning//Proceedings of the 40th International Conference on Machine Learning. Honolulu,USA: PMLR: 7596-7616
Der Kiureghian A and Ditlevsen O. 2009. Aleatory or epistemic? Does it matter? Structural Safety,31(2): 105-112 [DOI: 10.1016/j.strusafe.2008.06.020]
Ding H W,Huang N N and Cui X H. 2024. Leveraging GANs data augmentation for imbalanced medical image classification. Applied Soft Computing,165: #112050 [DOI: 10.1016/j.asoc.2024.112050]
El-Din D M,Hassanein A E and Hassanien E E. 2024. An adaptive and late multifusion framework in contextual representation based on evidential deep learning and Dempster-Shafer theory. Knowledge and Information Systems,66(11): 6881-6932 [DOI: 10.1007/s10115-024-02150-2]
Fang C W,Li X,Li Z Y,Jiao L C and Zhang D W. 2023. Interactive dual-model learning for semi-supervised medical image segmentation. Acta Automatica Sinica,49(4): 805-819
方超伟,李雪,李钟毓,焦李成,张鼎文. 2023. 基于双模型交互学习的半监督医学图像分割. 自动化学报,49(4): 805-819 [DOI: 10.16383/j.aas.c210667]
Fidon L,Aertsen M,Kofler F,Bink A,David A L,Deprest T,et al. 2024. A dempster-Shafer approach to trustworthy AI with application to fetal brain MRI segmentation. IEEE Transactions on Pattern Analysis and Machine Intelligence,46(5): 3784-3795 [DOI: 10.1109/tpami.2023.3346330]
Fu W,Chen Y F,Liu W,Yue X D and Ma C. 2023. Evidence reconciled neural network for out-of-distribution detection in medical images//Proceedings of the 26th International Conference on Medical Image Computing and Computer-Assisted Intervention. Vancouver,Canada: Springer: 305-315 [DOI: 10.1007/978-3-031-43898-1_30]
Gal Y and Ghahramani Z. 2016. Dropout as a Bayesian approximation: representing model uncertainty in deep learning//Proceedings of the 33rd International Conference on Machine Learning. New York,USA: PMLR: 1050-1059
Gan Y C,Dai Z C,Wu L C,Liu W Y and Chen L D. 2023. Deep reinforcement learning and dempster-Shafer theory: a unified approach to imbalanced classification//Proceedings of the 3rd International Conference on Computer Science,Electronic Information Engineering and Intelligent Control Technology (CEI). Wuhan,China: IEEE: 67-72 [DOI: 10.1109/cei60616.2023.10528127]
Gao Y B,Gao Z Y,Gao X,Liu Y Y,Wang B M and Zhuang X H. 2024. Evidential concept embedding models: towards reliable concept explanations for skin disease diagnosis//Proceedings of the 27th International Conference on Medical Image Computing and Computer-Assisted Intervention. Marrakesh,Morocco: Springer: 308-317 [DOI: 10.1007/978-3-031-72117-5_29]
Ghesu F C,Georgescu B,Gibson E,Guendel S,Kalra M K,Singh R,et al. 2019. Quantifying and leveraging classification uncertainty for chest radiograph assessment//Proceedings of the 22nd International Conference on Medical Image Computing and Computer-Assisted Intervention. Shenzhen,China: Springer: 676-684 [DOI: 10.1007/978-3-030-32226-7_75]
Ghoshal B,Swift S and Tucker A. 2021. Bayesian deep active learning for medical image analysis//Proceedings of the 19th International Conference on Artificial Intelligence in Medicine. [s.l.]: Springer: 36-42 [DOI: 10.1007/978-3-030-77211-6_4]
Grina F,Elouedi Z and Lefevre E. 2023. Evidential generative adversarial networks for handling imbalanced learning//Proceedings of the 17th European Conference on Symbolic and Quantitative Approaches with Uncertainty. Arras,France: Springer: 257-269 [DOI: 10.1007/978-3-031-45608-4_20]
