Article(id=1244321222117409754, tenantId=1146029695717560320, journalId=1244284848500682798, issueId=1244321215637209904, articleNumber=null, orderNo=null, doi=10.16156/j.1004-7220.2025.05.031, pmid=null, cstr=null, oa=null, hot=null, price=null, onlineType=0, articleFormat=0, articleType=null, articleTypeStr=null, receivedDate=1736179200000, receivedDateStr=2025-01-07, revisedDate=1739203200000, revisedDateStr=2025-02-11, acceptedDate=null, acceptedDateStr=null, onlineDate=1774598897722, onlineDateStr=2026-03-27, pubDate=1759248000000, pubDateStr=2025-10-01, doiRegisterDate=null, doiRegisterDateStr=null, onlineIssueDate=1774598897722, onlineIssueDateStr=2026-03-27, onlineJustAcceptDate=null, onlineJustAcceptDateStr=null, onlineFirstDate=null, onlineFirstDateStr=null, sourceXml=null, magXml=null, createTime=1774598897722, creator=13701087609, updateTime=1774598897722, updator=13701087609, issue=Issue{id=1244321215637209904, tenantId=1146029695717560320, journalId=1244284848500682798, year='2025', volume='40', issue='5', pageStart='1079', pageEnd='1366', issueExtLink='null', onlineDate='null', pubDate='null', beforeIssueId=null, nextIssueId=null, price=null, status=1, issueComplete=1, articleOrder=1, issueType=1, specialIssue=null, createTime=1774598896178, creator=13701087609, updateTime=1774599509568, updator=13701087609, preIssue=null, nextIssue=null, ext={EN=IssueExt(id=1244323788452639476, tenantId=1146029695717560320, journalId=1244284848500682798, issueId=1244321215637209904, language=EN, specialIssueTitle=, coverIllustrator=null, specialIssueEditor=, specialIssueAbout=), CN=IssueExt(id=1244323788452639477, tenantId=1146029695717560320, journalId=1244284848500682798, issueId=1244321215637209904, language=CN, specialIssueTitle=, coverIllustrator=null, specialIssueEditor=, specialIssueAbout=)}, issueFiles=null}, startPage=1318, endPage=1324, ext={EN=ArticleExt(id=1244321222473925624, articleId=1244321222117409754, tenantId=1146029695717560320, journalId=1244284848500682798, language=EN, title=Determination of Sputum Suction Timing in Mechanical Ventilation Based on Transfer Learning and Breath Sounds Recognition, columnId=1244321216404767539, journalTitle=Journal of Medical Biomechanics, columnName=Original Articles, runingTitle=null, highlight=null, articleAbstract=
Objective

To propose a transfer learning-based method for breath sound feature recognition and autonomous determination of sputum suction timing.

Methods

An electronic stethoscope was used to collect breath sounds from the main airways of clinically ventilated patients before and after sputum suction, with pre-suction breath sounds labeled as requiring suction. The collected data underwent high-pass filtering and wavelet soft-threshold denoising, followed by the extraction of log-Mel spectrograms. A VGGish model pretrained on the Audio Set dataset was then employed to extract feature vectors from these spectrograms, which were subsequently classified using a support vector machine to determine whether suction was required.

Results

The precision, recall and F1 score for recognition of breath sounds requiring sputum suction were 86.73%, 93.06% and 89.78%, respectively.

Conclusions

The proposed breath sound recognition method based on transfer learning effectively determines the timing of sputum suction and shows a significant clinical potential.

, correspAuthors=Jiangzhen GUO, Chunjing TAO, 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=Shuai WANG, Jiangzhen GUO, Chunjing TAO), CN=ArticleExt(id=1244321226278158578, articleId=1244321222117409754, tenantId=1146029695717560320, journalId=1244284848500682798, language=CN, title=基于迁移学习与呼吸音特征识别的机械通气吸痰时机判别, columnId=1244321216576734006, journalTitle=医用生物力学, columnName=论著, runingTitle=null, highlight=null, articleAbstract=
目的

提出基于迁移学习的呼吸音特征识别与吸痰时机自主判断方法。

方法

使用电子听诊器采集临床机械通气患者在吸痰前、后的主气道呼吸音,将吸痰前的呼吸音标记为“需要吸痰”类别。对采集数据进行高通滤波和小波软阈值滤波去噪,提取对数梅尔频谱图,并利用在Audio Set数据集上预训练的VGGish模型提取对数梅尔频谱图的特征向量,最后通过支持向量机分类器对特征向量进行分类,判断是否需要吸痰。

