Article(id=1251893510224425537, tenantId=1146029695717560320, journalId=1251234473337991274, issueId=1251893504037831074, articleNumber=null, orderNo=null, doi=10.3969/j.issn.1003-3114.2025.05.012, pmid=null, cstr=null, oa=null, hot=null, price=null, onlineType=0, articleFormat=0, articleType=null, articleTypeStr=null, receivedDate=1745251200000, receivedDateStr=2025-04-22, revisedDate=null, revisedDateStr=null, acceptedDate=null, acceptedDateStr=null, onlineDate=1776404271894, onlineDateStr=2026-04-17, pubDate=1758124800000, pubDateStr=2025-09-18, doiRegisterDate=null, doiRegisterDateStr=null, onlineIssueDate=1776404271894, onlineIssueDateStr=2026-04-17, onlineJustAcceptDate=null, onlineJustAcceptDateStr=null, onlineFirstDate=null, onlineFirstDateStr=null, sourceXml=null, magXml=null, createTime=1776404271894, creator=13701087609, updateTime=1776404271894, updator=13701087609, issue=Issue{id=1251893504037831074, tenantId=1146029695717560320, journalId=1251234473337991274, year='2025', volume='51', issue='5', pageStart='877', pageEnd='1134', issueExtLink='null', onlineDate='null', pubDate='null', beforeIssueId=null, nextIssueId=null, price=null, status=1, issueComplete=1, articleOrder=1, issueType=1, specialIssue=null, createTime=1776404270419, creator=13701087609, updateTime=1776404832543, updator=13701087609, preIssue=null, nextIssue=null, ext={EN=IssueExt(id=1251895861849043019, tenantId=1146029695717560320, journalId=1251234473337991274, issueId=1251893504037831074, language=EN, specialIssueTitle=, coverIllustrator=null, specialIssueEditor=, specialIssueAbout=), CN=IssueExt(id=1251895861849043020, tenantId=1146029695717560320, journalId=1251234473337991274, issueId=1251893504037831074, language=CN, specialIssueTitle=, coverIllustrator=null, specialIssueEditor=, specialIssueAbout=)}, issueFiles=null}, startPage=993, endPage=1007, ext={EN=ArticleExt(id=1251893512770368122, articleId=1251893510224425537, tenantId=1146029695717560320, journalId=1251234473337991274, language=EN, title=Precise Skin Cancer Detection Based on Gamma Transform and Wavelet Convolution, columnId=1251893508886446519, journalTitle=Radio Communications Technology, columnName=Special Topic:Frontiers in Intelligent Communication, Storage, and Information Processing Technologies, runingTitle=null, highlight=null, articleAbstract=

Skin cancer and melanocytic nevus share numerous similarities, which can result in a misdiagnosis by dermatologists. To improve the screening accuracy of early skin cancer patients, the Gamma Transform Block (GMTB) based on Gamma Transform (GT) and Wavelet Convolution Block (WTCB) based on Wavelet Transform (WT) are proposed. Furthermore, the Space-Frequency Transform Network (SFTNet) for capturing fine-grained features of skin cancer is innovatively proposed based on the Detection Transformer(DETR) architecture. SFTNet-based skin cancer screening system can effectively improve disease detection accuracy because it enhances the sample image at different channels and reduces over-fitting effect during the model training process. Simulation results on HAM10000 dataset show that the accuracy of this system can reach 85.5%, which underscores the significant clinical value of our approach in skin cancer assisted diagnosis.

, correspAuthors=Lihong LI, 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, authorCompany=null, fund=null, authors=null, authorsList=Shiqiang BAI, Yiren CAI, Lihong LI), CN=ArticleExt(id=1251893522908001283, articleId=1251893510224425537, tenantId=1146029695717560320, journalId=1251234473337991274, language=CN, title=基于伽马变换与小波卷积的皮肤癌精确检测, columnId=1251893509079384505, journalTitle=无线电通信技术, columnName=专题:智能通信、存储与信息处理技术前沿, runingTitle=null, highlight=null, articleAbstract=

皮肤癌和黑素细胞痣的相似性较高,极易被皮肤科医师误诊。为提高早期皮肤癌患者的筛查准确度,基于伽马变换(Gamma Transform,GT)算法和小波变换(Wavelet Transform,WT)算法,分别构建伽马变换模块(Gamma Transform Block,GMTB)与小波卷积模块(Wavelet Convolution Block,WTCB),并在Detection Transformer(DETR)架构基础上创新性地提出用于捕捉皮肤癌细粒度特征的空频变换网络(Space-Frequency Transform Network,SFTNet)。包含SFT-Net的皮肤癌筛选系统能够增强样本图像的不同通道,减少模型训练过程中的过拟合效应,进而有效提升疾病检测精度。基于HAM10000数据集的仿真实验结果表明,系统正确率达85.5%,在皮肤癌辅助诊断方面具有重要的临床应用价值。

, correspAuthors=李丽宏, authorNote=null, correspAuthorsNote=
李丽宏 女,(1973—),博士,教授。主要研究方向:图像处理、计算机视觉。
, copyrightStatement=null, copyrightOwner=null, extLink=null, articleAbsUrl=null, sourceXml=d0oILx3olaE2vsuNTquQFg==, magXml=yot2YvryED08nlfuTgXw6g==, pdfUrl=null, pdf=P05/F1e91Nsyu6+VuuNoBA==, pdfFileSize=13780887, pdfExtLink=null, richHtmlUrl=null, mobilePdfUrl=null, reviewReport=null, pdfFirstPage=null, abstractGraph=J+DCkp1knKwkBfSonfJHVA==, abstractGraphContent=null, abstractVideo=null, citation=null, cebUrl=null, magXmlContent=S0Ca9f9ey8MT0/1Qbk/6Tg==, mapNumber=null, authorCompany=null, fund=null, authors=

白世强 男,(1999—),硕士研究生。主要研究方向:智能医学工程、医学影像感知。

蔡益人 男,(2000—),硕士研究生。主要研究方向:小样本分类。

, authorsList=白世强, 蔡益人, 李丽宏)}, authors=[Author(id=1251895520701137120, tenantId=1146029695717560320, journalId=1251234473337991274, articleId=1251893510224425537, orderNo=0, firstName=null, middleName=null, lastName=null, nameCn=null, orcid=null, stid=null, country=null, authorPic=null, dead=0, email=null, emailSecond=null, emailThird=null, correspondingAuthor=0, authorType=1, ext={EN=AuthorExt(id=1251895520822771944, tenantId=1146029695717560320, journalId=1251234473337991274, articleId=1251893510224425537, authorId=1251895520701137120, language=EN, stringName=Shiqiang BAI, firstName=Shiqiang, middleName=null, lastName=BAI, prefix=null, suffix=null, authorComment=null, nameInitials=null, affiliation=null, department=null, xref=null, address=College of Information and Electrical Engineering, Hebei University of Engineering, Handan 056038, China, bio=null, bioImg=null, bioContent=null, aboutCorrespAuthor=null), CN=AuthorExt(id=1251895520969572592, tenantId=1146029695717560320, journalId=1251234473337991274, articleId=1251893510224425537, authorId=1251895520701137120, language=CN, stringName=白世强, firstName=null, middleName=null, lastName=null, prefix=null, suffix=null, authorComment=null, nameInitials=null, affiliation=null, department=null, xref=null, address=河北工程大学 信息与电气工程学院,河北 邯郸 056038, bio={"content":"

白世强 男,(1999—),硕士研究生。主要研究方向:智能医学工程、医学影像感知。

"}, bioImg=null, bioContent=

白世强 男,(1999—),硕士研究生。主要研究方向:智能医学工程、医学影像感知。

, aboutCorrespAuthor=null)}, companyList=[AuthorCompany(id=1251895520554336469, tenantId=1146029695717560320, journalId=1251234473337991274, articleId=1251893510224425537, xref=null, ext=[AuthorCompanyExt(id=1251895520562725078, tenantId=1146029695717560320, journalId=1251234473337991274, articleId=1251893510224425537, companyId=1251895520554336469, language=EN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=College of Information and Electrical Engineering, Hebei University of Engineering, Handan 056038, China), AuthorCompanyExt(id=1251895520571113688, tenantId=1146029695717560320, journalId=1251234473337991274, articleId=1251893510224425537, companyId=1251895520554336469, language=CN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=河北工程大学 信息与电气工程学院,河北 邯郸 056038)])]), Author(id=1251895521070235893, tenantId=1146029695717560320, journalId=1251234473337991274, articleId=1251893510224425537, orderNo=1, firstName=null, middleName=null, lastName=null, nameCn=null, orcid=null, stid=null, country=null, authorPic=null, dead=0, email=null, emailSecond=null, emailThird=null, correspondingAuthor=0, authorType=1, ext={EN=AuthorExt(id=1251895521162510589, tenantId=1146029695717560320, journalId=1251234473337991274, articleId=1251893510224425537, authorId=1251895521070235893, language=EN, stringName=Yiren CAI, firstName=Yiren, middleName=null, lastName=CAI, prefix=null, suffix=null, authorComment=null, nameInitials=null, affiliation=null, department=null, xref=null, address=College of Information and Electrical Engineering, Hebei University of Engineering, Handan 056038, China, bio=null, bioImg=null, bioContent=null, aboutCorrespAuthor=null), CN=AuthorExt(id=1251895521246396675, tenantId=1146029695717560320, journalId=1251234473337991274, articleId=1251893510224425537, authorId=1251895521070235893, language=CN, stringName=蔡益人, firstName=null, middleName=null, lastName=null, prefix=null, suffix=null, authorComment=null, nameInitials=null, affiliation=null, department=null, xref=null, address=河北工程大学 信息与电气工程学院,河北 邯郸 056038, bio={"content":"