Guo Z,Wan Z L,Zhang Q S,Zhao X J,Zhang Q,Kaplan L M,et al. 2024. A survey on uncertainty reasoning and quantification in belief theory and its application to deep learning. Information Fusion,101: #101987 [DOI: 10.1016/j.inffus.2023.101987]
Han K,Sheng V S,Song Y Q,Liu Y,Qiu C J,Ma S Q,et al. 2024. Deep semi-supervised learning for medical image segmentation: a review. Expert Systems with Applications,245: #123052 [DOI: 10.1016/j.eswa.2023.123052]
He Y P,Bi Y L,Li L J,Pun C M,Jiao W P and Jin Z. 2024. Mutual evidential deep learning for semi-supervised medical image segmentation//Proceedings of 2024 IEEE International Conference on Bioinformatics and Biomedicine (BIBM). Lisbon,Portugal: IEEE: 2010-2017 [DOI: 10.1109/BIBM62325.2024.10822008]
He Y P,Li L J,Zhan T X,Pun C M,Jiao W P and Jin Z. 2025a. Co-evidential fusion with information volume for semi-supervised medical image segmentation. Pattern Recognition,166: #111639 [DOI: 10.1016/j.patcog.2025.111639]
He Y P,Li L J,Zhan T X,Pun C M,Jiao W P and Jin Z. 2025b. Evidential prototype learning for semi-supervised medical image segmentation//Proceedings of the 31st ACM SIGKDD Conference on Knowledge Discovery and Data Mining V.2. Toronto,Canada: Association for Computing Machinery: 908-919 [DOI: 10.1145/3711896.3736944]
Hemmer P,Kühl N and Schöffer J. 2020. DEAL: deep evidential active learning for image classification//Proceedings of the 19th IEEE International Conference on Machine Learning and Applications (ICMLA). Miami,USA: IEEE: 865-870 [DOI: 10.1109/icmla51294.2020.00141]
Hoarau A,Lemaire V,Le Gall Y,Dubois J C and Martin A. 2024. Evidential uncertainty sampling strategies for active learning. Machine Learning,113(9): 6453-6474 [DOI: 10.1007/s10994-024-06567-2]
Huang B,Yang F,Yin M X,Mo X Y and Zhong C. 2020. A review of multimodal medical image fusion techniques. Computational and Mathematical Methods in Medicine,2020(1): #8279342 [DOI: 10.1155/2020/8279342]
Huang L,Denøeux T,Tonnelet D,Decazes P and Ruan S. 2021a. Deep PET/CT fusion with Dempster-Shafer theory for lymphoma segmentation//Proceedings of the 12th International Workshop on Machine Learning in Medical Imaging. Strasbourg,France: Springer: 30-39 [DOI: 10.1007/978-3-030-87589-3_4]
Huang L,Ruan S,Decazes P and Denœux T. 2025. Deep evidential fusion with uncertainty quantification and reliability learning for multimodal medical image segmentation. Information Fusion,113: #102648 [DOI: 10.1016/j.inffus.2024.102648]
Huang L,Ruan S and Denœux T. 2021b. Covid-19 classification with deep neural network and belief functions//Proceedings of the 5th International Conference on Biological Information and Biomedical Engineering. Hangzhou,China: Association for Computing Machinery: #3 [DOI: 10.1145/3469678.3469719]
Huang L,Ruan S and Denœux T. 2023. Semi-supervised multiple evidence fusion for brain tumor segmentation. Neurocomputing,535: 40-52 [DOI: 10.1016/j.neucom.2023.02.047]
Huang Q G,Chen Y F,Liu Y Q,Yue X D and Ma C. 2024. Trusted multi-view pancreatic tumor subtypes diagnosis through augmented opinion//Proceedings of 2024 IEEE International Conference on Bioinformatics and Biomedicine (BIBM). Lisbon,Portugal: IEEE: 3319-3322 [DOI: 10.1109/bibm62325.2024.10822127]
Jiao R S,Zhang Y C,Ding L,Xue B S,Zhang J C,Cai R,et al. 2024. Learning with limited annotations: a survey on deep semi-supervised learning for medical image segmentation. Computers in Biology and Medicine,169: #107840 [DOI: 10.1016/j.compbiomed.2023.107840]