结果

对需要吸痰的呼吸音识别的精确率、召回率和F1分数分别为86.73%、93.06%和89.78%。

结论

基于迁移学习的呼吸音识别方法能够有效判断吸痰时机,在临床应用中具有潜在价值。

, correspAuthors=郭江真, 陶春静, authorNote=null, correspAuthorsNote=
郭江真,副教授,E-mail:
陶春静,教授,E-mail:

*

为共同通信作者
, copyrightStatement=null, copyrightOwner=《医用生物力学》编辑部, extLink=null, articleAbsUrl=null, sourceXml=AyeZMR6AadPsFMPZZ9VBlw==, magXml=m6XNOsnU0XOYB54mM6o90Q==, pdfUrl=null, pdf=IVOGId5oZUjYWs3o073CHQ==, pdfFileSize=4143716, pdfExtLink=null, richHtmlUrl=null, mobilePdfUrl=null, reviewReport=null, pdfFirstPage=null, abstractGraph=gvyOVvYFLKuBWvMEA0KHAw==, abstractGraphContent=null, abstractVideo=null, citation=null, cebUrl=null, magXmlContent=3WUMrxXFm3zck5rrYbFedw==, mapNumber=null, authorCompany=null, fund=null, authors=

作者贡献声明:

汪帅负责模型搭建与测试,论文撰写与修改;郭江真与陶春静负责资源协调、研究监督、项目统筹与论文审核等工作。

, authorsList=汪帅, 郭江真, 陶春静)}, authors=[Author(id=1244321226626285850, tenantId=1146029695717560320, journalId=1244284848500682798, articleId=1244321222117409754, 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=1244321226735337764, tenantId=1146029695717560320, journalId=1244284848500682798, articleId=1244321222117409754, authorId=1244321226626285850, language=EN, stringName=Shuai WANG, firstName=Shuai, middleName=null, lastName=WANG, prefix=null, suffix=null, authorComment=null, nameInitials=null, affiliation=null, department=null, xref=null, address=School of Engineering Medicine; Beijing Advanced Innovation Center for Biomedical Engineering, Beihang University, Beijing 100191, China, bio=null, bioImg=null, bioContent=null, aboutCorrespAuthor=null), CN=AuthorExt(id=1244321226840195376, tenantId=1146029695717560320, journalId=1244284848500682798, articleId=1244321222117409754, authorId=1244321226626285850, language=CN, stringName=汪帅, firstName=null, middleName=null, lastName=null, prefix=null, suffix=null, authorComment=null, nameInitials=null, affiliation=null, department=null, xref=null, address=北京航空航天大学 医学科学与工程学院;北京生物医学工程高精尖创新中心,北京 100191, bio=null, bioImg=null, bioContent=null, aboutCorrespAuthor=null)}, companyList=[AuthorCompany(id=1244321226521428238, tenantId=1146029695717560320, journalId=1244284848500682798, articleId=1244321222117409754, xref=null, ext=[AuthorCompanyExt(id=1244321226525622543, tenantId=1146029695717560320, journalId=1244284848500682798, articleId=1244321222117409754, companyId=1244321226521428238, language=EN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=School of Engineering Medicine; Beijing Advanced Innovation Center for Biomedical Engineering, Beihang University, Beijing 100191, China), AuthorCompanyExt(id=1244321226534011152, tenantId=1146029695717560320, journalId=1244284848500682798, articleId=1244321222117409754, companyId=1244321226521428238, language=CN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=北京航空航天大学 医学科学与工程学院;北京生物医学工程高精尖创新中心,北京 100191)])]), Author(id=1244321226961830200, tenantId=1146029695717560320, journalId=1244284848500682798, articleId=1244321222117409754, orderNo=1, firstName=null, middleName=null, lastName=null, nameCn=null, orcid=null, stid=null, country=null, authorPic=null, dead=0, email=jzguo@buaa.edu.cn, emailSecond=null, emailThird=null, correspondingAuthor=1, authorType=1, ext={EN=AuthorExt(id=1244321227133796674, tenantId=1146029695717560320, journalId=1244284848500682798, articleId=1244321222117409754, authorId=1244321226961830200, language=EN, stringName=Jiangzhen GUO, firstName=Jiangzhen, middleName=null, lastName=GUO, prefix=null, suffix=null, authorComment=null, nameInitials=null, affiliation=null, department=null, xref=null, address=School of Engineering Medicine; Beijing Advanced Innovation Center for Biomedical Engineering, Beihang