蔡益人 男,(2000—),硕士研究生。主要研究方向:小样本分类。

"}, bioImg=null, bioContent=

蔡益人 男,(2000—),硕士研究生。主要研究方向:小样本分类。

, aboutCorrespAuthor=null)}, companyList=[AuthorCompany(id=1251895520554336469, tenantId=1146029695717560320, journalId=1251234473337991274, articleId=1251893510224425537, xref=null, ext=[AuthorCompanyExt(id=1251895520562725078, tenantId=1146029695717560320, journalId=1251234473337991274, articleId=1251893510224425537, companyId=1251895520554336469, language=EN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=College of Information and Electrical Engineering, Hebei University of Engineering, Handan 056038, China), AuthorCompanyExt(id=1251895520571113688, tenantId=1146029695717560320, journalId=1251234473337991274, articleId=1251893510224425537, companyId=1251895520554336469, language=CN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=河北工程大学 信息与电气工程学院,河北 邯郸 056038)])]), Author(id=1251895521330282761, tenantId=1146029695717560320, journalId=1251234473337991274, articleId=1251893510224425537, orderNo=2, firstName=null, middleName=null, lastName=null, nameCn=null, orcid=null, stid=null, country=null, authorPic=null, dead=0, email=null, emailSecond=null, emailThird=null, correspondingAuthor=1, authorType=1, ext={EN=AuthorExt(id=1251895521430946065, tenantId=1146029695717560320, journalId=1251234473337991274, articleId=1251893510224425537, authorId=1251895521330282761, language=EN, stringName=Lihong LI, firstName=Lihong, middleName=null, lastName=LI, prefix=null, suffix=null, authorComment=null, nameInitials=null, affiliation=null, department=null, xref=*, address=College of Information and Electrical Engineering, Hebei University of Engineering, Handan 056038, China, bio=null, bioImg=null, bioContent=null, aboutCorrespAuthor=null), CN=AuthorExt(id=1251895521527415065, tenantId=1146029695717560320, journalId=1251234473337991274, articleId=1251893510224425537, authorId=1251895521330282761, language=CN, stringName=李丽宏, firstName=null, middleName=null, lastName=null, prefix=null, suffix=null, authorComment=null, nameInitials=null, affiliation=null, department=null, xref=*, address=河北工程大学 信息与电气工程学院,河北 邯郸 056038, bio=null, bioImg=null, bioContent=null, aboutCorrespAuthor=null)}, companyList=[AuthorCompany(id=1251895520554336469, tenantId=1146029695717560320, journalId=1251234473337991274, articleId=1251893510224425537, xref=null, ext=[AuthorCompanyExt(id=1251895520562725078, tenantId=1146029695717560320, journalId=1251234473337991274, articleId=1251893510224425537, companyId=1251895520554336469, language=EN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=College of Information and Electrical Engineering, Hebei University of Engineering, Handan 056038, China), AuthorCompanyExt(id=1251895520571113688, tenantId=1146029695717560320, journalId=1251234473337991274, articleId=1251893510224425537, companyId=1251895520554336469, language=CN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=河北工程大学 信息与电气工程学院,河北 邯郸 056038)])])], keywords=[Keyword(id=1251895521732935970, tenantId=1146029695717560320, journalId=1251234473337991274, articleId=1251893510224425537, language=EN, orderNo=1, keyword=deep learning), Keyword(id=1251895521825210666, tenantId=1146029695717560320, journalId=1251234473337991274, articleId=1251893510224425537, language=EN, orderNo=2, keyword=medical image processing), Keyword(id=1251895521938456878, tenantId=1146029695717560320, journalId=1251234473337991274, articleId=1251893510224425537, language=EN, orderNo=3, keyword=skin cancer detection), Keyword(id=1251895522060091701, tenantId=1146029695717560320, journalId=1251234473337991274, articleId=1251893510224425537, language=EN, orderNo=4, keyword=GT), Keyword(id=1251895522164949306, tenantId=1146029695717560320, journalId=1251234473337991274, articleId=1251893510224425537, language=EN, orderNo=5, keyword=wavelet convolution), Keyword(id=1251895522261418302, tenantId=1146029695717560320, journalId=1251234473337991274, articleId=1251893510224425537, language=EN, orderNo=6, keyword=SFTNet), Keyword(id=1251895522332721472, tenantId=1146029695717560320, journalId=1251234473337991274, articleId=1251893510224425537, language=CN, orderNo=1, keyword=深度学习), Keyword(id=1251895522429190471, tenantId=1146029695717560320, journalId=1251234473337991274, articleId=1251893510224425537, language=CN, orderNo=2, keyword=医学影像处理), Keyword(id=1251895522517270861, tenantId=1146029695717560320, journalId=1251234473337991274, articleId=1251893510224425537, language=CN, orderNo=3, keyword=皮肤癌检测), Keyword(id=1251895522643099987, tenantId=1146029695717560320, journalId=1251234473337991274, articleId=1251893510224425537, language=CN, orderNo=4, keyword=伽马变换), Keyword(id=1251895522743763288, tenantId=1146029695717560320, journalId=1251234473337991274, articleId=1251893510224425537, language=CN, orderNo=5, keyword=小波卷积), Keyword(id=1251895522836037981, tenantId=1146029695717560320, journalId=1251234473337991274, articleId=1251893510224425537, language=CN, orderNo=6, keyword=空频变换网络)], refs=[Reference(id=1251895529500787265, tenantId=1146029695717560320, journalId=1251234473337991274, articleId=1251893510224425537, doi=null, pmid=null, pmcid=null, year=2025-04-18, volume=null, issue=null, pageStart=null, pageEnd=null, url=null, language=null, rfNumber=[1], rfOrder=0, authorNames=穆亚楠, 陈星会, 何璐, journalName=null, refType=null, unstructuredReference=穆亚楠,陈星会,何璐,.常见非黑色素瘤皮肤癌新兴治疗的研究进展[J/OL]. (2024-11-18)[2025-04-18]. http://kns.cnki.Net/kcms/detail/61.1415.R.20241115.2350.007.html., articleTitle=常见非黑色素瘤皮肤癌新兴治疗的研究进展, refAbstract=null), Reference(id=1251895529647587915, tenantId=1146029695717560320, journalId=1251234473337991274, articleId=1251893510224425537, doi=null, pmid=null, pmcid=null, year=2024, volume=null, issue=null, pageStart=null, pageEnd=null, url=null, language=null, rfNumber=[2], rfOrder=1, authorNames=崔勇, 高兴华, journalName=皮肤病性病学, refType=null, unstructuredReference=崔勇,高兴华.皮肤病性病学[M]. 10版.北京:人民卫生出版社,2024., articleTitle=null, refAbstract=null), Reference(id=1251895529811165782, tenantId=1146029695717560320, journalId=1251234473337991274, articleId=1251893510224425537, doi=null, pmid=null, pmcid=null, year=2023, volume=null, issue=null, pageStart=null, pageEnd=null, url=null, language=null, rfNumber=[3], rfOrder=2, authorNames=刘道远, journalName=皮肤镜诊断常见非黑色素瘤皮肤癌的临床价值研究, refType=null, unstructuredReference=刘道远.皮肤镜诊断常见非黑色素瘤皮肤癌的临床价值研究[D].昆明:昆明医科大学,2023., articleTitle=null, refAbstract=null), Reference(id=1251895529924411998, tenantId=1146029695717560320, journalId=1251234473337991274, articleId=1251893510224425537, doi=null, pmid=null, pmcid=null, year=2021, volume=null, issue=null, pageStart=null, pageEnd=null, url=null, language=null, rfNumber=[4], rfOrder=3, authorNames=邓凡智, journalName=皮肤镜在基底细胞癌诊断中的应用价值, refType=null, unstructuredReference=邓凡智.皮肤镜在基底细胞癌诊断中的应用价值[D].蚌埠:蚌埠医学院,2021., articleTitle=null, refAbstract=null), Reference(id=1251895530041852518, tenantId=1146029695717560320, journalId=1251234473337991274, articleId=1251893510224425537, doi=null, pmid=null, pmcid=null, year=2021, volume=50, issue=4, pageStart=428, pageEnd=431, url=null, language=null, rfNumber=[5], rfOrder=4, authorNames=崔莉, 王贵佐, 马小萍, journalName=陕西医学杂志, refType=null, unstructuredReference=崔莉,王贵佐,马小萍,.皮肤镜检查在黑素细胞痣、脂溢性角化病和基底细胞癌中的诊断价值[J].陕西医学杂志,2021,50(4):428-431., articleTitle=皮肤镜检查在黑素细胞痣、脂溢性角化病和基底细胞癌中的诊断价值, refAbstract=null), Reference(id=1251895530159293038, tenantId=1146029695717560320, journalId=1251234473337991274, articleId=1251893510224425537, doi=null, pmid=null, pmcid=null, year=2019, volume=26, issue=4, pageStart=204, pageEnd=209, url=null, language=null, rfNumber=[6], rfOrder=5, authorNames=江浩波, 孙志平, 陈海针, journalName=皮肤性病诊疗学杂志, refType=null, unstructuredReference=江浩波,孙志平,陈海针,.皮肤镜在脂溢性角化、日光性角化和基底细胞癌诊断中的应用[J].皮肤性病诊疗学杂志,2019,26(4):204-209., articleTitle=皮肤镜在脂溢性角化、日光性角化和基底细胞癌诊断中的应用, refAbstract=null), Reference(id=1251895530268344948, tenantId=1146029695717560320, journalId=1251234473337991274, articleId=1251893510224425537, doi=null, pmid=null, pmcid=null, year=2016, volume=null, issue=null, pageStart=21, pageEnd=37, url=null, language=null, rfNumber=[7], rfOrder=6, authorNames=LIU W, ANGUEVOV D, ERHAN D, journalName=null, refType=null, unstructuredReference=LIU W,ANGUEVOV D,ERHAN D,et al. SSD:Single Shot MultiBox Detector[C]//European Conference on Computer Vision. Cham:Springer,2016:21-37., articleTitle=SSD:Single Shot MultiBox Detector, refAbstract=null), Reference(id=1251895530394174075, tenantId=1146029695717560320, journalId=1251234473337991274, articleId=1251893510224425537, doi=null, pmid=null, pmcid=null, year=2014, volume=null, issue=null, pageStart=580, pageEnd=587, url=null, language=null, rfNumber=[8], rfOrder=7, authorNames=GIRSHICK R, DONAHUE J, DARRELL T, journalName=null, refType=null, unstructuredReference=GIRSHICK R, DONAHUE J, DARRELL T, et al. Rich Feature Hierarchies for Accurate Object Detection and Semantic Segmentation[C]//Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition. Columbus:IEEE,2014:580-587., articleTitle=Rich Feature Hierarchies for Accurate Object Detection and Semantic Segmentation, refAbstract=null), Reference(id=1251895530499031679, tenantId=1146029695717560320, journalId=1251234473337991274, articleId=1251893510224425537, doi=null, pmid=null, pmcid=null, year=2015, volume=null, issue=null, pageStart=1440, pageEnd=1448, url=null, language=null, rfNumber=[9], rfOrder=8, authorNames=GIRSHICK R, journalName=null, refType=null, unstructuredReference=GIRSHICK R. Fast R-CNN[C]//Proceedings of the 2015 IEEE International Conference on Computer Vision. Santiago:IEEE,2015:1440-1448., articleTitle=Fast R-CNN, refAbstract=null), Reference(id=1251895530620666502, tenantId=1146029695717560320, journalId=1251234473337991274, articleId=1251893510224425537, doi=null, pmid=null, pmcid=null, year=2015, volume=28, issue=null, pageStart=1137, pageEnd=1149, url=null, language=null, rfNumber=[10], rfOrder=9, authorNames=REN S Q, HE K M, GIRSHICK R, journalName=Advances in Neural Information Processing Systems. IEEE, refType=null, unstructuredReference=REN S Q,HE K M,GIRSHICK R,et al. Faster R-CNN:Towards Real-time Object Detection with Region Proposal Networks[J]. Advances in Neural Information Processing Systems. IEEE,2015,28:1137-1149., articleTitle=Faster R-CNN:Towards Real-time Object Detection with Region Proposal Networks, refAbstract=null), Reference(id=1251895530712941194, tenantId=1146029695717560320, journalId=1251234473337991274, articleId=1251893510224425537, doi=null, pmid=null, pmcid=null, year=2016, volume=null, issue=null, pageStart=779, pageEnd=788, url=null, language=null, rfNumber=[11], rfOrder=10, authorNames=REDMON J, DIVVALA S, GIRSHICK R, journalName=null, refType=null, unstructuredReference=REDMON J,DIVVALA S,GIRSHICK R,et al. You Only Look Once:Unified,Real-time Object Detection[C]//Proceedings of the 2016 IEEE Conference on Computer Vision and Pattern Recognition. Las Vegas:IEEE,2016:779-788., articleTitle=You Only Look Once:Unified,Real-time Object Detection, refAbstract=null), Reference(id=1251895530872324752, tenantId=1146029695717560320, journalId=1251234473337991274, articleId=1251893510224425537, doi=null, pmid=null, pmcid=null, year=2017, volume=30, issue=null, pageStart=5998, pageEnd=6008, url=null, language=null, rfNumber=[12], rfOrder=11, authorNames=VASWANI A, SHAZEER N, PARMAR N, journalName=Advances in Neural Information Processing Systems, refType=null, unstructuredReference=VASWANI A,SHAZEER N,PARMAR N,et al. Attention is All You Need[J]. Advances in Neural Information Processing Systems,2017,30:5998-6008., articleTitle=Attention is All You Need, refAbstract=null), Reference(id=1251895530952016532, tenantId=1146029695717560320, journalId=1251234473337991274, articleId=1251893510224425537, doi=null, pmid=null, pmcid=null, year=2020, volume=null, issue=null, pageStart=213, pageEnd=229, url=null, language=null, rfNumber=[13], rfOrder=12, authorNames=CARION N, MASSA F, SYNNAEVE G, journalName=null, refType=null, unstructuredReference=CARION N,MASSA F,SYNNAEVE G,et al. End-to-End Object Detection with Transformers[C]//Proceedings of the 2020 European Conference on Computer Vision. Glasgow:Springer,2020:213-229., articleTitle=End-to-End Object Detection with Transformers, refAbstract=null), Reference(id=1251895531082039963, tenantId=1146029695717560320, journalId=1251234473337991274, articleId=1251893510224425537, doi=null, pmid=null, pmcid=null, year=2024, volume=39, issue=11, pageStart=3597, pageEnd=3606, url=null, language=null, rfNumber=[14], rfOrder=13, authorNames=赵嘉晖, 温杰, 蔡星娟, journalName=控制与决策, refType=null, unstructuredReference=赵嘉晖,温杰,蔡星娟,.基于高维多目标优化的小样本皮肤癌检测[J].控制与决策, 2024,39(11):3597-3606., articleTitle=基于高维多目标优化的小样本皮肤癌检测, refAbstract=null), Reference(id=1251895531207869092, tenantId=1146029695717560320, journalId=1251234473337991274, articleId=1251893510224425537, doi=null, pmid=null, pmcid=null, year=2024, volume=null, issue=null, pageStart=null, pageEnd=null, url=null, language=null, rfNumber=[15], rfOrder=14, authorNames=王煜坤, journalName=皮肤影像及人工智能辅助皮肤疾病精准诊疗新探索, refType=null, unstructuredReference=王煜坤.皮肤影像及人工智能辅助皮肤疾病精准诊疗新探索[D].北京:北京协和医学院,2024., articleTitle=null, refAbstract=null), Reference(id=1251895531291755179, tenantId=1146029695717560320, journalId=1251234473337991274, articleId=1251893510224425537, doi=null, pmid=null, pmcid=null, year=2020, volume=26, issue=8, pageStart=1229, pageEnd=1234, url=null, language=null, rfNumber=[16], rfOrder=15, authorNames=TSCHANDL P, RINNER C, APALLA Z, journalName=Nature Medicine, refType=null, unstructuredReference=TSCHANDL P,RINNER C,APALLA Z,et al. Human computer Collaboration for Skin Cancer Recognition[J]. Nature Medicine,2020,26(8):1229-1234., articleTitle=Human computer Collaboration for Skin Cancer Recognition, refAbstract=null), Reference(id=1251895531384029874, tenantId=1146029695717560320, journalId=1251234473337991274, articleId=1251893510224425537, doi=null, pmid=null, pmcid=null, year=2020, volume=null, issue=null, pageStart=null, pageEnd=null, url=null, language=null, rfNumber=[17], rfOrder=16, authorNames=GONZALEZ R C, WOODS R E, journalName=Digital Image Processing, refType=null, unstructuredReference=GONZALEZ R C,WOODS R E. Digital Image Processing[M].