Jøsang A. 2016. Subjective Logic. Cham,Germany: Springer [DOI: 10.1007/978-3-319-42337-1]
Jøsang A,Cho J H and Chen F. 2018. Uncertainty characteristics of subjective opinions//Proceedings of the 21st International Conference on Information Fusion (FUSION). Cambridge,UK: IEEE: 1998-2005 [DOI: 10.23919/icif.2018.8455454]
Karimi D and Gholipour A. 2023. Improving calibration and out-of-distribution detection in deep models for medical image segmentation. IEEE Transactions on Artificial Intelligence,4(2): 383-397 [DOI: 10.1109/TAI.2022.3159510]
Kendall A and Gal Y. 2017. What uncertainties do we need in Bayesian deep learning for computer vision?//Proceedings of the 31st International Conference on Neural Information Processing Systems. Long Beach,USA: Curran Associates Inc.: 5580-5590
Khosravi B,Faghani S and Ashraf-Ganjouei A. 2022. Uncertainty quantification in COVID-19 detection using evidential deep learning. medRxiv: 2022-05 [DOI: 10.1101/2022.05.29.22275732]
Koetzier L R,Wu J,Mastrodicasa D,Lutz A,Chung M,Koszek W A,et al. 2024. Generating synthetic data for medical imaging. Radiology,312(3): #e232471 [DOI: 10.1148/radiol.232471]
Lakshminarayanan B,Pritzel A and Blundell C. 2017. Simple and scalable predictive uncertainty estimation using deep ensembles//Proceedings of the 31st International Conference on Neural Information Processing Systems. Long Beach,USA: Curran Associates Inc.: 6402-6413
Lei T,Sun R,Du X G,Fu H Z,Zhang C Q and Nandi A K. 2023. SGU-Net: shape-guided ultralight network for abdominal image segmentation. IEEE Journal of Biomedical and Health Informatics,27(3): 1431-1442 [DOI: 10.1109/JBHI.2023.3238183]
Li F C,Gao S S,Liu Z,Zhang C M and Zhou Y F. 2025. Multimodal medical image fusion with progressive feature extraction and frequency domain information complementation. Journal of Image and Graphics,30(5): 1510-1527
李夫辰,高珊珊,刘峥,张彩明,周元峰. 2025. 渐进特征提取和频域信息补充的多模态医学图像融合. 中国图象图形学报,30(5): 1510-1527 [DOI: 10.11834/jig.240509]
Li H,Nan Y,Del Ser J and Yang G. 2023. Region-based evidential deep learning to quantify uncertainty and improve robustness of brain tumor segmentation. Neural Computing and Applications,35(30): 22071-22085 [DOI: 10.1007/s00521-022-08016-4]
Li P H,Kong X Y,He Y Z,Liu Y,Peng X,Li Z H,et al. 2022. Recent developments in application of single-cell RNA sequencing in the tumour immune microenvironment and cancer therapy. Military Medical Research,9(1): #52 [DOI: 10.1186/s40779-022-00414-y]
Liang X Y,Fu P H,Qian Y H,Guo Q and Liu G Q. 2025a. Trusted multi-view classification via evolutionary multi-view fusion//Proceedings of the 13th International Conference on Learning Representations. Singapore,Singapore: ICLR: 57147-57173
Liang X Y,Wang S J,Qian Y H,Guo Q,Du L,Jiang B B,et al. 2025b. Trusted multi-view classification with expert knowledge constraints//Proceedings of the 42nd International Conference on Machine Learning. Vancouver,Canada: PMLR: 37409-37426
Liu W,Chen Y F and Yue X D. 2024. Building trust in decision with conformalized multi-view deep classification//Proceedings of the 32nd ACM International Conference on Multimedia. Melbourne,Australia: Association for Computing Machinery: 7278-7287 [DOI: 10.1145/3664647.3681297]
Liu W,Chen Y F and Yue X D. 2025a. Enhancing testing-time robustness for trusted multi-view classification in the wild//Proceedings of 2025 Computer Vision and Pattern Recognition Conference. Nashville,USA: IEEE: 15508-15517 [DOI: 10.1109/cvpr52734.2025.01445]