University, Beijing 100191, China, bio=null, bioImg=null, bioContent=null, aboutCorrespAuthor=null), CN=AuthorExt(id=1244321227297374552, tenantId=1146029695717560320, journalId=1244284848500682798, articleId=1244321222117409754, authorId=1244321226961830200, language=CN, stringName=郭江真, firstName=null, middleName=null, lastName=null, prefix=null, suffix=null, authorComment=null, nameInitials=null, affiliation=null, department=null, xref=*, address=北京航空航天大学 医学科学与工程学院;北京生物医学工程高精尖创新中心,北京 100191, bio=null, bioImg=null, bioContent=null, aboutCorrespAuthor=null)}, companyList=[AuthorCompany(id=1244321226521428238, tenantId=1146029695717560320, journalId=1244284848500682798, articleId=1244321222117409754, xref=null, ext=[AuthorCompanyExt(id=1244321226525622543, tenantId=1146029695717560320, journalId=1244284848500682798, articleId=1244321222117409754, companyId=1244321226521428238, language=EN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=School of Engineering Medicine; Beijing Advanced Innovation Center for Biomedical Engineering, Beihang University, Beijing 100191, China), AuthorCompanyExt(id=1244321226534011152, tenantId=1146029695717560320, journalId=1244284848500682798, articleId=1244321222117409754, companyId=1244321226521428238, language=CN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=北京航空航天大学 医学科学与工程学院;北京生物医学工程高精尖创新中心,北京 100191)])]), Author(id=1244321227398037856, tenantId=1146029695717560320, journalId=1244284848500682798, articleId=1244321222117409754, orderNo=2, firstName=null, middleName=null, lastName=null, nameCn=null, orcid=null, stid=null, country=null, authorPic=null, dead=0, email=chunjingtao@buaa.edu.cn, emailSecond=null, emailThird=null, correspondingAuthor=1, authorType=1, ext={EN=AuthorExt(id=1244321227532255601, tenantId=1146029695717560320, journalId=1244284848500682798, articleId=1244321222117409754, authorId=1244321227398037856, language=EN, stringName=Chunjing TAO, firstName=Chunjing, middleName=null, lastName=TAO, prefix=null, suffix=null, authorComment=null, nameInitials=null, affiliation=null, department=null, xref=null, address=School of Engineering Medicine; Beijing Advanced Innovation Center for Biomedical Engineering, Beihang University, Beijing 100191, China, bio=null, bioImg=null, bioContent=null, aboutCorrespAuthor=null), CN=AuthorExt(id=1244321229079953801, tenantId=1146029695717560320, journalId=1244284848500682798, articleId=1244321222117409754, authorId=1244321227398037856, language=CN, stringName=陶春静, firstName=null, middleName=null, lastName=null, prefix=null, suffix=null, authorComment=null, nameInitials=null, affiliation=null, department=null, xref=*, address=北京航空航天大学 医学科学与工程学院;北京生物医学工程高精尖创新中心,北京 100191, bio=null, bioImg=null, bioContent=null, aboutCorrespAuthor=null)}, companyList=[AuthorCompany(id=1244321226521428238, tenantId=1146029695717560320, journalId=1244284848500682798, articleId=1244321222117409754, xref=null, ext=[AuthorCompanyExt(id=1244321226525622543, tenantId=1146029695717560320, journalId=1244284848500682798, articleId=1244321222117409754, companyId=1244321226521428238, language=EN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=School of Engineering Medicine; Beijing Advanced Innovation Center for Biomedical Engineering, Beihang University, Beijing 100191, China), AuthorCompanyExt(id=1244321226534011152, tenantId=1146029695717560320, journalId=1244284848500682798, articleId=1244321222117409754, companyId=1244321226521428238, language=CN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=北京航空航天大学 医学科学与工程学院;北京生物医学工程高精尖创新中心,北京 100191)])])], keywords=[Keyword(id=1244321229218365848, tenantId=1146029695717560320, journalId=1244284848500682798, articleId=1244321222117409754, language=EN, orderNo=1, keyword=mechanical