北京:电子工业出版社,2020., articleTitle=null, refAbstract=null), Reference(id=1251895531446944440, tenantId=1146029695717560320, journalId=1251234473337991274, articleId=1251893510224425537, doi=null, pmid=null, pmcid=null, year=2022, volume=null, issue=null, pageStart=null, pageEnd=null, url=null, language=null, rfNumber=[18], rfOrder=17, authorNames=禹晶, 肖创柏, 廖庆敏, journalName=数字图像处理, refType=null, unstructuredReference=禹晶,肖创柏,廖庆敏.数字图像处理[M].北京:清华大学出版社,2022., articleTitle=null, refAbstract=null), Reference(id=1251895531564384961, tenantId=1146029695717560320, journalId=1251234473337991274, articleId=1251893510224425537, doi=null, pmid=null, pmcid=null, year=1822, volume=null, issue=null, pageStart=null, pageEnd=null, url=null, language=null, rfNumber=[19], rfOrder=18, authorNames=FOURIER J B J, journalName=Théorie Analytique de la Chaleur, refType=null, unstructuredReference=FOURIER J B J. Théorie Analytique de la Chaleur[M]. Paris:Firmin Didot,1822., articleTitle=null, refAbstract=null), Reference(id=1251895531656659653, tenantId=1146029695717560320, journalId=1251234473337991274, articleId=1251893510224425537, doi=null, pmid=null, pmcid=null, year=1974, volume=23, issue=1, pageStart=90, pageEnd=93, url=null, language=null, rfNumber=[20], rfOrder=19, authorNames=AHMED N, NATARAJAN T, RAO K R, journalName=IEEE Transactions on Computers, refType=null, unstructuredReference=AHMED N, NATARAJAN T, RAO K R. Discrete Cosine Transform[J]. IEEE Transactions on Computers,1974,23(1):90-93., articleTitle=Discrete Cosine Transform, refAbstract=null), Reference(id=1251895531816043214, tenantId=1146029695717560320, journalId=1251234473337991274, articleId=1251893510224425537, doi=null, pmid=null, pmcid=null, year=1972, volume=20, issue=2, pageStart=140, pageEnd=146, url=null, language=null, rfNumber=[21], rfOrder=20, authorNames=DEVANEY A J, journalName=IEEE Transactions on Audio and Electroacoustics, refType=null, unstructuredReference=DEVANEY A J. The Slant Transform: An Orthogonal Transform for Signal Processing[J]. IEEE Transactions on Audio and Electroacoustics,1972,20(2):140-146., articleTitle=The Slant Transform: An Orthogonal Transform for Signal Processing, refAbstract=null), Reference(id=1251895531933483728, tenantId=1146029695717560320, journalId=1251234473337991274, articleId=1251893510224425537, doi=null, pmid=null, pmcid=null, year=1923, volume=25, issue=1, pageStart=5, pageEnd=24, url=null, language=null, rfNumber=[22], rfOrder=21, authorNames=WALSH J L, journalName=Transactions of the American Mathematical Society, refType=null, unstructuredReference=WALSH J L. A Closed Set of Normal Orthogonal Functions[J]. Transactions of the American Mathematical Society,1923,25(1):5-24., articleTitle=A Closed Set of Normal Orthogonal Functions, refAbstract=null), Reference(id=1251895532004786901, tenantId=1146029695717560320, journalId=1251234473337991274, articleId=1251893510224425537, doi=null, pmid=null, pmcid=null, year=1992, volume=null, issue=null, pageStart=null, pageEnd=null, url=null, language=null, rfNumber=[23], rfOrder=22, authorNames=DAUBECHIES I, journalName=Ten Lectures on Wavelets, refType=null, unstructuredReference=DAUBECHIES I. Ten Lectures on Wavelets[M]. Philadelphia:Society for Industrial and Applied Mathematics, 1992., articleTitle=null, refAbstract=null), Reference(id=1251895532092867291, tenantId=1146029695717560320, journalId=1251234473337991274, articleId=1251893510224425537, doi=null, pmid=null, pmcid=null, year=2022, volume=60, issue=3, pageStart=21, pageEnd=23, url=null, language=null, rfNumber=[24], rfOrder=23, authorNames=李艳玲, journalName=中国现代医生, refType=null, unstructuredReference=李艳玲.皮肤镜在颜面部皮肤肿瘤筛检中的应用[J].中国现代医生,2022,60(3):21-23., articleTitle=皮肤镜在颜面部皮肤肿瘤筛检中的应用, refAbstract=null), Reference(id=1251895532193530594, tenantId=1146029695717560320, journalId=1251234473337991274, articleId=1251893510224425537, doi=null, pmid=null, pmcid=null, year=2018, volume=5, issue=null, pageStart=1, pageEnd=10, url=null, language=null, rfNumber=[25], rfOrder=24, authorNames=TSCHANDL P, ROSENDAHL C, KITTLER H, journalName=Scientific Data, refType=null, unstructuredReference=TSCHANDL P, ROSENDAHL C, KITTLER H. The HAM10000 Dataset, a Large Collection of Multi-source Dermatoscopic Images of Common Pigmented Skin Lesions[J]. Scientific Data,2018,5:1-10., articleTitle=The HAM10000 Dataset, a Large Collection of Multi-source Dermatoscopic Images of Common Pigmented Skin Lesions, refAbstract=null), Reference(id=1251895533774783209, tenantId=1146029695717560320, journalId=1251234473337991274, articleId=1251893510224425537, doi=null, pmid=null, pmcid=null, year=2008, volume=9, issue=86, pageStart=2579, pageEnd=2605, url=null, language=null, rfNumber=[26], rfOrder=25, authorNames=LAURENS V D M, HINTON G E, journalName=Journal of Machine Learning Research, refType=null, unstructuredReference=LAURENS V D M,HINTON G E. Visualizing Data Using t-SNE[J]. Journal of Machine Learning Research,2008,9(86):2579-2605., articleTitle=Visualizing Data Using t-SNE, refAbstract=null), Reference(id=1251895533858669296, tenantId=1146029695717560320, journalId=1251234473337991274, articleId=1251893510224425537, doi=null, pmid=null, pmcid=null, year=1998, volume=86, issue=11, pageStart=2278, pageEnd=2324, url=null, language=null, rfNumber=[27], rfOrder=26, authorNames=LECUN Y, BOTTOU L, BENGIO Y, journalName=Proceedings of the IEEE, refType=null, unstructuredReference=LECUN Y,BOTTOU L,BENGIO Y,et al. Gradient-based Learning Applied to Document Recognition[J]. Proceedings of the IEEE,1998,86(11):2278-2324., articleTitle=Gradient-based Learning Applied to Document Recognition, refAbstract=null), Reference(id=1251895533938361077, tenantId=1146029695717560320, journalId=1251234473337991274, articleId=1251893510224425537, doi=null, pmid=null, pmcid=null, year=2012, volume=null, issue=null, pageStart=1097, pageEnd=1105, url=null, language=null, rfNumber=[28], rfOrder=27, authorNames=KRIZHEVSKY A, SUTSKEVER I, HINTON G E, journalName=null, refType=null, unstructuredReference=KRIZHEVSKY A,SUTSKEVER I,HINTON G E. Imagenet Classification with Deep Convolutional Neural Networks[C]//Advances in Neural Information Processing Systems. Lake Tahoe:Curran Associates Inc.,2012:1097-1105., articleTitle=Imagenet Classification with Deep Convolutional Neural Networks, refAbstract=null), Reference(id=1251895534106133242, tenantId=1146029695717560320, journalId=1251234473337991274, articleId=1251893510224425537, doi=null, pmid=null, pmcid=null, year=2015, volume=null, issue=null, pageStart=1, pageEnd=14, url=null, language=null, rfNumber=[29], rfOrder=28, authorNames=SIMONYAN K, ZISSERMAN A, journalName=null, refType=null, unstructuredReference=SIMONYAN K,ZISSERMAN A. Very Deep Convolutional Networks for Large-scale Image Recognition[C]//Proceedings of the 3rd International Conference on Learning Representations (ICLR 2015). San Diego:OpenReview, 2015:1-14., articleTitle=Very Deep Convolutional Networks for Large-scale Image Recognition, refAbstract=null), Reference(id=1251895534219379460, tenantId=1146029695717560320, journalId=1251234473337991274, articleId=1251893510224425537, doi=null, pmid=null, pmcid=null, year=2015, volume=null, issue=null, pageStart=1, pageEnd=9, url=null, language=null, rfNumber=[30], rfOrder=29, authorNames=SZEGEDY C, LIU W, JIA Y Q, journalName=null, refType=null, unstructuredReference=SZEGEDY C,LIU W,JIA Y Q,et al. Going Deeper with Convolutions[C]//Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition. Boston:IEEE, 2015:1-9., articleTitle=Going Deeper with Convolutions, refAbstract=null), Reference(id=1251895534299071241, tenantId=1146029695717560320, journalId=1251234473337991274, articleId=1251893510224425537, doi=null, pmid=null, pmcid=null, year=2015, volume=null, issue=null, pageStart=1026, pageEnd=1034, url=null, language=null, rfNumber=[31], rfOrder=30, authorNames=HE K M, ZHANG X Y, REN S Q, journalName=null, refType=null, unstructuredReference=HE K M,ZHANG X Y,REN S Q,et al. Delving Deep into Rectifiers:Surpassing Human-level Performance on Imagenet Classification[C]//Proceedings of the IEEE International Conference on Computer Vision. Santiago:IEEE, 2015:1026-1034., articleTitle=Delving Deep into Rectifiers:Surpassing Human-level Performance on Imagenet Classification, refAbstract=null), Reference(id=1251895534382957324, tenantId=1146029695717560320, journalId=1251234473337991274, articleId=1251893510224425537, doi=null, pmid=null, pmcid=null, year=2021, volume=null, issue=null, pageStart=191, pageEnd=195, url=null, language=null, rfNumber=[32], rfOrder=31, authorNames=YANG J C, SHI R, NI B B, journalName=null, refType=null, unstructuredReference=YANG J C,SHI R,NI B B. MedMNIST Classification Decathlon:A Lightweight Automl Benchmark for Medical Image Analysis[C]//IEEE 18th International Symposium on Biomedical Imaging(ISBI). Piscataway:IEEE,2021:191-195., articleTitle=MedMNIST Classification Decathlon:A Lightweight Automl Benchmark for Medical Image Analysis, refAbstract=null), Reference(id=1251895534454260496, tenantId=1146029695717560320, journalId=1251234473337991274, articleId=1251893510224425537, doi=null, pmid=null, pmcid=null, year=2023, volume=10, issue=1, pageStart=1, pageEnd=41, url=null, language=null, rfNumber=[33], rfOrder=32, authorNames=YANG J C, SHI R, WEI D L, journalName=Scientific Data, refType=null, unstructuredReference=YANG J C, SHI R, WEI D L, et al. MedMNIST v2: A Large-scale Lightweight Benchmark for 2D and 3D Biomedical Image Classification[J]. Scientific Data,2023, 10(1):1-41., articleTitle=MedMNIST v2: A Large-scale Lightweight Benchmark for 2D and 3D Biomedical Image Classification, refAbstract=null), Reference(id=1251895534542340886, tenantId=1146029695717560320, journalId=1251234473337991274, articleId=1251893510224425537, doi=null, pmid=null, pmcid=null, year=2015, volume=null, issue=null, pageStart=2962, pageEnd=2970, url=null, language=null, rfNumber=[34], rfOrder=33, authorNames=FEURER M, KLEIN A, EGGENSPERGER K, journalName=null, refType=null, unstructuredReference=FEURER M,KLEIN A,EGGENSPERGER K,et al. Efficient and Robust Automated Machine Learning[C]//Advances in Neural Information Processing Systems. Montreal:Curran Associates Inc.,2015:2962-2970., articleTitle=Efficient and Robust Automated Machine Learning, refAbstract=null), Reference(id=1251895534663975708, tenantId=1146029695717560320, journalId=1251234473337991274, articleId=1251893510224425537, doi=null, pmid=null, pmcid=null, year=2018, volume=null, issue=null, pageStart=1946, pageEnd=1956, url=null, language=null, rfNumber=[35], rfOrder=34, authorNames=JIN H F, SONG Q Q, HU X, journalName=null, refType=null, unstructuredReference=JIN H F, SONG Q Q, HU X. Auto-keras: An Efficient Neural Architecture Search System[C]//Proceedings of the 25th ACM SIGKDD International Conference on Knowledge Discovery & Data Mining. Anchorage:ACM, 2018:1946-1956., articleTitle=Auto-keras: An Efficient Neural Architecture Search System, refAbstract=null), Reference(id=1251895534752056097, tenantId=1146029695717560320, journalId=1251234473337991274, articleId=1251893510224425537, doi=null, pmid=null, pmcid=null, year=2025-04-10, volume=null, issue=null, pageStart=null, pageEnd=null, url=null, language=null, rfNumber=[36], rfOrder=35, authorNames=ARAVIND S, TSUNG-YI L, NIKI P, journalName=null, refType=null, unstructuredReference=ARAVIND S, TSUNG-YI L, NIKI P, et al. Bottleneck Transformers for Visual Recognition[EB/OL].[2025-04-10].https://arxiv.org/pdf/2101.11605., articleTitle=Bottleneck Transformers for Visual Recognition, refAbstract=null), Reference(id=1251895534831747877, tenantId=1146029695717560320, journalId=1251234473337991274, articleId=1251893510224425537, doi=null, pmid=null, pmcid=null, year=2020, volume=null, issue=null, pageStart=1580, pageEnd=1589, url=null, language=null, rfNumber=[37], rfOrder=36, authorNames=HAN K, WANG Y H, TIAN Q, journalName=null, refType=null, unstructuredReference=HAN K,WANG Y H,TIAN Q,et al. GhostNet:More Features from Cheap Operations[C]//2020 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR). Piscataway:IEEE,2020:1580-1589., articleTitle=GhostNet:More Features from Cheap Operations, refAbstract=null), Reference(id=1251895534928216875, tenantId=1146029695717560320, journalId=1251234473337991274, articleId=1251893510224425537, doi=null, pmid=null, pmcid=null, year=2025-04-18, volume=null, issue=null, pageStart=null, pageEnd=null, url=null, language=null, rfNumber=[38], rfOrder=37, authorNames=HOWARD A G, ZHU M L, CHEN B, journalName=null, refType=null, unstructuredReference=HOWARD A G,ZHU M L,CHEN B,et al. MobileNets:Efficient Convolutional Neural Networks for Mobile Vision Applications[EB/OL]. (2017-04-17)[2025-04-18]. https://arxiv.org/pdf/1704.04861., articleTitle=MobileNets:Efficient Convolutional Neural Networks for Mobile Vision Applications, refAbstract=null), Reference(id=1251895535012102959, tenantId=1146029695717560320, journalId=1251234473337991274, articleId=1251893510224425537, doi=null, pmid=null, pmcid=null, year=2018, volume=null, issue=null, pageStart=4510, pageEnd=4520, url=null, language=null, rfNumber=[39], rfOrder=38, authorNames=SANDLER M, HOWARD A, ZHU M, journalName=null, refType=null, unstructuredReference=SANDLER M, HOWARD A, ZHU M, et al. MobileNetV2:Inverted Residuals and Linear Bottlenecks[C]//2018 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR). Piscataway:IEEE,2018:4510-4520., articleTitle=MobileNetV2:Inverted Residuals and Linear Bottlenecks, refAbstract=null), Reference(id=1251895535112766261, tenantId=1146029695717560320, journalId=1251234473337991274, articleId=1251893510224425537, doi=null, pmid=null, pmcid=null, year=2019, volume=null, issue=null, pageStart=1314, pageEnd=1324, url=null, language=null, rfNumber=[40], rfOrder=39, authorNames=HOWARD A, SANDLER M, CHU G, journalName=null, refType=null, unstructuredReference=HOWARD A,SANDLER M,CHU G,et al. Searching for MobileNetV3[C]//2019 IEEE/CVF International Conference on Computer Vision (ICCV). New York: IEEE, 2019:1314-1324., articleTitle=Searching for MobileNetV3, refAbstract=null), Reference(id=1251895535196652347, tenantId=1146029695717560320, journalId=1251234473337991274, articleId=1251893510224425537, doi=null, pmid=null, pmcid=null, year=2025-04-18, volume=null, issue=null, pageStart=null, pageEnd=null, url=null, language=null, rfNumber=[41], rfOrder=40, authorNames=WADEKAR S N, CHAURASIA A, journalName=null, refType=null, unstructuredReference=WADEKAR S N, CHAURASIA A. MobileViTv3: MobileFriendly