Liu W,Chen Y F and Yue X D. 2025b. Enhancing multi-view classification reliability with adaptive rejection//Proceedings of the 39th AAAI Conference on Artificial Intelligence. Philadelphia,USA: AAAI: 18969-18977 [DOI: 10.1609/aaai.v39i18.34088]
Liu W,Chen Y F,Yue X D,Zhang C Q and Xie S R. 2023a. Safe multi-view deep classification//Proceedings of the 37th AAAI Conference on Artificial Intelligence. Washington USA: AAAI: 8870-8878 [DOI: 10.1609/aaai.v37i7.26066]
Liu W,Chen Y F,Yue X D,Zhang C Q and Xie S R. 2025c. Enhancing reliability in medical image classification of imperfect views. IEEE Transactions on Circuits and Systems for Video Technology,36(2): 2290-2303 [DOI: 10.1109/TCSVT.2025.3610762]
Liu W,Yue X D,Chen Y F and Denoeux T. 2022. Trusted multi-view deep learning with opinion aggregation//Proceedings of the 36th AAAI Conference on Artificial Intelligence. [s.l.]: AAAI: 7585-7593 [DOI: 10.1609/aaai.v36i7.20724]
Liu Y B,Li H N,Luo T,Zhang C Q,Xiao Z T,Wei Y,et al. 2023b. Structural attention graph neural network for diagnosis and prediction of COVID-19 severity. IEEE Transactions on Medical Imaging,42(2): 557-567 [DOI: 10.1109/TMI.2022.3226575]
Liu Y Y,Gao Z Y,Shi N N,Wu F P,Shi Y X,Chen Q C,et al. 2025d. MERIT: multi-view evidential learning for reliable and interpretable liver fibrosis staging. Medical Image Analysis,102: #103507 [DOI: 10.1016/j.media.2025.103507]
Long J W,Ren Y,Yang C X,Ren P C and Zeng Z Q. 2024. MDT: semi-supervised medical image segmentation with mixup-decoupling training. Physics in Medicine and Biology,69(6): #065012 [DOI: 10.1088/1361-6560/ad2715]
Nagahama A. 2023. Learning and predicting the unknown class using evidential deep learning. Scientific Reports,13(1): #14904 [DOI: 10.1038/s41598-023-40649-w]
Noor M H M and Ige A O. 2025. A survey on state-of-the-art deep learning applications and challenges. Engineering Applications of Artificial Intelligence,159: #111225 [DOI: 10.1016/j.engappai.2025.111225]
Pacal I,Kunduracioglu I,Alma M H,Deveci M,Kadry S,Nedoma J,et al. 2024. A systematic review of deep learning techniques for plant diseases. Artificial Intelligence Review,57(11): #304 [DOI: 10.1007/s10462-024-10944-7]
Pan Y S,Ma H J,Xia Y and Zhang Y N. 2025. Application of content generation in medical images. Journal of Image and Graphics,30(6): 1985-2000
潘永生,马豪杰,夏勇,张艳宁. 2025. 医学影像中的生成技术. 中国图象图形学报,30(6): 1985-2000 [DOI: 10.11834/jig.240459]
Pang S J,Zhao Y Z,Zhao R W and Feng R. 2025. Towards multi-label open set diagnosis for medical images//Proceedings of the 22nd IEEE International Symposium on Biomedical Imaging (ISBI). Houston,USA: IEEE: 1-5 [DOI: 10.1109/isbi60581.2025.10980926]
Pei J B,Men A,Liu Y,Zhuang X H and Chen Q C. 2024. Evidential multi-source-free unsupervised domain adaptation. IEEE Transactions on Pattern Analysis and Machine Intelligence,46(8): 5288-5305 [DOI: 10.1109/tpami.2024.3361978]
Pei X D,Zuo K,Li Y and Pang Z B. 2023. A review of the application of multi-modal deep learning in medicine: bibliometrics and future directions. International Journal of Computational Intelligence Systems,16(1): #44 [DOI: 10.1007/s44196-023-00225-6]
Penso C,Frenkel L and Goldberger J. 2024. Confidence calibration of a medical imaging classification system that is robust to label noise. IEEE Transactions on Medical Imaging,43(6): 2050-2060 [DOI: 10.1109/TMI.2024.3353762]