ventilation), Keyword(id=1244321229386138018, tenantId=1146029695717560320, journalId=1244284848500682798, articleId=1244321222117409754, language=EN, orderNo=2, keyword=breath sound), Keyword(id=1244321229507772846, tenantId=1146029695717560320, journalId=1244284848500682798, articleId=1244321222117409754, language=EN, orderNo=3, keyword=transfer learning), Keyword(id=1244321229629407671, tenantId=1146029695717560320, journalId=1244284848500682798, articleId=1244321222117409754, language=EN, orderNo=4, keyword=determination of sputum suction time), Keyword(id=1244321229734265284, tenantId=1146029695717560320, journalId=1244284848500682798, articleId=1244321222117409754, language=CN, orderNo=1, keyword=机械通气), Keyword(id=1244321229855900114, tenantId=1146029695717560320, journalId=1244284848500682798, articleId=1244321222117409754, language=CN, orderNo=2, keyword=呼吸音), Keyword(id=1244321229969146334, tenantId=1146029695717560320, journalId=1244284848500682798, articleId=1244321222117409754, language=CN, orderNo=3, keyword=迁移学习), Keyword(id=1244321230074003947, tenantId=1146029695717560320, journalId=1244284848500682798, articleId=1244321222117409754, language=CN, orderNo=4, keyword=吸痰时机判断)], refs=[Reference(id=1244321231562982003, tenantId=1146029695717560320, journalId=1244284848500682798, articleId=1244321222117409754, doi=null, pmid=null, pmcid=null, year=2013, volume=8, issue=2, pageStart=76, pageEnd=82, url=null, language=null, rfNumber=[1], rfOrder=0, authorNames=STEFAN MS, SHIEH MS, PEKOW PS, journalName=J Hosp Med, refType=null, unstructuredReference=STEFAN MS, SHIEH MS, PEKOW PS, et al. Epidemiology and outcomes of acute respiratory failure in the United States, 2001 to 2009: A national survey[J]. J Hosp Med, 2013, 8(2): 76-82., articleTitle=Epidemiology and outcomes of acute respiratory failure in the United States, 2001 to 2009: A national survey, refAbstract=null), Reference(id=1244321231676228220, tenantId=1146029695717560320, journalId=1244284848500682798, articleId=1244321222117409754, doi=null, pmid=null, pmcid=null, year=2013, volume=188, issue=2, pageStart=220, pageEnd=230, url=null, language=null, rfNumber=[2], rfOrder=1, authorNames=BEZZI M, journalName=Am J Respir Crit Care Med, refType=null, unstructuredReference=BEZZI M. Evolution of mortality over time in patients receiving mechanical ventilation[J]. Am J Respir Crit Care Med, 2013, 188(2): 220-230., articleTitle=Evolution of mortality over time in patients receiving mechanical ventilation, refAbstract=null), Reference(id=1244321231772697221, tenantId=1146029695717560320, journalId=1244284848500682798, articleId=1244321222117409754, doi=null, pmid=null, pmcid=null, year=2018, volume=37, issue=1, pageStart=44, pageEnd=56, url=null, language=null, rfNumber=[3], rfOrder=2, authorNames=FISK AC, journalName=Dimens Crit Care Nurs, refType=null, unstructuredReference=FISK AC. The effects of endotracheal suctioning in the pediatric population: An integrative review[J]. Dimens Crit Care Nurs, 2018, 37(1): 44-56., articleTitle=The effects of endotracheal suctioning in the pediatric population: An integrative review, refAbstract=null), Reference(id=1244321231873360524, tenantId=1146029695717560320, journalId=1244284848500682798, articleId=1244321222117409754, doi=null, pmid=null, pmcid=null, year=2009, volume=28, issue=5, pageStart=281, pageEnd=302, url=null, language=null, rfNumber=[4], rfOrder=3, authorNames=GARDNER D, SHIRLAND L, journalName=Neonatal Netw, refType=null, unstructuredReference=GARDNER D, SHIRLAND L. Evidence-based guideline for suctioning the