Vision Transformer with Simple and Effective Fusion of Local, Global and Input Features[EB/OL]. (2022-09-01)[2025-04-18]. https://arxiv.org/pdf/2209.15159v2., articleTitle=MobileViTv3: MobileFriendly Vision Transformer with Simple and Effective Fusion of Local, Global and Input Features, refAbstract=null), Reference(id=1251895535276344127, tenantId=1146029695717560320, journalId=1251234473337991274, articleId=1251893510224425537, doi=null, pmid=null, pmcid=null, year=2025-04-18, volume=null, issue=null, pageStart=null, pageEnd=null, url=null, language=null, rfNumber=[42], rfOrder=41, authorNames=NUGROHO A A, SLAMET I, SUGIYANTO, journalName=null, refType=null, unstructuredReference=NUGROHO A A,SLAMET I,SUGIYANTO. Skins Cancer Identification System of HAM10000 Skin Cancer Dataset Using Convolutional Neural Network[EB/OL]. (2019-12-27)[2025-04-18]. https://doi.org/10.1063/1.5141652., articleTitle=Skins Cancer Identification System of HAM10000 Skin Cancer Dataset Using Convolutional Neural Network, refAbstract=null), Reference(id=1251895535372813124, tenantId=1146029695717560320, journalId=1251234473337991274, articleId=1251893510224425537, doi=null, pmid=null, pmcid=null, year=2025-04-18, volume=null, issue=null, pageStart=null, pageEnd=null, url=null, language=null, rfNumber=[43], rfOrder=42, authorNames=RATUL A R, MOZAFFARI M H, LEE W S, journalName=null, refType=null, unstructuredReference=RATUL A R,MOZAFFARI M H, LEE W S,et al. Skin Lesions Classification Using Deep Learning Based on Dilated Convolution[EB/OL]. (2020-06-05)[2025-04-18]. https://doi.org/10.1101/860700., articleTitle=Skin Lesions Classification Using Deep Learning Based on Dilated Convolution, refAbstract=null), Reference(id=1251895535469282122, tenantId=1146029695717560320, journalId=1251234473337991274, articleId=1251893510224425537, doi=null, pmid=null, pmcid=null, year=2021, volume=9, issue=null, pageStart=112193, pageEnd=112205, url=null, language=null, rfNumber=[44], rfOrder=43, authorNames=THURNHOFER H K, LÓPEZ R E, DOMÍNGUEZ E, journalName=IEEE Access, refType=null, unstructuredReference=THURNHOFER H K, LÓPEZ R E, DOMÍNGUEZ E, et al. Skin Lesion Classification by Ensembles of Deep Convolutional Networks and Regularly Spaced Shifting[J]. IEEE Access,2021,9:112193-112205., articleTitle=Skin Lesion Classification by Ensembles of Deep Convolutional Networks and Regularly Spaced Shifting, refAbstract=null), Reference(id=1251895535578334033, tenantId=1146029695717560320, journalId=1251234473337991274, articleId=1251893510224425537, doi=null, pmid=null, pmcid=null, year=2024, volume=5, issue=4, pageStart=711, pageEnd=721, url=null, language=null, rfNumber=[45], rfOrder=44, authorNames=PATIL S N, PATIL H D, ADHIYA K P, journalName=International Research Journal of Multidisciplinary Scope, refType=null, unstructuredReference=PATIL S N,PATIL H D,ADHIYA K P,et al. Densenet201 for Skin Melanoma Classification: A Comprehensive Performance Evaluation and Analysis[J]. International Research Journal of Multidisciplinary Scope,2024,5(4):711-721., articleTitle=Densenet201 for Skin Melanoma Classification: A Comprehensive Performance Evaluation and Analysis, refAbstract=null), Reference(id=1251895535704163160, tenantId=1146029695717560320, journalId=1251234473337991274, articleId=1251893510224425537, doi=null, pmid=null, pmcid=null, year=2025, volume=15, issue=null, pageStart=1, pageEnd=12, url=null, language=null, rfNumber=[46], rfOrder=45, authorNames=LIU H, DOU Y, WANG K, journalName=Scientific Reports, refType=null, unstructuredReference=LIU H,DOU Y,WANG K,et al. A Skin Disease Classification Model Based on Multi Scale Combined Efficient Channel Attention Module[J]. Scientific Reports,2025, 15:1-12., articleTitle=A Skin Disease Classification Model Based on Multi Scale Combined Efficient Channel Attention Module, refAbstract=null), Reference(id=1251895535796437852, tenantId=1146029695717560320, journalId=1251234473337991274, articleId=1251893510224425537, doi=null, pmid=null, pmcid=null, year=2023, volume=null, issue=null, pageStart=null, pageEnd=null, url=null, language=null, rfNumber=[47], rfOrder=46, authorNames=胡吉红, journalName=基于轻量级网络的皮肤病检测算法研究, refType=null, unstructuredReference=胡吉红.基于轻量级网络的皮肤病检测算法研究[D].兰州:西北师范大学,2023., articleTitle=null, refAbstract=null), Reference(id=1251895535897101154, tenantId=1146029695717560320, journalId=1251234473337991274, articleId=1251893510224425537, doi=null, pmid=null, pmcid=null, year=2019, volume=null, issue=null, pageStart=242, pageEnd=247, url=null, language=null, rfNumber=[48], rfOrder=47, authorNames=SAE L W, WETTAYAPRASIT W, AIYARAK P, journalName=null, refType=null, unstructuredReference=SAE L W,WETTAYAPRASIT W,AIYARAK P. Convolutional Neural Networks Using MobileNet for Skin Lesion Classification[C]//2019 IEEE Conference on Computer Vision and Pattern Recognition Workshops (CVPRW). Long Beach:IEEE,2019:242-247., articleTitle=Convolutional Neural Networks Using MobileNet for Skin Lesion Classification, refAbstract=null), Reference(id=1251895536010347367, tenantId=1146029695717560320, journalId=1251234473337991274, articleId=1251893510224425537, doi=null, pmid=null, pmcid=null, year=2021, volume=null, issue=null, pageStart=165, pageEnd=176, url=null, language=null, rfNumber=[49], rfOrder=48, authorNames=CHATURVEDI S S, GUPTA K, PRASAD P S, journalName=null, refType=null, unstructuredReference=CHATURVEDI S S,GUPTA K,PRASAD P S. Skin Lesion Analyser:An Efficient Seven-way Multi-class Skin Cancer Classification Using MobileNet[C]//Advanced Machine Learning Technologies and Applications: Proceedings of AMLTA 2020. Singapore:Springer,2021:165-176., articleTitle=Skin Lesion Analyser:An Efficient Seven-way Multi-class Skin Cancer Classification Using MobileNet, refAbstract=null), Reference(id=1251895536085844844, tenantId=1146029695717560320, journalId=1251234473337991274, articleId=1251893510224425537, doi=null, pmid=null, pmcid=null, year=2025-04-18, volume=null, issue=null, pageStart=null, pageEnd=null, url=null, language=null, rfNumber=[50], rfOrder=49, authorNames=朱秋慧, 杨靖, 黄若愚, journalName=null, refType=null, unstructuredReference=朱秋慧,杨靖,黄若愚,.基于部分卷积的多尺度特征卷积神经网络模型[J/OL]. (2025-03-11)[2025-04-18].http://kns.cnki.net/kcms/detail/13.1099.TN.20250310.1707.012.html., articleTitle=基于部分卷积的多尺度特征卷积神经网络模型, refAbstract=null)], funds=null, companyList=[AuthorCompany(id=1251895520554336469, tenantId=1146029695717560320, journalId=1251234473337991274, articleId=1251893510224425537, xref=null, ext=[AuthorCompanyExt(id=1251895520562725078, tenantId=1146029695717560320, journalId=1251234473337991274, articleId=1251893510224425537, companyId=1251895520554336469, language=EN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=College of Information and Electrical Engineering, Hebei University of Engineering, Handan 056038, China), AuthorCompanyExt(id=1251895520571113688, tenantId=1146029695717560320, journalId=1251234473337991274, articleId=1251893510224425537, companyId=1251895520554336469, language=CN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=河北工程大学 信息与电气工程学院,河北 邯郸 056038)])], figs=[ArticleFig(id=1251895523045753189, tenantId=1146029695717560320, journalId=1251234473337991274, articleId=1251893510224425537, language=EN, label=Fig. 1, caption=Principle of DETR, figureFileSmall=6Ib8uwVa8Of7eoYpfibc1Q==, figureFileBig=J+DCkp1knKwkBfSonfJHVA==, tableContent=null), ArticleFig(id=1251895523129639274, tenantId=1146029695717560320, journalId=1251234473337991274, articleId=1251893510224425537, language=CN, label=图1, caption=DETR原理, figureFileSmall=6Ib8uwVa8Of7eoYpfibc1Q==, figureFileBig=J+DCkp1knKwkBfSonfJHVA==, tableContent=null), ArticleFig(id=1251895523284828533, tenantId=1146029695717560320, journalId=1251234473337991274, articleId=1251893510224425537, language=EN, label=Fig. 2, caption=Relationship between frequency domain,spatial domain and complex frequency domain of an image, figureFileSmall=WEWFkRFtZZ+I7nK7WemuHg==, figureFileBig=u1gpmo2PsqyQx4MARPkYbw==, tableContent=null), ArticleFig(id=1251895524887052666, tenantId=1146029695717560320, journalId=1251234473337991274, articleId=1251893510224425537, language=CN, label=图2, caption=图像的频率域、空间域和复频域之间的关系, figureFileSmall=WEWFkRFtZZ+I7nK7WemuHg==, figureFileBig=u1gpmo2PsqyQx4MARPkYbw==, tableContent=null), ArticleFig(id=1251895525004493185, tenantId=1146029695717560320, journalId=1251234473337991274, articleId=1251893510224425537, language=EN, label=Fig. 3, caption=Structure of GMTB, figureFileSmall=kfecoVXBkMOGDu+IHW99sg==, figureFileBig=EkdNVAKWP77jwKo0NUJzaQ==, tableContent=null), ArticleFig(id=1251895525084184965, tenantId=1146029695717560320, journalId=1251234473337991274, articleId=1251893510224425537, language=CN, label=图3, caption=GMTB结构示意, figureFileSmall=kfecoVXBkMOGDu+IHW99sg==, figureFileBig=EkdNVAKWP77jwKo0NUJzaQ==, tableContent=null), ArticleFig(id=1251895525168071052, tenantId=1146029695717560320, journalId=1251234473337991274, articleId=1251893510224425537, language=EN, label=Fig. 4, caption=Calculation method of two-dimensional FDWT for image, figureFileSmall=a9rObqWxl6VkF+l8fw8QgQ==, figureFileBig=/xu7Dk4xCK7memiDenoAzw==, tableContent=null), ArticleFig(id=1251895525281317268, tenantId=1146029695717560320, journalId=1251234473337991274, articleId=1251893510224425537, language=CN, label=图4, caption=对图像进行二维FDWT的计算方法, figureFileSmall=a9rObqWxl6VkF+l8fw8QgQ==, figureFileBig=/xu7Dk4xCK7memiDenoAzw==, tableContent=null), ArticleFig(id=1251895525373591959, tenantId=1146029695717560320, journalId=1251234473337991274, articleId=1251893510224425537, language=EN, label=Fig. 5, caption=Calculation method of two-dimensional FIDWT for image, figureFileSmall=+A9nh7Z63pnr+ZXujG/rJA==, figureFileBig=Q/mO9LXcKXHTPA4LZGVsCA==, tableContent=null), ArticleFig(id=1251895525507809694, tenantId=1146029695717560320, journalId=1251234473337991274, articleId=1251893510224425537, language=CN, label=图5, caption=对图像进行二维FIDWT的计算方法, figureFileSmall=+A9nh7Z63pnr+ZXujG/rJA==, figureFileBig=Q/mO9LXcKXHTPA4LZGVsCA==, tableContent=null), ArticleFig(id=1251895525595890085, tenantId=1146029695717560320, journalId=1251234473337991274, articleId=1251893510224425537, language=EN, label=Fig. 6, caption=Structure of WTCB, figureFileSmall=7JcRCsaYDmR4tNneZWmRHg==, figureFileBig=crPt0QKZmGZ+NwmfrdBFFg==, tableContent=null), ArticleFig(id=1251895525784633769, tenantId=1146029695717560320, journalId=1251234473337991274, articleId=1251893510224425537, language=CN, label=图6, caption=WTCB的结构示意, figureFileSmall=7JcRCsaYDmR4tNneZWmRHg==, figureFileBig=crPt0QKZmGZ+NwmfrdBFFg==, tableContent=null), ArticleFig(id=1251895525885297072, tenantId=1146029695717560320, journalId=1251234473337991274, articleId=1251893510224425537, language=EN, label=Fig. 7, caption=Model structure of dermoscopy image processing system, figureFileSmall=zpbEA7T1BPJZCd6TtcaHzg==, figureFileBig=tjWHwExnv8SJTg0kQCsx7g==, tableContent=null), ArticleFig(id=1251895525969183156, tenantId=1146029695717560320, journalId=1251234473337991274, articleId=1251893510224425537, language=CN, label=图7, caption=皮肤镜图像处理系统模型结构, figureFileSmall=zpbEA7T1BPJZCd6TtcaHzg==, figureFileBig=tjWHwExnv8SJTg0kQCsx7g==, tableContent=null), ArticleFig(id=1251895526074040761, tenantId=1146029695717560320, journalId=1251234473337991274, articleId=1251893510224425537, language=EN, label=Fig. 8, caption=HAM10000 dermoscopy dataset division, figureFileSmall=fCsPkbE4UjUvxOceenSpcA==, figureFileBig=h7XnP3BMXzvRHOGDx/ElZA==, tableContent=null), ArticleFig(id=1251895526166315452, tenantId=1146029695717560320, journalId=1251234473337991274, articleId=1251893510224425537, language=CN, label=图8, caption=HAM10000皮肤镜数据集的划分方法, figureFileSmall=fCsPkbE4UjUvxOceenSpcA==, figureFileBig=h7XnP3BMXzvRHOGDx/ElZA==, tableContent=null), ArticleFig(id=1251895526258590145, tenantId=1146029695717560320, journalId=1251234473337991274, articleId=1251893510224425537, language=EN, label=Fig. 9, caption=Dimensionality reduction scatter plots of training dataset,validation dataset and test datase, figureFileSmall=YJGdEzi5qJa9smNebmLFCg==, figureFileBig=0sbHQBUSWSsBhUyCy56wzw==, tableContent=null), ArticleFig(id=1251895526338281924, tenantId=1146029695717560320, journalId=1251234473337991274, articleId=1251893510224425537, language=CN, label=图9, caption=训练集、验证集、测试集的降维散点, figureFileSmall=YJGdEzi5qJa9smNebmLFCg==, figureFileBig=0sbHQBUSWSsBhUyCy56wzw==, tableContent=null), ArticleFig(id=1251895526468305356, tenantId=1146029695717560320, journalId=1251234473337991274, articleId=1251893510224425537, language=EN, label=Fig. 10, caption=Change curve of evaluation index in training process and verification process, figureFileSmall=/q0eq+w3QR0VN1XQo3AV1Q==, figureFileBig=jtBcn7Ptib8RFMXvyfhEcg==, tableContent=null), ArticleFig(id=1251895526573162962, tenantId=1146029695717560320, journalId=1251234473337991274, articleId=1251893510224425537, language=CN, label=图10, caption=训练过程与验证过程中评价指标的变化曲线, figureFileSmall=/q0eq+w3QR0VN1XQo3AV1Q==, figureFileBig=jtBcn7Ptib8RFMXvyfhEcg==, tableContent=null), ArticleFig(id=1251895526665437655, tenantId=1146029695717560320, journalId=1251234473337991274, articleId=1251893510224425537, language=EN, label=Fig. 11, caption=Two-dimensional heat map drawn by confusion matrix of test dataset, figureFileSmall=WU6tyUjeFLK8phdDL+rlYw==, figureFileBig=6Aw8y5Xk+dJPqIzyI6Br5Q==, tableContent=null), ArticleFig(id=1251895526736740830, tenantId=1146029695717560320, journalId=1251234473337991274, articleId=1251893510224425537, language=CN, label=图11, caption=测试集混淆矩阵的二维热力图, figureFileSmall=WU6tyUjeFLK8phdDL+rlYw==, figureFileBig=6Aw8y5Xk+dJPqIzyI6Br5Q==, tableContent=null), ArticleFig(id=1251895526833209830, tenantId=1146029695717560320, journalId=1251234473337991274, articleId=1251893510224425537, language=EN, label=Fig. 12, caption=Data comparison of different models under the same index, figureFileSmall=eMfcr2GjSduefvDxxrK3EA==, figureFileBig=1mHVhCA91IhPrjdZbv4Pzg==, tableContent=null), ArticleFig(id=1251895526942261742, tenantId=1146029695717560320, journalId=1251234473337991274, articleId=1251893510224425537, language=CN, label=图12, caption=不同模型在同一指标下的数据对比, figureFileSmall=eMfcr2GjSduefvDxxrK3EA==, figureFileBig=1mHVhCA91IhPrjdZbv4Pzg==, tableContent=null), ArticleFig(id=1251895527059702262, tenantId=1146029695717560320, journalId=1251234473337991274, articleId=1251893510224425537, language=EN, label=Tab. 1, caption=