Qi Y X,Lin L,Zhang B,Zhang J Y and Wang J J. 2025. Multi-modal evidential fusion network for trustworthy PET/CT tumor segmentation. Knowledge-Based Systems,324: #113838 [DOI: 10.1016/j.knosys.2025.113838]
Qiu Z C,Miller D J and Kesidis G. 2017. A maximum entropy framework for semisupervised and active learning with unknown and label-scarce classes. IEEE Transactions on Neural Networks and Learning Systems,28(4): 917-933 [DOI: 10.1109/TNNLS.2016.2514401]
Qu J E,Chen Y F,Yue X D,Fu W and Huang Q G. 2024. Hyper-opinion evidential deep learning for out-of-distribution detection//Proceedings of the 38th International Conference on Neural Information Processing Systems. Vancouver,Canada: Curran Associates Inc.: #2687
Rayed M E,Islam S M S,Niha S I,Jim J R,Kabir M M and Mridha M F. 2024. Deep learning for medical image segmentation: state-of-the-art advancements and challenges. Informatics in Medicine Unlocked,47: #101504 [DOI: 10.1016/j.imu.2024.101504]
Ren K,Zou K,Liu X J,Chen Y D,Yuan X D,Shen X J,et al. 2023. Uncertainty-informed mutual learning for joint medical image classification and segmentation//Proceedings of the 26th International Conference on Medical Image Computing and Computer-Assisted Intervention. Vancouver,Canada: Springer: 35-45 [DOI: 10.1007/978-3-031-43901-8_4]
Roy A G,Ren J,Azizi S,Loh A,Natarajan V,Mustafa B,et al. 2022. Does your dermatology classifier know what it doesn’t know? Detecting the long-tail of unseen conditions. Medical Image Analysis,75: #102274 [DOI: 10.1016/j.media.2021.102274]
Sensoy M,Kaplan L M and Kandemir M. 2018. Evidential deep learning to quantify classification uncertainty//Proceedings of the 32nd International Conference on Neural Information Processing Systems. Montréal,Canada: Curran Associates Inc.: 3183-3193
Shafer G. 1976. A Mathematical Theory of Evidence. Princeton,USA: Princeton University Press
Sharifani K and Amini M. 2023. Machine learning and deep learning: a review of methods and applications. World Information Technology and Engineering Journal,10(7): 3897-3904
Shi J B,Li C,Gong T L and Fu H Z. 2024. E2-MIL: an explainable and evidential multiple instance learning framework for whole slide image classification. Medical Image Analysis,97: #103294 [DOI: 10.1016/j.media.2024.103294]
Suganyadevi S,Seethalakshmi V and Balasamy K. 2022. A review on deep learning in medical image analysis. International Journal of Multimedia Information Retrieval,11(1): 19-38 [DOI: 10.1007/s13735-021-00218-1]
Sun S C,Chen Y F,Yue X D,Ma C and Zhuang X H. 2026. REDNet: reliable evidential discounting network for multi-modality medical image segmentation. IEEE Transactions on Medical Imaging,45(1): 6-27 [DOI: 10.1109/tmi.2025.3591124]
Takahashi S,Sakaguchi Y,Kouno N,Takasawa K,Ishizu K,Akagi Y,et al. 2024. Comparison of vision transformers and convolutional neural networks in medical image analysis: a systematic review. Journal of Medical Systems,48(1): #84 [DOI: 10.1007/s10916-024-02105-8]
Tan H S,Wang K C and Mcbeth R. 2024a. Uncertainty-error correlations in evidential deep learning models for biomedical segmentation//Proceedings of the 29th International Conference on Technologies and Applications of Artificial Intelligence. Hsinchu,China: Springer: 91-105 [DOI: 10.1007/978-981-96-4589-3_7]
Tan H S,Wang K C and McBeth R. 2024b. Deep evidential learning for radiotherapy dose prediction. Computers in Biology and Medicine,182: #109172 [DOI: 10.1016/j.compbiomed.2024.109172]