intubated neonate and infant[J]. Neonatal Netw, 2009, 28(5): 281-302., articleTitle=Evidence-based guideline for suctioning the intubated neonate and infant, refAbstract=null), Reference(id=1244321231982412436, tenantId=1146029695717560320, journalId=1244284848500682798, articleId=1244321222117409754, doi=null, pmid=null, pmcid=null, year=2018, volume=39, issue=4, pageStart=797, pageEnd=808, url=null, language=null, rfNumber=[5], rfOrder=4, authorNames=METERSKY ML, KALIL AC, journalName=Clin Chest Med, refType=null, unstructuredReference=METERSKY ML, KALIL AC. Management of ventilator-associated pneumonia: Guidelines[J]. Clin Chest Med, 2018, 39(4): 797-808., articleTitle=Management of ventilator-associated pneumonia: Guidelines, refAbstract=null), Reference(id=1244321233576247971, tenantId=1146029695717560320, journalId=1244284848500682798, articleId=1244321222117409754, doi=null, pmid=null, pmcid=null, year=2020, volume=46, issue=6, pageStart=1170, pageEnd=1179, url=null, language=null, rfNumber=[6], rfOrder=5, authorNames=FERNANDO SM, TRAN A, CHENG W, journalName=Intensive Care Med, refType=null, unstructuredReference=FERNANDO SM, TRAN A, CHENG W, et al. Diagnosis of ventilator-associated pneumonia in critically ill adult patients—A systematic review and meta-analysis[J]. Intensive Care Med, 2020, 46(6): 1170-1179., articleTitle=Diagnosis of ventilator-associated pneumonia in critically ill adult patients—A systematic review and meta-analysis, refAbstract=null), Reference(id=1244321233748214451, tenantId=1146029695717560320, journalId=1244284848500682798, articleId=1244321222117409754, doi=null, pmid=null, pmcid=null, year=2020, volume=46, issue=5, pageStart=888, pageEnd=906, url=null, language=null, rfNumber=[7], rfOrder=6, authorNames=PAPAZIAN L, KLOMPAS M, LUYT C-E, journalName=Intensive Care Med, refType=null, unstructuredReference=PAPAZIAN L, KLOMPAS M, LUYT C-E. Ventilator-associated pneumonia in adults: A narrative review[J]. Intensive Care Med, 2020, 46(5): 888-906., articleTitle=Ventilator-associated pneumonia in adults: A narrative review, refAbstract=null), Reference(id=1244321233890820802, tenantId=1146029695717560320, journalId=1244284848500682798, articleId=1244321222117409754, doi=null, pmid=null, pmcid=null, year=2021, volume=null, issue=null, pageStart=null, pageEnd=null, url=null, language=null, rfNumber=[8], rfOrder=7, authorNames=严玉娇, journalName=null, refType=null, unstructuredReference=严玉娇. 成人危重症患者气道管理最佳证据的临床应用[D]. 武汉: 长江大学, 2021., articleTitle=成人危重症患者气道管理最佳证据的临床应用, refAbstract=null), Reference(id=1244321233995678418, tenantId=1146029695717560320, journalId=1244284848500682798, articleId=1244321222117409754, doi=null, pmid=null, pmcid=null, year=2010, volume=null, issue=null, pageStart=393, pageEnd=397, url=null, language=null, rfNumber=[9], rfOrder=8, authorNames=LI S, YI L, journalName=null, refType=null, unstructuredReference=LI S, YI L. Feature extraction of lung sounds based on bispectrum analysis[C]//Proceedings of the 2010 Third International Symposium on Information Processing. Qingdao: IEEE, 2010: 393-397., articleTitle=Feature extraction of lung sounds based on bispectrum analysis, refAbstract=null), Reference(id=1244321234197005026, tenantId=1146029695717560320, journalId=1244284848500682798, articleId=1244321222117409754, doi=null, pmid=null, pmcid=null, year=2016, volume=29, issue=1, pageStart=45, pageEnd=51, url=null, language=null, rfNumber=[10], rfOrder=9, authorNames=姚小静, 王洪, 刘盛雄, journalName=航天医学与医学工程, refType=null, unstructuredReference=姚小静, 王洪, 刘盛雄. 基于遗传BP神经网络的肺音分类识别算法研究[J]. 