Comparison of the main advantages and disadvantages of existing target detection methods

, figureFileSmall=null, figureFileBig=null, tableContent=
技术方案主要优点主要缺点
CNN结构简单检测性能不佳
SSD速度与精度的均衡小目标检测不灵敏
R-CNN精度较高训练过程缓慢
YOLO快速、实时检测目标数量受限,精度较差
DETR端到端检测注意力机制消耗过量成本
), ArticleFig(id=1251895527143588347, tenantId=1146029695717560320, journalId=1251234473337991274, articleId=1251893510224425537, language=CN, label=表1, caption=

现有目标检测方法的主要优缺点对比

, figureFileSmall=null, figureFileBig=null, tableContent=
技术方案主要优点主要缺点
CNN结构简单检测性能不佳
SSD速度与精度的均衡小目标检测不灵敏
R-CNN精度较高训练过程缓慢
YOLO快速、实时检测目标数量受限,精度较差
DETR端到端检测注意力机制消耗过量成本
), ArticleFig(id=1251895527248445958, tenantId=1146029695717560320, journalId=1251234473337991274, articleId=1251893510224425537, language=EN, label=Tab. 2, caption=

Evaluation indices of three groups of ablation experiments

, figureFileSmall=null, figureFileBig=null, tableContent=
分组模型TPTNFPAccuracyPrecisionF1AP
1ResNet181 444285610.724 10.720 20.837 30.020 9
ResNet341 493285120.748 20.744 60.853 60.020 5
ResNet501 34106640.668 80.668 80.801 60.027 5
2WTCB-R181 495195100.748 00.745 60.854 00.019 0
WTCB-R341 506294990.754 70.751 10.858 00.020 0
WTCB-R501 34106640.668 80.668 80.802 00.021 0
3SFTNet-R181 651363540.826 60.823 40.903 20.051 7
SFTNet-R341 708442970.855 10.851 90.920 00.021 2
SFTNet-R501 683473220.843 10.839 40.912 70.019 9
SFTNet-R1011 617423880.810 50.806 50.892 90.021 1
), ArticleFig(id=1251895527349109266, tenantId=1146029695717560320, journalId=1251234473337991274, articleId=1251893510224425537, language=CN, label=表2, caption=

3组消融实验中的各项评价指标

, figureFileSmall=null, figureFileBig=null, tableContent=
分组模型TPTNFPAccuracyPrecisionF1AP
1ResNet181 444285610.724 10.720 20.837 30.020 9
ResNet341 493285120.748 20.744 60.853 60.020 5
ResNet501 34106640.668 80.668 80.801 60.027 5
2WTCB-R181 495195100.748 00.745 60.854 00.019 0
WTCB-R341 506294990.754 70.751 10.858 00.020 0
WTCB-R501 34106640.668 80.668 80.802 00.021 0
3SFTNet-R181 651363540.826 60.823 40.903 20.051 7
SFTNet-R341 708442970.855 10.851 90.920 00.021 2
SFTNet-R501 683473220.843 10.839 40.912 70.019 9
SFTNet-R1011 617423880.810 50.806 50.892 90.021 1
), ArticleFig(id=1251895527474938393, tenantId=1146029695717560320, journalId=1251234473337991274, articleId=1251893510224425537, language=EN, label=Tab. 3, caption=

Comparison table of test dataset accuracy between SFTNet and the dataset benchmark model

, figureFileSmall=null, figureFileBig=null, tableContent=
来源ResNet18ResNet50Auto-sklearnAuto-kerasGoogle AutoML VisionSFTNet(ours)
文献[32]0.7270.7190.7340.7560.766
文献[33]0.7540.7310.7190.7490.768
本文0.7240.6690.855
), ArticleFig(id=1251895527583990303, tenantId=1146029695717560320, journalId=1251234473337991274, articleId=1251893510224425537, language=CN, label=表3, caption=

本文模型与数据集基准模型的测试集正确率对比表

, figureFileSmall=null, figureFileBig=null, tableContent=
来源ResNet18ResNet50Auto-sklearnAuto-kerasGoogle AutoML VisionSFTNet(ours)
文献[32]0.7270.7190.7340.7560.766
文献[33]0.7540.7310.7190.7490.768
本文0.7240.6690.855
), ArticleFig(id=1251895527672070692, tenantId=1146029695717560320, journalId=1251234473337991274, articleId=1251893510224425537, language=EN, label=Tab. 4, caption=

Test evaluation indexes of different models on HAM10000 skin disease test dataset

, figureFileSmall=null, figureFileBig=null, tableContent=
对比来源核心模型或主要方法AccuracyPrecisionRecallF1
文献[42]LeNet0.7800.4140.4140.414
文献[43]VGG0.8500.8500.8500.850
文献[44]GoogLeNet0.8050.6850.5810.594
文献[45]ResNet0.8520.8490.9750.908
文献[46]BoTNet0.8150.7260.6860.692
文献[47]GhostNet0.8370.8370.8320.833
文献[48]MobileNet0.7930.6800.8000.763
文献[49]MobileNet0.8320.8300.8300.754
文献[44]MobileNet-v20.8190.7140.6590.664
文献[44]MobileNet-v2 & GoogLeNet-v20.8350.7660.6560.672
文献[46]MobileNet-v30.8430.7980.6710.693
文献[47]MobileNet-v30.8380.8390.8340.833
文献[47]MobileNet-v3 & CA0.8470.8480.8450.846
文献[47]MobileNet-v3 & SE0.8480.8460.8490.846
文献[47]Attention MobileNet-v30.8500.8490.8420.841
文献[46]MobileViT-v30.8510.8020.7220.738
本文SFTNet0.8550.8520.8510.851
), ArticleFig(id=1251895527764345385, tenantId=1146029695717560320, journalId=1251234473337991274, articleId=1251893510224425537, language=CN, label=表4, caption=

不同模型在HAM10000皮肤病测试数据集上的评价指标

, figureFileSmall=null, figureFileBig=null, tableContent=
对比来源核心模型或主要方法AccuracyPrecisionRecallF1
文献[42]LeNet0.7800.4140.4140.414
文献[43]VGG0.8500.8500.8500.850
文献[44]GoogLeNet0.8050.6850.5810.594
文献[45]ResNet0.8520.8490.9750.908
文献[46]BoTNet0.8150.7260.6860.692
文献[47]GhostNet0.8370.8370.8320.833
文献[48]MobileNet0.7930.6800.8000.763
文献[49]MobileNet0.8320.8300.8300.754
文献[44]MobileNet-v20.8190.7140.6590.664
文献[44]MobileNet-v2 & GoogLeNet-v20.8350.7660.6560.672
文献[46]MobileNet-v30.8430.7980.6710.693
文献[47]MobileNet-v30.8380.8390.8340.833
文献[47]MobileNet-v3 & CA0.8470.8480.8450.846
文献[47]MobileNet-v3 & SE0.8480.8460.8490.846
文献[47]Attention MobileNet-v30.8500.8490.8420.841
文献[46]MobileViT-v30.8510.8020.7220.738
本文SFTNet0.8550.8520.8510.851
)], attaches=null, journal=Journal(id=1251231494887223395, delFlag=0, nameCn=无线电通信技术, nameEn=Radio Communications Technology, nameHistory1=null, nameHistory2=null, issn=1003-3114, eissn=, cn=13-1099/TN, coden=null, periodic=1, language=CN, oaType=1, 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=, officePhone=, editDirector=null, officeDirector=null, officeDirectorPhone=null, officeStaffNum=null, officeEmpNum=null, coverPicUrl=veWCdfK9mJVXm/uFgI4wQA==, journalPrice=null, startedYear=null, abbrevIsoEn=Radio Communications Technology, journalRemark=null, publicationField=null, createdTime=1776246435141, updatedTime=1776397604574, createdBy=18614031015, updatedBy=13701087609, firstLetterCn=R, firstLetterEn=R, subjectCode=Engineering, subjectName=工程, subjectCodeEn=Engineering, subjectNameEn=null, picCn=veWCdfK9mJVXm/uFgI4wQA==, picEn=OSQVHuARoHUd1TQ4ONLQrQ==, jcr=null, cjcr=null, exts=[JournalExt(id=1251865545604285354, language=CN, name=无线电通信技术, 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=1776397604609, updatedTime=1776397604609, createdBy=13701087609, updatedBy=13701087609, submissionGuidelinesUrl=, submissionAuthorUrl=https://wxdt.cbpt.cnki.net/index.aspx?t=1, submissionEditorUrl=https://wxdt.cbpt.cnki.net/index.aspx?t=3, submissionReviewUrl=https://wxdt.cbpt.cnki.net/index.aspx?t=2, submissionCeEditorUrl=, submissionAeEditorUrl=, option={"copyright":""}), JournalExt(id=1251865545646228395, language=EN, name=Radio Communications Technology, 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=1776397604619, updatedTime=1776397604619, createdBy=13701087609, updatedBy=13701087609, submissionGuidelinesUrl=, submissionAuthorUrl=https://wxdt.cbpt.cnki.net/index.aspx?t=1, submissionEditorUrl=https://wxdt.cbpt.cnki.net/index.aspx?t=3, submissionReviewUrl=https://wxdt.cbpt.cnki.net/index.aspx?t=2, submissionCeEditorUrl=, submissionAeEditorUrl=, option={"copyright":""})], databaseList=null, tenantJournalId=1251234473337991274, websiteList=[Website(id=1251257283515203650, webName=null, webTitle=null, webDomain=null, webCopyrigh=null, webIpcNo=null, seoTitle=null, seoKeywords=null, seoDescription=null, tenantJournalId=null, journalId=1251234473337991274, 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/wxdtxjs/CN, language=CN, createTime=1776252583627, createBy=18614031015, updateTime=1776253691546, updateBy=18614031015, name=无线电通信技术-中文, tplId=1146099689490845704, title=无线电通信技术, delFlag=0, indexPage=/home, props=[WebsiteProps(id=1251262047678313076, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1251257283515203650, code=articleTextType, value=kx, createTime=1776253719491, updateTime=1776253719491, creator=18614031015, updator=18614031015), WebsiteProps(id=1251262047653147249, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1251257283515203650, code=banner, value=null, createTime=1776253719485, updateTime=1776253719485, creator=18614031015, updator=18614031015), WebsiteProps(id=1251262047707673207, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1251257283515203650, code=grayFlag, value=0, createTime=1776253719498, updateTime=1776253719498, creator=18614031015, updator=18614031015), WebsiteProps(id=1251262047644758640, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1251257283515203650, code=logo, value=https://castjournals.cast.org.cn/joweb/wxdtxjs/CN/file/pic?fileId=sk5LMh+QbAm+98l18HjovQ==, createTime=1776253719483, updateTime=1776253719483, creator=18614031015, updator=18614031015), WebsiteProps(id=1251262047720256121, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1251257283515203650, code=minRunFlag, value=0, createTime=1776253719501, updateTime=1776253719501, creator=18614031015, updator=18614031015), WebsiteProps(id=1251262047669924467, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1251257283515203650, code=picServerUrl, value=https://castjournals.cast.org.cn/joweb/wxdtxjs/CN/file/pic, createTime=1776253719489, updateTime=1776253719489, creator=18614031015, updator=18614031015), WebsiteProps(id=1251262047716061816, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1251257283515203650, code=silenceFlag, value=0, createTime=1776253719500, updateTime=1776253719500, creator=18614031015, updator=18614031015), WebsiteProps(id=1251262047661535858, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1251257283515203650, code=staticResourcePath, value=https://castjournals.cast.org.cn/joweb/cast_kjdb_cn_619/, createTime=1776253719487, updateTime=1776253719487, creator=18614031015, updator=18614031015), WebsiteProps(id=1251262047682507381, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1251257283515203650, code=themeColor, value=null, createTime=1776253719492, updateTime=1776253719492, creator=18614031015, updator=18614031015), WebsiteProps(id=1251262047690895990, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1251257283515203650, code=themeStyle, value=null, createTime=1776253719494, updateTime=1776253719494, creator=18614031015, updator=18614031015)]), Website(id=1251257283607478339, webName=null, webTitle=null, webDomain=null, webCopyrigh=null, webIpcNo=null, seoTitle=null, seoKeywords=null, seoDescription=null, tenantJournalId=null, journalId=1251234473337991274, 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/wxdtxjs/EN, language=EN, createTime=1776252583648, createBy=18614031015, updateTime=1776253687916, updateBy=18614031015, name=无线电通信技术-英文, tplId=1146101810881728533, title=Radio Communications Technology, delFlag=0, indexPage=/home, props=[WebsiteProps(id=1251262071707484468, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1251257283607478339, code=articleTextType, value=kx, createTime=1776253725220, updateTime=1776253725220, creator=18614031015, updator=18614031015), WebsiteProps(id=1251262071690707249, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1251257283607478339, code=banner, value=null, createTime=1776253725216, updateTime=1776253725216, creator=18614031015, updator=18614031015), WebsiteProps(id=1251262071724261687, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1251257283607478339, code=grayFlag, value=0, createTime=1776253725224, updateTime=1776253725224, creator=18614031015, updator=18614031015), WebsiteProps(id=1251262071682318640, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1251257283607478339, code=logo, value=https://castjournals.cast.org.cn/joweb/wxdtxjs/EN/file/pic?fileId=sk5LMh+QbAm+98l18HjovQ==, createTime=1776253725214, updateTime=1776253725214, creator=18614031015, updator=18614031015), WebsiteProps(id=1251262071732650297, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1251257283607478339, code=minRunFlag, value=0, createTime=1776253725226, updateTime=1776253725226, creator=18614031015, updator=18614031015), WebsiteProps(id=1251262071703290163, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1251257283607478339, code=picServerUrl, value=https://castjournals.cast.org.cn/joweb/wxdtxjs/EN/file/pic, createTime=1776253725219, updateTime=1776253725219, creator=18614031015, updator=18614031015), WebsiteProps(id=1251262071728455992, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1251257283607478339, code=silenceFlag, value=0, createTime=1776253725225, updateTime=1776253725225, creator=18614031015, updator=18614031015), WebsiteProps(id=1251262071694901554, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1251257283607478339, code=staticResourcePath, value=https://castjournals.cast.org.cn/joweb/cast_kjdb_en_623/, createTime=1776253725217, updateTime=1776253725217, creator=18614031015, updator=18614031015), WebsiteProps(id=1251262071711678773, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1251257283607478339, code=themeColor, value=null, createTime=1776253725221, updateTime=1776253725221, creator=18614031015, updator=18614031015), WebsiteProps(id=1251262071720067382, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1251257283607478339, code=themeStyle, value=null, createTime=1776253725223, updateTime=1776253725223, creator=18614031015, updator=18614031015)])], journalTitle=无线电通信技术, weixinUrl=null, journalUrl=https://wxdt.cbpt.cnki.net/, iacademicId=null, status=1, seqNo=null, journalTitleEn=Radio Communications Technology, journalPhotoCn=veWCdfK9mJVXm/uFgI4wQA==, journalPhotoEn=OSQVHuARoHUd1TQ4ONLQrQ==, journalFirstLetter=R, 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), detailUrlCn=https://castjournals.cast.org.cn/joweb/wxdtxjs/CN/10.3969/j.issn.1003-3114.2025.05.012, detailUrlEn=https://castjournals.cast.org.cn/joweb/wxdtxjs/EN/10.3969/j.issn.1003-3114.2025.05.012, pdfUrlCn=https://castjournals.cast.org.cn/joweb/wxdtxjs/CN/PDF/10.3969/j.issn.1003-3114.2025.05.012, pdfUrlEn=https://castjournals.cast.org.cn/joweb/wxdtxjs/EN/PDF/10.3969/j.issn.1003-3114.2025.05.012, aliStartDate=null, aliEndDate=null, collectionFlag=false, citedCount=null, citedUrl=null, reference=null)
收藏切换
基于伽马变换与小波卷积的皮肤癌精确检测
收藏切换
PDF下载
白世强 , 蔡益人 , 李丽宏 *
无线电通信技术 | 专题:智能通信、存储与信息处理技术前沿 2025,51(5): 993-1007
收起
收藏切换
无线电通信技术 | 专题:智能通信、存储与信息处理技术前沿 2025, 51(5): 993-1007
基于伽马变换与小波卷积的皮肤癌精确检测
全屏
白世强, 蔡益人, 李丽宏*
作者信息
  • 河北工程大学 信息与电气工程学院,河北 邯郸 056038
  • 白世强 男,(1999—),硕士研究生。主要研究方向:智能医学工程、医学影像感知。