Ulmer D,Meijerink L,Cin􀅣 G. 2020. Trust issues: uncertainty estimation does not enable reliable OOD detection on medical tabular data//Proceedings of the Machine Learning for Health Workshop,ML4H@NeurIPS 2020. [s.l.]: PMLR: 341-354
Wang M,Lin T,Lin A D,Yu K,Peng Y Y,Wang L Y,et al. 2025. Enhancing diagnostic accuracy in rare and common fundus diseases with a knowledge-rich vision-language model. Nature Communications,16(1): #5528 [DOI: 10.1038/s41467-025-60577-9]
Wang M,Lin T,Wang L Y,Lin A D,Zou K,Xu X X,et al. 2023. Uncertainty-inspired open set learning for retinal anomaly identification. Nature Communications,14(1): #6757 [DOI: 10.1038/s41467-023-42444-7]
Wu Q,Chen Y F,Liu W,Yue X D and Zhuang X H. 2024. Deep closing: enhancing topological connectivity in medical tubular segmentation. IEEE Transactions on Medical Imaging,43(11): 3990-4003 [DOI: 10.1109/TMI.2024.3405982]
Xia T,Dang T,Han J,Qendro L and Mascolo C. 2024. Uncertainty-aware health diagnostics via class-balanced evidential deep learning. IEEE Journal of Biomedical and Health Informatics,28(11): 6417-6428 [DOI: 10.1109/jbhi.2024.3360002]
Xia T,Han J,Qendro L,Dang T and Mascolo C. 2022. Hybrid-EDL: improving evidential deep learning for uncertainty quantification on imbalanced data//Proceedings of Workshop on Trustworthy and Socially Responsible Machine Learning,NeurIPS 2022. OrleansNew,USA: Curran Associates Inc.: 1-16
Xiao Z Y,Sun H and Liu F. 2024. Semi-supervised CT image segmentation via contrastive learning based on entropy constraints. Biomedical Engineering Letters,14(5): 1023-1035 [DOI: 10.1007/s13534-024-00387-y]
Xu L,Wu H,He C M,Wang J,Zhang C Q,Nie F P,et al. 2022a. Multi-modal sequence learning for Alzheimer’s disease progression prediction with incomplete variable-length longitudinal data. Medical Image Analysis,82: #102643 [DOI: 10.1016/j.media.2022.102643]
Xu S X,Chen Y F,Ma C and Yue X D. 2022b. Deep evidential fusion network for medical image classification. International Journal of Approximate Reasoning,150: 188-198 [DOI: 10.1016/j.ijar.2022.08.013]
Xu Z K,Yue X D,Lv Y,Liu W and Li Z H. 2023. Trusted fine-grained image classification through hierarchical evidence fusion//Proceedings of the 37th AAAI Conference on Artificial Intelligence. Washington,USA: AAAI: 10657-10665 [DOI: 10.1609/aaai.v37i9.26265]
Yan M,Lin Y,Peng X and Zeng Z. 2025. mixDA: mixup domain adaptation for glaucoma detection on fundus images. Neural Computing and Applications,37(11): 7541-7560 [DOI: 10.1007/s00521-023-08572-3]
Yao H Y,Han Z B,Fu H Z,Peng X,Hu Q H and Zhang C Q. 2024. Out-of-distribution detection with diversification (provably)//Proceedings of the 38th International Conference on Neural Information Processing Systems. Vancouver,Canada: Curran Associates Inc.: #714
Yeung M,Sala E,Schönlieb C B and Rundo L. 2022. Unified focal loss: generalising dice and cross entropy-based losses to handle class imbalanced medical image segmentation. Computerized Medical Imaging and Graphics,95: #102026 [DOI: 10.1016/j.compmedimag.2021.102026]
Zhang O,Delbrouck J B and Rubin D L. 2021. Out of distribution detection for medical images//Proceedings of the 3rd International Workshop on Uncertainty for Safe Utilization of Machine Learning in Medical Imaging,and Perinatal Imaging,Placental and Preterm Image Analysis. Strasbourg,France: Springer: 102-111 [DOI: 10.1007/978-3-030-87735-4_10]