航天医学与医学工程, 2016, 29(1): 45-51., articleTitle=基于遗传BP神经网络的肺音分类识别算法研究, refAbstract=null), Reference(id=1244321234335417070, tenantId=1146029695717560320, journalId=1244284848500682798, articleId=1244321222117409754, doi=null, pmid=null, pmcid=null, year=2006, volume=null, issue=null, pageStart=385, pageEnd=388, url=null, language=null, rfNumber=[11], rfOrder=10, authorNames=LI Z, DU M, journalName=null, refType=null, unstructuredReference=LI Z, DU M. HHT based lung sound crackle detection and classification[C]//Proceedings of International Symposium on Intelligent Signal Processing & Communication Systems. Hong Kong: IEEE, 2006: 385-388., articleTitle=HHT based lung sound crackle detection and classification, refAbstract=null), Reference(id=1244321234465440503, tenantId=1146029695717560320, journalId=1244284848500682798, articleId=1244321222117409754, doi=null, pmid=null, pmcid=null, year=2012, volume=null, issue=29, pageStart=2693, pageEnd=2698, url=null, language=null, rfNumber=[12], rfOrder=11, authorNames=LU L, MA S, FENG S, journalName=Procedia Eng, refType=null, unstructuredReference=LU L, MA S, FENG S, et al. Envelope extraction of anaesthesia breathing sound signal on Hilbert huang transform[J]. Procedia Eng, 2012(29): 2693-2698., articleTitle=Envelope extraction of anaesthesia breathing sound signal on Hilbert huang transform, refAbstract=null), Reference(id=1244321234587075328, tenantId=1146029695717560320, journalId=1244284848500682798, articleId=1244321222117409754, doi=null, pmid=null, pmcid=null, year=2013, volume=null, issue=null, pageStart=1, pageEnd=4, url=null, language=null, rfNumber=[13], rfOrder=12, authorNames=ABBASI S, DERAKHSHANFAR R, ABBASI A, journalName=null, refType=null, unstructuredReference=ABBASI S, DERAKHSHANFAR R, ABBASI A, et al. Classification of normal and abnormal lung sounds using neural network and support vector machines[C]//Proceedings of the 2013 21st Iranian Conference on Electrical Engineering (ICEE). Mashhad: IEEE, 2013: 1-4., articleTitle=Classification of normal and abnormal lung sounds using neural network and support vector machines, refAbstract=null), Reference(id=1244321234721293070, tenantId=1146029695717560320, journalId=1244284848500682798, articleId=1244321222117409754, doi=null, pmid=null, pmcid=null, year=2014, volume=null, issue=null, pageStart=2138, pageEnd=2141, url=null, language=null, rfNumber=[14], rfOrder=13, authorNames=UYSAL S, UYSAL H, BOLAT B, journalName=null, refType=null, unstructuredReference=UYSAL S, UYSAL H, BOLAT B, et al. Classification of normal and abnormal lung sounds using wavelet coefficients[C]//Proceedings of the 2014 22nd Signal Processing and Communications Applications Conference (SIU). Trabzon: IEEE, 2014: 2138-2141., articleTitle=Classification of normal and abnormal lung sounds using wavelet coefficients, refAbstract=null), Reference(id=1244321234855510806, tenantId=1146029695717560320, journalId=1244284848500682798, articleId=1244321222117409754, doi=null, pmid=null, pmcid=null, year=2014, volume=null, issue=15, pageStart=223, pageEnd=null, url=null, language=null, rfNumber=[15], rfOrder=14, authorNames=PALANIAPPAN R, SUNDARAJ K, SUNDARAJ S, journalName=BMC Bioinformatics, refType=null, unstructuredReference=PALANIAPPAN R, SUNDARAJ K, SUNDARAJ S. A comparative study of the SVM and K-NN machine learning algorithms for the diagnosis of respiratory pathologies using pulmonary acoustic signals[J]. BMC Bioinformatics, 2014(15): 223., articleTitle=A comparative study of the SVM and K-NN machine learning algorithms for the diagnosis of respiratory pathologies using pulmonary acoustic signals, refAbstract=null), Reference(id=1244321234964562721, tenantId=1146029695717560320, journalId=1244284848500682798, articleId=1244321222117409754, doi=null, pmid=null, pmcid=null, year=2014, volume=35, issue=10, pageStart=5, pageEnd=null, url=null, language=null, rfNumber=[16], rfOrder=15, authorNames=刘国栋, 许静, journalName=通信学报, refType=null, unstructuredReference=刘国栋, 许静. 