    蔡益人 男,(2000—),硕士研究生。主要研究方向:小样本分类。

通讯作者:

李丽宏 女,(1973—),博士,教授。主要研究方向:图像处理、计算机视觉。
Precise Skin Cancer Detection Based on Gamma Transform and Wavelet Convolution
Shiqiang BAI, Yiren CAI, Lihong LI*
Affiliations
  • College of Information and Electrical Engineering, Hebei University of Engineering, Handan 056038, China
出版时间: 2025-09-18 doi: 10.3969/j.issn.1003-3114.2025.05.012
文章导航
收藏切换

皮肤癌和黑素细胞痣的相似性较高,极易被皮肤科医师误诊。为提高早期皮肤癌患者的筛查准确度,基于伽马变换(Gamma Transform,GT)算法和小波变换(Wavelet Transform,WT)算法,分别构建伽马变换模块(Gamma Transform Block,GMTB)与小波卷积模块(Wavelet Convolution Block,WTCB),并在Detection Transformer(DETR)架构基础上创新性地提出用于捕捉皮肤癌细粒度特征的空频变换网络(Space-Frequency Transform Network,SFTNet)。包含SFT-Net的皮肤癌筛选系统能够增强样本图像的不同通道,减少模型训练过程中的过拟合效应,进而有效提升疾病检测精度。基于HAM10000数据集的仿真实验结果表明,系统正确率达85.5%,在皮肤癌辅助诊断方面具有重要的临床应用价值。

深度学习  /  医学影像处理  /  皮肤癌检测  /  伽马变换  /  小波卷积  /  空频变换网络

Skin cancer and melanocytic nevus share numerous similarities, which can result in a misdiagnosis by dermatologists. To improve the screening accuracy of early skin cancer patients, the Gamma Transform Block (GMTB) based on Gamma Transform (GT) and Wavelet Convolution Block (WTCB) based on Wavelet Transform (WT) are proposed. Furthermore, the Space-Frequency Transform Network (SFTNet) for capturing fine-grained features of skin cancer is innovatively proposed based on the Detection Transformer(DETR) architecture. SFTNet-based skin cancer screening system can effectively improve disease detection accuracy because it enhances the sample image at different channels and reduces over-fitting effect during the model training process. Simulation results on HAM10000 dataset show that the accuracy of this system can reach 85.5%, which underscores the significant clinical value of our approach in skin cancer assisted diagnosis.