Zhang Z X,Zhou H,Shi X R,Ran R,Tian C N and Zhou F. 2024a. Quality-driven deep cross-supervised learning network for semi-supervised medical image segmentation. Computers in Biology and Medicine,176: #108609 [DOI: 10.1016/j.compbiomed.2024.108609]
Zhang Z X,Zhou H,Shi X R,Ran R,Tian C N and Zhou F. 2024b. An evidential-enhanced tri-branch consistency learning method for semi-supervised medical image segmentation. IEEE Transactions on Instrumentation and Measurement,73: #2535713 [DOI: 10.1109/tim.2024.3488143]
Zhao R W,Xu Y Q,Zhang Y and Feng R. 2024. Evidential open set recognition for imbalanced medical images via multi-level data augmentation//Proceedings of 2024 IEEE International Conference on Bioinformatics and Biomedicine (BIBM). Lisbon,Portugal: IEEE: 4028-4031 [DOI: 10.1109/bibm62325.2024.10822558]
Zhao S K,Zhang X A,Pan L P,Xu X B and Pelusi D. 2025. Evidential deep active learning for semi-supervised classification [EB/OL]. [2025-09-15]. https://arxiv.org/pdf/2505.20691.pdf
Zhao X M,Shi P L,Wang D D,Fu Y Y,Zhang S Q and Fang J X. 2025. Distribution-aware mean teacher networks for semi-supervised medical image segmentation. Journal of Image and Graphics,30(2): 0575-0588
赵小明,石培炼,王丹丹,付有瑶,张石清,方江雄. 2025. 分布感知均值教师网络的半监督医学影像分割. 中国图象图形学报,30(2): 575-588 [DOI: 10.11834/jig.240208]
Zhao Y Q,Rao Y,Dong S P and Zhang J Y. 2020. Survey on deep learning object detection. Journal of image and graphics,25(4): 629-654
赵永强,饶元,董世鹏,张君毅. 2020. 深度学习目标检测方法综述. 中国图象图形学报,25(4): 629-654 [DOI: 10.11834/jig.190307]
Zhu H F,Qu Y,Ma Y T and Pei Y C. 2025. Evidential deep learning with reweighted margin adjustment for uncertainty-driven cervical OCT image diagnosis//Proceedings of 2025 IEEE International Conference on Acoustics,Speech and Signal Processing (ICASSP). Hyderabad,India: IEEE: 1-5 [DOI: 10.1109/icassp49660.2025.10890637]
Zhu T Y,Li K Z,Herrero P and Georgiou P. 2023. Personalized blood glucose prediction for type 1 diabetes using evidential deep learning and meta-learning. IEEE Transactions on Biomedical Engineering,70(1): 193-204 [DOI: 10.1109/tbme.2022.3187703]
Zou K,Lin T,Han Z B,Wang M,Yuan X D,Chen H Y,et al. 2024. Confidence-aware multi-modality learning for eye disease screening. Medical Image Analysis,96: #103214 [DOI: 10.1016/j.media.2024.103214]
Zou K,Yuan X D,Shen X J,Wang M and Fu H Z. 2022. TBraTS: trusted brain tumor segmentation//Proceedings of the 25th International Conference on Medical Image Computing and Computer-Assisted Intervention. Singapore,Singapore: Springer: 503-513 [DOI: 10.1007/978-3-031-16452-1_48]
2026年第31卷第7期
PDF下载
33
17
引用本文
BibTeX
文章信息
doi: 10.11834/jig.250475
  • 接收时间:2025-09-29
  • 首发时间:2026-10-30
  • 出版时间:2026-07-16
补充材料
相关文章
文章信息
作者
出版历史
  • 收稿日期:2025-09-29
  • 修回日期:2025-12-05
基金
National Natural Science Foundation of China (62472315,62476165)
国家自然科学基金项目(62472315,62476165)
作者信息
    1 同济大学计算机科学与技术学院,上海 201804
    2 上海大学人工智能研究院,上海 200444

通讯作者:

参考文献
分享链接
https://castjournals.cast.org.cn/joweb/zgtxtxxb/CN/10.11834/jig.250475
分享至
全文二维码

扫描看全文

引用本文
BibTeX
本文的引用情况
2种不同金属材料的力学参数
科
Family
属数
Number of
genus
种数
Number of
species
占总种数比例
Percentage of
total species (%)
属
Genus
种数
Number of
species
占总种数比例
Percentage of total
species (%)
鹅膏菌科Amanitaceae 2 11 5.26 鹅膏菌属 Amanita 10 4.78
小菇科 Mycenaceae 2 12 5.74 丝盖伞属 Inocybe 5 2.39
多孔菌科 Polyporaceae 8 14 6.70 蜡蘑属 Laccaria 5 2.39
红菇科 Russulaceae 3 23 11.00 小皮伞属 Marasmius 6 2.87
小菇属 Mycena 11 5.26
光柄菇属 Pluteus 5 2.39
红菇属 Russula 17 8.13
栓菌属 Trametes 5 2.39
关闭全屏