基于SVM方法的神经网络呼吸音识别算法[J]. 通信学报, 2014, 35(10): 5., articleTitle=基于SVM方法的神经网络呼吸音识别算法, refAbstract=null), Reference(id=1244321235119751975, tenantId=1146029695717560320, journalId=1244284848500682798, articleId=1244321222117409754, doi=null, pmid=null, pmcid=null, year=2017, volume=null, issue=null, pageStart=null, pageEnd=null, url=null, language=null, rfNumber=[17], rfOrder=16, authorNames=POREVA A, KARPLYUK Y, VAITYSHYN V, journalName=null, refType=null, unstructuredReference=POREVA A, KARPLYUK Y, VAITYSHYN V. Machine learning techniques application for lung diseases diagnosis[C]//Proceedings of 2017 5th IEEE Workshop on Advances in Information, Electronic and Electrical Engineering (AIEEE). Kyiv: IEEE, 2017: 1754210., articleTitle=Machine learning techniques application for lung diseases diagnosis, refAbstract=null), Reference(id=1244321235258164021, tenantId=1146029695717560320, journalId=1244284848500682798, articleId=1244321222117409754, doi=null, pmid=null, pmcid=null, year=2016, volume=null, issue=7, pageStart=3, pageEnd=null, url=null, language=null, rfNumber=[18], rfOrder=17, authorNames=张晓燕, journalName=电子测试, refType=null, unstructuredReference=张晓燕. 基于BP神经网络的肺音识别与诊断研究[J]. 电子测试, 2016(7): 3., articleTitle=基于BP神经网络的肺音识别与诊断研究, refAbstract=null), Reference(id=1244321235404964674, tenantId=1146029695717560320, journalId=1244284848500682798, articleId=1244321222117409754, doi=null, pmid=null, pmcid=null, year=2018, volume=null, issue=88, pageStart=58, pageEnd=69, url=null, language=null, rfNumber=[19], rfOrder=18, authorNames=BARDOU D, ZHANG K, AHMAD S, journalName=Artif Intell Med, refType=null, unstructuredReference=BARDOU D, ZHANG K, AHMAD S, et al. Lung sounds classification using convolutional neuralnetworks[J]. Artif Intell Med, 2018(88): 58-69., articleTitle=Lung sounds classification using convolutional neuralnetworks, refAbstract=null), Reference(id=1244321235509822279, tenantId=1146029695717560320, journalId=1244284848500682798, articleId=1244321222117409754, doi=null, pmid=null, pmcid=null, year=2016, volume=null, issue=null, pageStart=1, pageEnd=4, url=null, language=null, rfNumber=[20], rfOrder=19, authorNames=CHEN Q, ZHANG W, XIANG T, journalName=null, refType=null, unstructuredReference=CHEN Q, ZHANG W, XIANG T, et al. Automatic heart and lung sounds classification using convolutional neural networks[C]//Proceedings of Signal & Information Processing Association Summit & Conference (APSIPA). Jeju: IEEE, 2016: 1-4., articleTitle=Automatic heart and lung sounds classification using convolutional neural networks, refAbstract=null), Reference(id=1244321235614679890, tenantId=1146029695717560320, journalId=1244284848500682798, articleId=1244321222117409754, doi=null, pmid=null, pmcid=null, year=2016, volume=null, issue=null, pageStart=1, pageEnd=5, url=null, language=null, rfNumber=[21], rfOrder=20, authorNames=LIN L, XU W, HONG Q, journalName=null, refType=null, unstructuredReference=LIN L, XU W, HONG Q, et al. Classification between normal and adventitious lung sounds using deep neural network[C]//Proceedings of the 2016 10th International Symposium on Chinese Spoken Language Processing (ISCSLP). Tianjin: IEEE, 2016: 1-5., articleTitle=Classification between normal and adventitious lung sounds using deep neural