deep learning  /  medical image processing  /  skin cancer detection  /  GT  /  wavelet convolution  /  SFTNet
白世强, 蔡益人, 李丽宏. 基于伽马变换与小波卷积的皮肤癌精确检测. 无线电通信技术, 2025 , 51 (5) : 993 -1007 . DOI: 10.3969/j.issn.1003-3114.2025.05.012
Shiqiang BAI, Yiren CAI, Lihong LI. Precise Skin Cancer Detection Based on Gamma Transform and Wavelet Convolution[J]. Radio Communications Technology, 2025 , 51 (5) : 993 -1007 . DOI: 10.3969/j.issn.1003-3114.2025.05.012
近年来,因紫外线灼伤或化妆品使用不当导致的皮肤癌患者数量不断增加。相较于常见的肺癌和肝癌,皮肤癌更容易被群众忽视,最终因患者发现不及时而错过最佳治疗期[1]。相较于普通的体格检查,皮肤镜具有高精度的优点,正在逐步成为皮肤癌筛查的重要方式。由于我国皮肤性病科医师数量不足,与日俱增的皮肤镜检查和海量的结果报告往往让医师捉襟见肘。本文以深度学习领域中的目标检测技术为核心,提出基于GT与小波卷积的皮肤癌检测系统,为皮肤镜检查结果报告不及时的问题提供新的解决方案。
皮肤性病科的检查项目通常包含体格检查、组织病理学检查、皮肤影像学检查、实验室诊断、变应原检测等[2]。皮肤镜又称皮表透光显微镜,是一种利用光学技术深入观测皮肤角质层以下结构的仪器。同其他影像学检查一样,皮肤镜检查不会对患者机体产生侵入性损伤,因而在众多方案中被优先考虑。文献[3-6]的研究结果表明,皮肤镜在非黑色素瘤皮肤癌、基底细胞癌、黑素细胞痣、脂溢性角化、日光性角化等常见皮肤疾病的临床诊断中具有重要的应用价值。
皮肤镜检查的结果通常以图像形式呈现。因此,对患者的皮肤镜检查结果进行预筛分析的过程可以建模为一个图像检测系统。在深度学习领域中,现有的目标检测技术可大致分为2类。其中一类是基于卷积神经网络(Convolutional Neural Network,CNN)的目标检测技术,该类的典型代表有单发多框检测(Single Shot Multibox Detector,SSD)模型[7]、区域卷积神经网络(Region-based Convolutional Neural Network,R-CNN)系列模型[8-10]以及You Only Look Once(YOLO)系列模型[11]。另一类则是基于注意力机制和Transformer[12],典型代表为Facebook AI Research团队[13]所提出的DETR架构(如图1所示)及其衍生模型。上述各方法的主要优缺点如表1所示。
当需要检测的每张图像样本中仅考虑一个类别时,可以将其简化为图像分类问题。虽然皮肤镜检查的结果可能显示某患者患有多种皮肤疾病,若仅设定其中一种关键疾病作为检测对象,也能够通过图像分类算法完成既定目标。
现有的智能医学影像检测技术中,多以CNN为结构核心。赵嘉晖等[14]提出基于CCNet注意力机制和高维多目标优化算法的小样本皮肤癌检测模型,并在ISIC2018、Derm7pt数据集上取得较好效果。王煜坤[15]在博士学位论文中同样利用CNN进行皮肤疾病的智能诊断。与前二者不同,Tschandl等[16]采用CNN与皮肤科医师协作的方法进行皮肤癌检测,进一步确保诊断结果的可靠性。
虽然上述研究在智能皮肤病识别领域都有明显突破,但均未能有效地解决模型过拟合问题,也未弥补降采样操作所带来的信息损失。因此,为克服现有技术的缺陷,提出GMTB和WTCB,并选取性能更优秀的DETR作为基础架构,以对皮肤癌患者进行更精确的筛选。
一幅图像可以被表示为空间域、频率域、复频域3种基本形式,其中的频率域和复频域均属于变换域。3种表示方法之间的关系如图2所示。
本文主要使用GT、空间滤波(Spatial Filtering,SF)、WT三种技术。
空间域是一幅图像本身所处的表达域,一般用于直接对像素进行处理。空间域图像处理技术主要包括直方图增强(Histogram Enhancement,HE)、灰度变换(Grey Level Transform,GLT)、SF三种[17]
HE主要指通过使用均衡化等方法改变灰度图像原有的灰度值分布规律,以增加或减少灰度图像的对比度。
GLT是一种通过对灰度图像中像素点的灰度值进行特定的函数运算,以调整并增强图像可视化效果的技术。设一幅灰度图像的空间域函数表达式是fxy),变换关系函数或变换算子是s=Tr),对其进行GLT后的表达式为[17]:
GLT常使用线性函数、对数函数、幂函数3种变换关系。基于幂函数的幂变换算法又称GT,其函数式为:
式中:r为输入灰度值,γ为变换参数,s为变换后的灰度值,c为常系数。当伽马参数小于1时,拉伸图像灰度直方图的左端并压缩其右端;当伽马参数大于1时,压缩图像灰度直方图的左端并拉伸其右端;当伽马参数等于1时,GT则退化为线性变换[18]。GLT的本质是对每个空间像素进行灰度调整。
SF是指在图像的空间域进行矩阵乘积运算,通过把每个像素点替换为该像素及其邻域像素的函数值来增强或抑制图像中的某个特定成分。设一幅数字图像在空间域的表达式为fxy),空间滤波器的核为kxy),则该图像的空间域卷积式为:
若对卷积核k进行反转,可得到如下SF表达式:
SF的本质是对图像像素的邻域进行空间卷积运算。
图像频率域处理技术的核心思想在于,使用傅里叶变换或拉普拉斯变换将原始图像由基本的空间域转变至对应的变换域,在频率域或复频域进行相应的处理后,再将处理结果反变换至空间域,以完成既定目标。傅里叶变换[19]以及后续的余弦变换[20]、斜变换[21]、沃尔什-哈达玛变换[22]、WT[23]等方法均属于线性变换。
当一维连续函数ft)的傅里叶变换Fw)满足条件时,可通过WT将已知信号分解为小基波和尺度函数的线性组合。连续WT及其逆变换为:
式中:a为伸缩因子,b为平移因子。在此基础上取:
可得到函数ft)的一维离散小波变换(Discrete Wavelet Transform,DWT)系数。其表达式为:
作为时频分析领域的基础理论之一,WT在现代信号处理中发挥着不可替代的作用。
已知患者的皮肤镜检查结果为RGB格式图像。设该彩色数字图像宽为W pixel,高为H pixel,像素点在图像中所对应的位置坐标为(xy)。
读取患者的皮肤镜检查结果,对彩色图像进行通道分解,得到R通道、G通道、B通道。设(xy)在3个通道中所对应的灰度值大小分别为rxy)、gxy)、bxy),且单个通道中的灰度最大值是M,灰度均值和灰度方差分别是ms,对3个通道分别进行灰度归一化和图像标准化,可得:
设定伽马参数,使用幂函数对标准化后的灰度图像进行逐像素GT,依次得到:
最终将变换后的3个通道合并,作为GMTB的输出。
GMTB的结构如图3所示。
GMTB主要由通道拆分、单通道GT、通道整合3个部分组成。设模块输入一幅大小是w×h的彩色皮肤镜图像,张量形状为3×w×h。在通道拆分过程中,仅使用赋值语句提取张量中的一个维度;在通道整合过程中,只完成张量拼接,并未进行实质性运算。因此,GMTB的计算复杂度主要由单通道GT阶段所决定。
在模块的GT阶段,需根据式(2)分别对输入图像的3个通道进行逐像素变换计算,GT阶段的时间复杂度为Ow×h)。同时,该过程会对输入图像与变换输出结果加以保存,故GT阶段的空间复杂度同样为Ow×h)。
综上,GMTB的时间复杂度与空间复杂度均为Ow×h)。
设输入皮肤镜图像中某个通道的灰度函数为fxy)。通过指定一维函数φψ,并在x轴与y轴方向上进行乘积,可得二维尺度函数φxy)、二维纵向小波ΨHxy)、二维横向小波ΨVxy)、二维对角小波ΨDxy)。
由于上述4个二维函数均可分离,故能够采用快速离散小波变换(Fast Discrete Wavelet Transform,FDWT)算法进行求解。具体运算步骤如下:分别对一维函数φψ构建基函数hφhψ,使用hφhψ对输入图像fxy)逐行求解一维卷积,并逐列进行降采样,进而得到2幅水平方向上具有1/2分辨率的子图像。再次使用hφhψ基函数对上述2个子图像分别进行逐列一维卷积,并逐行进行降采样计算,最终输出4幅长宽均为原图像1/2的子图TDTVTHT。其中,TD包含对角线细节系数,TV包含垂直细节系数,TH包含水平细节系数,T包含图像的原始低频信息。二维FDWT的计算步骤如图4所示。
计算二维快速离散小波逆变换(Fast Inverse Discrete Wavelet Transform,FIDWT)时,与二维FDWT的流程反向对应。其具体计算步骤为:对输入的TDTVTHT四幅图像逐行进行升采样计算,分别使用hφhψhφhψ基函数对4个升采样结果子图逐列计算卷积,将结果分2组并各自相加。对相加后的2个图像分别按列进行升采样计算,并使用hφhψ基函数逐行卷积,最终将卷积后的结果相加并输出。二维FIDWT的计算步骤如图5所示。
WTCB主要使用小波卷积和SF两种技术。在WTCB之前,先对输入的皮肤镜彩色图像进行通道分解,得到R、G、B三个通道。选取单个通道作为原始输入,并对输入图像计算二维FDWT,得到低频-低频图像(LL)、低频-高频图像(LH)、高频-低频图像(HL)、高频-高频图像(HH)共4幅子图。在此基础之上,分别对LLLHHLHH四幅子图进行SF,并对滤波结果计算二维FIDWT,以获得包含丰富空频信息的图像特征。最终将小波逆变换结果与原始输入图像的卷积结果相加,作为WTCB的输出。WTCB的结构如图6所示。
WTCB主要由二维DWT、二维空间域卷积、二维IDWT三部分组成。设输入为一幅w×h大小的彩色皮肤镜图像,其张量封装形状是3×w×h。其中,对该图像计算二维DWT或二维IDWT时,完成一层小波分解所需要的时间复杂度可表示为Ow×h)。若二维卷积操作的卷积核大小是k×k,对DWT所输出的4幅子图分别进行单通道卷积的计算量可表示为k×k×;对原始输入图像进行单通道卷积的计算量可表示为k×k×w×h。将IDWT的输出图像和原图像的空间卷积结果相加,该步骤的复杂度同为Ow×h)。因此,WTCB的时间复杂度可表示为Ok×k×w×h)。
由于二维DWT会令输出图像的长宽均缩小至原图像大小的1/2,故LLLHHLHH四幅小波子图所占用的存储空间是。除此之外,原始输入图像、IDWT输出图像、原图像的空间卷积输出以及模块最终结果的矩阵大小均为w×h,即WTCB的空间复杂度是Ow×h)。
综上所述,WTCB的时间复杂度是Ok×k×w×h),而空间复杂度是Ow×h)。
选取DETR作为皮肤镜图像处理系统的基础架构,并在网络中使用GMTB和WTCB进行特征提取。该系统的图像处理流程表述如下:
①获取HAM10000皮肤镜数据集,将其划分为训练集、验证集、测试集3个部分,并依次读取至计算机内存中。
②将皮肤镜影像输入系统,对图像进行滤波、裁剪、标准化等操作,并封装为张量格式。
③使用GMTB分别调整R、G、B三个通道的灰度,并利用包含WTCB的主干网络对图像特征加以提取。
④利用图像的高度参数h和宽度参数w对输出特征进行位置编码,把原始图像与编码结果输入至改进后的Transformer结构。
⑤把Transformer的输出传至2个线性层,分别预测该患者的皮肤病类型和位置锚框。
⑥以可视化方式输出系统的检测结果。
完整的皮肤镜图像处理系统模型如图7所示。
皮肤癌患者的筛选流程表述如下[24]:
①嘱咐患者暴露并清洗检查区域,确保皮肤表面无污垢等残留物。
②将皮肤镜放置在患者的皮肤表面,调节镜头的焦距和放大倍数,从不同距离、不同角度观察患者皮肤的毛孔、血管等表面结构及皮下结构。
③使用皮肤镜对相应部位拍照或录像,记录患者皮肤的病变情况。
④将图像或视频数据保存至计算机中,利用智能皮肤镜图像处理系统初步判断该患者的疾病类型。
⑤反馈系统判别结果至诊间计算机中,辅助皮肤科医师诊断疾病。
⑥将筛查结果告知患者,并做出相应处理。
实验程序基于PyTorch架构编写,并在具有RTX3080显卡和64 GB内存的人工智能服务器上运行与仿真。为保证实验结果可复现,使用公开的HAM10000皮肤镜数据集[25]进行训练、验证、测试。该数据集共由10 015张皮肤镜图片组成,其中训练集包含7 007张图片,占比70%;验证集包含1 003张图片,占比10%;测试集包含2 005张图片,占比20%。HAM10000皮肤镜数据集的划分方法如图8所示。
使用t-SNE降维聚类算法[26]分别绘制训练集、验证集、测试集的二维散点图,结果如图9所示。
图9可知:在HAM10000数据集中,黑素细胞痣(图中浅蓝色点)的样本数量最多,且在整个平面内均有分布;暖色系所代表的光化性角化病和上皮内癌、基底细胞癌、良性角化病变的样本点则普遍集中于y>0平面区域,且与黑素细胞痣的样本点相互混杂。上述情况表明,在二维情况下,癌症图像样本与细胞痣图像样本之间相似性较大,这极大地增加了识别皮肤疾病的难度。
人类皮肤由表皮、真皮和皮下组织3个部分构成,皮肤中黑色素的含量和分布直接决定皮肤的颜色[2]。在皮肤性病科的临床实践中,皮肤颜色会因患者种族、年龄以及发病部位的不同而产生差异,但皮肤镜检查结果主要呈现为白色、粉色、红色和黑色。因此,可以直接对皮肤镜影像使用彩色图像处理技术,或将彩色图像拆为R、G、B三个通道分别进行灰度图像处理。
疾病的病理学形态是进行临床诊断的重要依据。黑素细胞痣具有清晰的边界,多呈现条状或巢状;基底细胞癌的边界不规则,同样呈现为条状或巢状;黑色素瘤的边界不清晰,且细胞形态多样,具有显著差异。因此,通过仔细观察病变部位的边界与纹理,可有效区分黑素细胞痣、基底细胞癌、黑色素瘤。
GT是一种灰度值变换算法,通过设定不同的伽马参数值,能够对图像中明暗部分进行增强或抑制,从而提升图像整体的对比度。由于患病部位与正常部位的颜色相近,仅使用肉眼难以区分颜色的细微变化,这给皮肤科医师带来巨大困扰。GT在有效增加二者对比度的同时,也更加突出病变区域的边缘,故采用GMTB更有利于识别边界不规则的基底细胞癌及黑色素瘤。
当指数位置的伽马参数满足条件γ<1时,可以提高图像中较暗区域的对比度,同时降低较亮区域的对比度,主要用于增大灰度级偏暗图像的灰度动态范围;当伽马参数满足条件γ>1时,能够提高图像中较亮区域的对比度,降低较暗区域的对比度,进而增加整体偏亮图像的灰度动态范围[18]。皮肤镜图像的关键区域是黑色的细胞痣或细胞癌,因此设定γ=0.9有利于改善图像中目标与背景的对比效果。与此同时,图像产生的颜色变化也可认为是对原数据集进行数据增广,以减小模型训练过程中的过拟合效应。
传统的CNN均采用空域卷积技术,以提取图像的空间域特征。WT则延续傅里叶变换和沃尔什-哈达玛变换的思想,其使用的小波函数在空间域和频率域均具有集中性与平滑性的特点,故WT在处理皮肤镜影像等非平稳信号时优势明显[18]。此外,WT能够根据二维图像自适应地调整空频窗口,并获得可变的空间分辨率和频域分辨率,从而捕捉到病变部位的纹理细节。在骨干网络中添加WTCB后,于传统的空域卷积基础之上补充了图像的频率域特征,同时也能够增大卷积核感受野,提升模型的细粒度分辨能力。因此,充分利用WT的多分辨率特性可以有效地识别黑色素瘤及其他疾病。
作为目标检测领域的一个重要分支,DETR主要由CNN、Transformer、目标检测头3个部分构成。其中,包含降采样的CNN可以有效获得图像的局部特征,并减少数据量;而Transformer具有优良的全局特征提取能力,很大程度上弥补了CNN在视野受限方面的不足。因此,具有混合结构的SFTNet模型比YOLO系列模型更具性能优势。
在图像处理系统中,骨干网络的关键作用是提取图像特征。而选取不当的CNN可能给系统带来严重的性能瓶颈。由于第一代CNN的LeNet-5[27]模型无法有效地提取复杂图像特征,因此选取Alex-Net[28]、VGG[29]、GoogLeNet[30]、ResNet[31]四种经典CNN作为本次消融实验的基础结构。训练与验证过程中,Accuracy、Precision、F1、AP四种模型评价指标的变化曲线如图10所示。
图10(a)图10(c)图10(e)图10(g)所示,选取AlexNet、VGG和GoogLeNet作为系统骨干网络时,三者在训练和验证过程中的Accuracy曲线、Precision曲线、F1曲线相近,且数值均远低于以ResNet为骨干网络时的曲线。由此可知,当数据集与基础参数保持相同时,AlexNet、VGG、GoogLeNet的特征提取效果相似,且ResNet模型的特征提取能力优于前述三者,即以ResNet为基础的骨干网络更适合于该系统的特征提取工作。当训练轮数epoch<5时,图10(d)图10(h)中的AP曲线快速下降,但在epoch>5后彼此相互重叠。这表明,不论使用何种CNN作为基础模型,训练或验证过程中的AP曲线变化趋势均无明显差异。
对比4幅训练曲线和4幅验证曲线可知,相较于平缓的训练过程曲线,验证过程中的Accuracy、Precision、F1以及AP指标均产生大范围波动。产生该现象的根本原因在于,小批量随机梯度下降算法令训练损失函数的优化方向随epoch的变化而变化,进而导致验证过程中各评价指标的数值不稳定。
模块消融实验共分3组。第1组和第2组分别使用ResNet18、ResNet34、ResNet50[31]作为系统骨干网络,第3组实验中加入ResNet101[31]模型进行对比。3组消融实验的各项评价指标如表2所示。
在上述3组消融实验中,以ResNet34为骨干网络时模型的各项评价指标均为最优,ResNet18次之。而以ResNet50为骨干网络时,模型的各项指标均为最差。分析原因在于,ResNet18结构的卷积层数量较少,无法有效地提取深层次特征;ResNet50结构的卷积层数量过多,导致模型出现轻度过拟合现象;第3组实验中使用的ResNet101结构则展现出更为严重的过拟合效应。因此,针对HAM10000皮肤镜图像数据集,卷积深度居于二者之间的Res-Net34模型可以拥有比其他模型更优秀的特征提取效果。
相较于未做出任何改进的第1组实验,加入WTCB后,WTCB-R18和WTCB-R34模型在TP、Accuracy、Precision、F1四个关键指标上均有提升。以ResNet18为骨干网络时,4项指标的提升幅度分别达到3.53%、3.3%、3.53%、2.03%;以ResNet34为骨干网络时,模型指标的提升幅度分别为0.87%、0.87%、0.87%、0.50%;而TN、FP、AP三项指标的数值则略有下降。除了ResNet50结构出现过拟合与训练瓶颈的现象,导致WTCB-R50模型并未发生实质性改变外,在原有ResNet结构的基础之上添加WTCB可带来明显的性能提升。
在上述实验的基础之上加入GMTB,并采用迁移学习策略,进行第3组消融实验。由第3组实验结果可得,模型的TP、Accuracy、Precision、F1四项指标均有大幅提升。其中,SFTNet-R18模型的指标提升幅度为10.43%、10.51%、10.43%、5.72%;SFT-Net-R34模型的提升幅度为13.41%、13.30%、13.42%、7.24%;SFTNet-R50模型的提升幅度达到25.5%、26.06%、25.51%、13.86%。对比第2组和第3组消融实验可知,在WTCB-ResNet模型中添加GMTB后,模型性能提升显著。模型性能获得有效提升的重要原因在于,GMTB改变了原图像中R、G、B三个通道的灰度值,进而增加图像对比度。
测试集混淆矩阵的二维热力图如图11所示。
图11中,混淆矩阵的行表示由皮肤科医师标注的真实疾病类别标签,混淆矩阵的列表示系统的诊断结果。阿拉伯数字1~7分别代表日光性角化病和原位癌、基底细胞癌、良性角化病变、皮肤纤维瘤、黑色素瘤、黑素细胞痣、血管病变7种常见皮肤疾病。
图11可知,混淆矩阵中各行的最大值均位于主对角线区域,表明系统的检测输出与专业皮肤科医师的标准诊断结果基本保持一致。除本类之外,各行中次高的元素值均分布在第6列,说明系统将部分患者的皮肤镜图像错误识别为黑素细胞痣。
对各行数据分别求和并计算正确率可知,智能皮肤癌筛选系统对日光性角化病和原位癌、基底细胞癌、良性角化病变、皮肤纤维瘤、黑色素瘤、黑素细胞痣、血管病变7种疾病的诊断正确率分别达到0.666 7、0.728 2、0.586、0.434 8、0.511 2、0.980 6、0.724 1。上述数据表明,系统对黑素细胞痣的检测正确率远超其他疾病,但皮肤纤维瘤的诊断精度较低。分析原因有二:一方面,黑素细胞痣患者的样本数量过多,且其他6种疾病患者的样本数量较少,这导致系统在训练过程中更倾向于学习皮肤痣样本的图像特征;另一方面,由于基底细胞癌、黑色素瘤的皮肤镜图像与黑素细胞痣的皮肤镜图像高度相似,而系统模型未能很好地分辨这3个疾病的细节。
本文模型与数据集基准模型的测试集正确率如表3所示。
由于MedMNIST数据集是HAM10000皮肤病数据集的延伸,因此可将二者的结果对比评价。在MedMNIST-v1[32]和MedMNIST-v2[33]两个数据集的原始论文中,作者均使用ResNet18、ResNet50、Auto-sklearn、Autokeras、Google AutoML Vision五种模型进行实验。其中,ResNet18和ResNet50为经典CNN,2个模型的平均正确率是0.733,与本文4.4节中消融实验的平均正确率0.696相接近。Auto-sklearn[34]、Auto-keras[35]、Google AutoML Vision均为自动调参模型,三者的平均正确率达到0.749,略超ResNet系列。本文在DETR架构的基础上加入GMTB与WTCB,模型正确率可达0.855。相较于文献[32-33],SFTNet模型的正确率指标平均值提升15.38%,改进效果明显。
不同深度学习模型在HAM10000皮肤病测试数据集上的评价指标参见表4。如表4所示,使用LeNet[27]、VGG[29]、GoogLeNet[30]、ResNet[31]、BoT-Net[35-36]、GhostNet[37]、MobileNet[38]、MobileNet-v2[39]、MobileNet-v3[40]、MobileViT-v3[41]共10种基础模型及其衍生版本进行性能对比实验。
表4可知,Nugroho等[42]对LeNet架构进行改进,该模型的Accuracy、Precision、Recall和F1指标分别为0.780、0.414、0.414、0.414,在所有模型中表现最差;而本文提出的SFTNet在HAM10000测试数据集上表现优秀,其Accuracy、Precision、Recall和F1分别达到0.855、0.852、0.851、0.851,在17个模型中最具优势。但Patil等[45]所使用的ResNet模型在该数据集上的Recall和F1为0.975和0.908,2项指标远高于其他模型。出现异常数值的原因可能是作者对测试数据集进行调整,或实验数据记录有误。
在均衡性层面,VGG、ResNet、GhostNet、MobileNet、MobileViT-v3、MobileNet-v3、SFTNet模型的4种评价指标之间方差较小,而LeNet、GoogLeNet、BoTNet、MobileNet四种模型的评价指标之间方差较大。这表明,前述7种模型的性能比后4种更加均衡。
进一步计算表4数据可得,MobileNet系列模型的4个评价指标均值分别是0.813、0.755、0.815、0.759;MobileNet-v2系列的4个评价指标均值分别是0.827、0.740、0.658、0.668;MobileNet-v3系列的均值则是0.845、0.836、0.808、0.812。由上述结果可知,从MobileNet到MobileNet-v2再到MobileNet-v3,正确率指标的数值逐步提升。这表明,随着Mo-bileNet系列架构的迭代升级,其模型性能也有所提高。除MobileNet-v3模型的召回率略小于MobileNet模型外,其Accuracy、Precision和F1三个评价指标均为最优,而MobileNet-v2模型在上述3个评价指标中表现均为最差。
不同深度学习模型在同一指标下的数据对比如图12所示。
图12可知,在Recall与F1方面,MobileViT-v3模型与BoTNet、MobileNet-v2、GoogLeNet、LeNet模型之间出现显著断层。针对正确率指标,11类模型的性能排序为SFTNet>ResNet>MobileViT-v3>VGG>MobileNet-v3>GhostNet>MobileNet-v2>BoT-Net>MobileNet>GoogLeNet>LeNet;而针对精确率指标,11类模型的性能排序为SFTNet>VGG>Res-Net>GhostNet>MobileNet-v3>MobileViT-v3>MobileNet>MobileNet-v2>BoTNet>GoogLeNet>LeNet。上述结果显示,不同模型在正确率和精确率层面上互有优劣。总体而言,SFTNet、ResNet、MobileNet-v3三类模型的性能较强,而GoogLeNet、LeNet的性能偏弱。模型的正确率过低暗示其极易出现误诊现象,并不适合应用在皮肤病诊断的临床场景。
本文创新性地提出基于GT和小波卷积的图像处理模块,并于DETR架构的基础上加以改进,在皮肤病检测方面获得较佳效果。仿真实验表明,GMTB能够提升皮肤镜图像的对比度,减少模型训练过程中的过拟合效应;WTCB在原有的空域信息基础上补充频域信息,增大卷积核感受野的同时捕捉更多图像细节[50]。该研究的不足之处在于,在提升系统整体检测性能的同时,GMTB和WTCB也增加了模型的训练成本,且系统针对皮肤纤维瘤和黑色素瘤的诊断正确率较低,即系统仍有一定改进空间。
由此可见,基于SFTNet的智能皮肤癌预筛系统不仅能够成功识别患者的皮肤疾病,也有效地减轻了皮肤性病科医师的工作负担。该系统可以对皮肤癌患者进行快速筛查和预警,适合在各级医疗单位推广与应用。
参考文献 引证文献
排序方式:
[1]
穆亚楠,陈星会,何璐,.常见非黑色素瘤皮肤癌新兴治疗的研究进展[J/OL]. (2024-11-18)[2025-04-18]. http://kns.cnki.Net/kcms/detail/61.1415.R.20241115.2350.007.html.
[2]
崔勇,高兴华.皮肤病性病学[M]. 10版.北京:人民卫生出版社,2024.
[3]
刘道远.皮肤镜诊断常见非黑色素瘤皮肤癌的临床价值研究[D].昆明:昆明医科大学,2023.
[4]
邓凡智.皮肤镜在基底细胞癌诊断中的应用价值[D].蚌埠:蚌埠医学院,2021.
[5]
崔莉,王贵佐,马小萍,.皮肤镜检查在黑素细胞痣、脂溢性角化病和基底细胞癌中的诊断价值[J].陕西医学杂志,2021,50(4):428-431.
[6]
江浩波,孙志平,陈海针,.皮肤镜在脂溢性角化、日光性角化和基底细胞癌诊断中的应用[J].皮肤性病诊疗学杂志,2019,26(4):204-209.
[7]
LIU W,ANGUEVOV D,ERHAN D,et al. SSD:Single Shot MultiBox Detector[C]//European Conference on Computer Vision. Cham:Springer,2016:21-37.
[8]
GIRSHICK R, DONAHUE J, DARRELL T, et al. Rich Feature Hierarchies for Accurate Object Detection and Semantic Segmentation[C]//Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition. Columbus:IEEE,2014:580-587.
[9]
GIRSHICK R. Fast R-CNN[C]//Proceedings of the 2015 IEEE International Conference on Computer Vision. Santiago:IEEE,2015:1440-1448.
[10]
REN S Q,HE K M,GIRSHICK R,et al. Faster R-CNN:Towards Real-time Object Detection with Region Proposal Networks[J]. Advances in Neural Information Processing Systems. IEEE,2015,28:1137-1149.
[11]
REDMON J,DIVVALA S,GIRSHICK R,et al. You Only Look Once:Unified,Real-time Object Detection[C]//Proceedings of the 2016 IEEE Conference on Computer Vision and Pattern Recognition. Las Vegas:IEEE,2016:779-788.
[12]
VASWANI A,SHAZEER N,PARMAR N,et al. Attention is All You Need[J]. Advances in Neural Information Processing Systems,2017,30:5998-6008.
[13]
CARION N,MASSA F,SYNNAEVE G,et al. End-to-End Object Detection with Transformers[C]//Proceedings of the 2020 European Conference on Computer Vision. Glasgow:Springer,2020:213-229.
[14]
赵嘉晖,温杰,蔡星娟,.基于高维多目标优化的小样本皮肤癌检测[J].控制与决策, 2024,39(11):3597-3606.
[15]
王煜坤.皮肤影像及人工智能辅助皮肤疾病精准诊疗新探索[D].北京:北京协和医学院,2024.
[16]
TSCHANDL P,RINNER C,APALLA Z,et al. Human computer Collaboration for Skin Cancer Recognition[J]. Nature Medicine,2020,26(8):1229-1234.
[17]
GONZALEZ R C,WOODS R E. Digital Image Processing[M].北京:电子工业出版社,2020.
[18]
禹晶,肖创柏,廖庆敏.数字图像处理[M].北京:清华大学出版社,2022.
[19]
FOURIER J B J. Théorie Analytique de la Chaleur[M]. Paris:Firmin Didot,1822.
[20]
AHMED N, NATARAJAN T, RAO K R. Discrete Cosine Transform[J]. IEEE Transactions on Computers,1974,23(1):90-93.
[21]
DEVANEY A J. The Slant Transform: An Orthogonal Transform for Signal Processing[J]. IEEE Transactions on Audio and Electroacoustics,1972,20(2):140-146.
[22]
WALSH J L. A Closed Set of Normal Orthogonal Functions[J]. Transactions of the American Mathematical Society,1923,25(1):5-24.
[23]
DAUBECHIES I. Ten Lectures on Wavelets[M]. Philadelphia:Society for Industrial and Applied Mathematics, 1992.
[24]
李艳玲.皮肤镜在颜面部皮肤肿瘤筛检中的应用[J].中国现代医生,2022,60(3):21-23.
[25]
TSCHANDL P, ROSENDAHL C, KITTLER H. The HAM10000 Dataset, a Large Collection of Multi-source Dermatoscopic Images of Common Pigmented Skin Lesions[J]. Scientific Data,2018,5:1-10.
[26]
LAURENS V D M,HINTON G E. Visualizing Data Using t-SNE[J]. Journal of Machine Learning Research,2008,9(86):2579-2605.
[27]
LECUN Y,BOTTOU L,BENGIO Y,et al. Gradient-based Learning Applied to Document Recognition[J]. Proceedings of the IEEE,1998,86(11):2278-2324.
[28]
KRIZHEVSKY A,SUTSKEVER I,HINTON G E. Imagenet Classification with Deep Convolutional Neural Networks[C]//Advances in Neural Information Processing Systems. Lake Tahoe:Curran Associates Inc.,2012:1097-1105.
[29]
SIMONYAN K,ZISSERMAN A. Very Deep Convolutional Networks for Large-scale Image Recognition[C]//Proceedings of the 3rd International Conference on Learning Representations (ICLR 2015). San Diego:OpenReview, 2015:1-14.
[30]
SZEGEDY C,LIU W,JIA Y Q,et al. Going Deeper with Convolutions[C]//Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition. Boston:IEEE, 2015:1-9.
[31]
HE K M,ZHANG X Y,REN S Q,et al. Delving Deep into Rectifiers:Surpassing Human-level Performance on Imagenet Classification[C]//Proceedings of the IEEE International Conference on Computer Vision. Santiago:IEEE, 2015:1026-1034.
[32]
YANG J C,SHI R,NI B B. MedMNIST Classification Decathlon:A Lightweight Automl Benchmark for Medical Image Analysis[C]//IEEE 18th International Symposium on Biomedical Imaging(ISBI). Piscataway:IEEE,2021:191-195.
[33]
YANG J C, SHI R, WEI D L, et al. MedMNIST v2: A Large-scale Lightweight Benchmark for 2D and 3D Biomedical Image Classification[J]. Scientific Data,2023, 10(1):1-41.
[34]
FEURER M,KLEIN A,EGGENSPERGER K,et al. Efficient and Robust Automated Machine Learning[C]//Advances in Neural Information Processing Systems. Montreal:Curran Associates Inc.,2015:2962-2970.
[35]
JIN H F, SONG Q Q, HU X. Auto-keras: An Efficient Neural Architecture Search System[C]//Proceedings of the 25th ACM SIGKDD International Conference on Knowledge Discovery & Data Mining. Anchorage:ACM, 2018:1946-1956.
[36]
ARAVIND S, TSUNG-YI L, NIKI P, et al. Bottleneck Transformers for Visual Recognition[EB/OL].[2025-04-10].https://arxiv.org/pdf/2101.11605.
[37]
HAN K,WANG Y H,TIAN Q,et al. GhostNet:More Features from Cheap Operations[C]//2020 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR). Piscataway:IEEE,2020:1580-1589.
[38]
HOWARD A G,ZHU M L,CHEN B,et al. MobileNets:Efficient Convolutional Neural Networks for Mobile Vision Applications[EB/OL]. (2017-04-17)[2025-04-18]. https://arxiv.org/pdf/1704.04861.
[39]
SANDLER M, HOWARD A, ZHU M, et al. MobileNetV2:Inverted Residuals and Linear Bottlenecks[C]//2018 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR). Piscataway:IEEE,2018:4510-4520.
[40]
HOWARD A,SANDLER M,CHU G,et al. Searching for MobileNetV3[C]//2019 IEEE/CVF International Conference on Computer Vision (ICCV). New York: IEEE, 2019:1314-1324.
[41]
WADEKAR S N, CHAURASIA A. MobileViTv3: MobileFriendly Vision Transformer with Simple and Effective Fusion of Local, Global and Input Features[EB/OL]. (2022-09-01)[2025-04-18]. https://arxiv.org/pdf/2209.15159v2.
[42]
NUGROHO A A,SLAMET I,SUGIYANTO. Skins Cancer Identification System of HAM10000 Skin Cancer Dataset Using Convolutional Neural Network[EB/OL]. (2019-12-27)[2025-04-18]. https://doi.org/10.1063/1.5141652.
[43]
RATUL A R,MOZAFFARI M H, LEE W S,et al. Skin Lesions Classification Using Deep Learning Based on Dilated Convolution[EB/OL]. (2020-06-05)[2025-04-18]. https://doi.org/10.1101/860700.
[44]
THURNHOFER H K, LÓPEZ R E, DOMÍNGUEZ E, et al. Skin Lesion Classification by Ensembles of Deep Convolutional Networks and Regularly Spaced Shifting[J]. IEEE Access,2021,9:112193-112205.
[45]
PATIL S N,PATIL H D,ADHIYA K P,et al. Densenet201 for Skin Melanoma Classification: A Comprehensive Performance Evaluation and Analysis[J]. International Research Journal of Multidisciplinary Scope,2024,5(4):711-721.
[46]
LIU H,DOU Y,WANG K,et al. A Skin Disease Classification Model Based on Multi Scale Combined Efficient Channel Attention Module[J]. Scientific Reports,2025, 15:1-12.
[47]
胡吉红.基于轻量级网络的皮肤病检测算法研究[D].兰州:西北师范大学,2023.
[48]
SAE L W,WETTAYAPRASIT W,AIYARAK P. Convolutional Neural Networks Using MobileNet for Skin Lesion Classification[C]//2019 IEEE Conference on Computer Vision and Pattern Recognition Workshops (CVPRW). Long Beach:IEEE,2019:242-247.
[49]
CHATURVEDI S S,GUPTA K,PRASAD P S. Skin Lesion Analyser:An Efficient Seven-way Multi-class Skin Cancer Classification Using MobileNet[C]//Advanced Machine Learning Technologies and Applications: Proceedings of AMLTA 2020. Singapore:Springer,2021:165-176.
[50]
朱秋慧,杨靖,黄若愚,.基于部分卷积的多尺度特征卷积神经网络模型[J/OL]. (2025-03-11)[2025-04-18].http://kns.cnki.net/kcms/detail/13.1099.TN.20250310.1707.012.html.
2025年第51卷第5期
PDF下载
68
30
引用本文
BibTeX
文章信息
doi: 10.3969/j.issn.1003-3114.2025.05.012
  • 接收时间:2025-04-22
  • 首发时间:2026-04-17
  • 出版时间:2025-09-18
补充材料
相关文章
文章信息
作者
出版历史
  • 收稿日期:2025-04-22
基金
作者信息
    河北工程大学 信息与电气工程学院,河北 邯郸 056038

通讯作者:

李丽宏 女,(1973—),博士,教授。主要研究方向:图像处理、计算机视觉。
参考文献
分享链接
https://castjournals.cast.org.cn/joweb/wxdtxjs/CN/10.3969/j.issn.1003-3114.2025.05.012
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

Family
属数
Number of
genus
种数
Number of
species
占总种数比例
Percentage of
total species (%)

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