network, refAbstract=null), Reference(id=1244321235727926104, tenantId=1146029695717560320, journalId=1244284848500682798, articleId=1244321222117409754, doi=null, pmid=null, pmcid=null, year=2017, volume=null, issue=null, pageStart=776, pageEnd=780, url=null, language=null, rfNumber=[22], rfOrder=21, authorNames=GEMMEKE JF, ELLIS DPW, FREEDMAN D, journalName=null, refType=null, unstructuredReference=GEMMEKE JF, ELLIS DPW, FREEDMAN D, et al. Audio set: An ontology and human-labeled dataset for audio events[C]//Proceedings of 2017-2017 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP). New Orleans: IEEE, 2017: 776-780., articleTitle=Audio set: An ontology and human-labeled dataset for audio events, refAbstract=null), Reference(id=1244321235824395101, tenantId=1146029695717560320, journalId=1244284848500682798, articleId=1244321222117409754, doi=null, pmid=null, pmcid=null, year=2022, volume=30, issue=4, pageStart=26, pageEnd=30, url=null, language=null, rfNumber=[23], rfOrder=22, authorNames=张志超, 李晓燕, journalName=电子设计工程, refType=null, unstructuredReference=张志超, 李晓燕. 基于VGGish网络的音频信息情感智能识别算法[J]. 电子设计工程, 2022, 30(4): 26-30., articleTitle=基于VGGish网络的音频信息情感智能识别算法, refAbstract=null), Reference(id=1244321235929252709, tenantId=1146029695717560320, journalId=1244284848500682798, articleId=1244321222117409754, doi=null, pmid=null, pmcid=null, year=2020, volume=null, issue=null, pageStart=null, pageEnd=null, url=null, language=null, rfNumber=[24], rfOrder=23, authorNames=廖辉强, journalName=null, refType=null, unstructuredReference=廖辉强. 基于语音帧自动标注和领域知识迁移的语音 情感识别研究[D]. 广州: 华南理工大学, 2020., articleTitle=基于语音帧自动标注和领域知识迁移的语音 情感识别研究, refAbstract=null), Reference(id=1244321236034110314, tenantId=1146029695717560320, journalId=1244284848500682798, articleId=1244321222117409754, doi=null, pmid=null, pmcid=null, year=2015, volume=null, issue=null, pageStart=1, pageEnd=14, url=null, language=null, rfNumber=[25], rfOrder=24, 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). Computational and Biological Learning Society, 2015: 1-14., articleTitle=Very deep convolutional networks for large-scale image recognition, refAbstract=null), Reference(id=1244321236122190706, tenantId=1146029695717560320, journalId=1244284848500682798, articleId=1244321222117409754, doi=null, pmid=null, pmcid=null, year=2023, volume=79, issue=1, pageStart=null, pageEnd=null, url=null, language=null, rfNumber=[26], rfOrder=25, authorNames=CINYOL F, BAYSAL U, KÖKSAL D, journalName=Biomed Signal Process Control, refType=null, unstructuredReference=CINYOL F, BAYSAL U, KÖKSAL D, et al. Incorporating support vector machine to the classification of respiratory sounds by convolutional neural network[J]. Biomed Signal Process Control, 2023, 79(1): 104093., articleTitle=Incorporating support vector machine to the classification of respiratory sounds by convolutional neural network, refAbstract=null), Reference(id=1244321236252214136, tenantId=1146029695717560320, journalId=1244284848500682798, articleId=1244321222117409754, doi=null, pmid=null, pmcid=null, year=2018, volume=26, issue=4, pageStart=888, pageEnd=900, url=null, language=null, rfNumber=[27], rfOrder=26, authorNames=孙艳丽, 杨娜, 张正涛, journalName=应用基础与工程科学学报, refType=null, unstructuredReference=孙艳丽, 杨娜, 张正涛, . 基于核主元分析和支持向量机的结构损伤识别研究[J]. 应用基础与工程科学学报, 2018, 26(4): 888-900., articleTitle=基于核主元分析和支持向量机的结构损伤识别研究, refAbstract=null), Reference(id=1244321236365460352, tenantId=1146029695717560320, journalId=1244284848500682798, articleId=1244321222117409754, doi=null, pmid=null, pmcid=null, year=1998, volume=null, issue=null, pageStart=null, pageEnd=null, url=null, language=null, rfNumber=[28], rfOrder=27, authorNames=VAPNIK VN, journalName=The nature of statistical learning theory, refType=null, unstructuredReference=VAPNIK VN. The nature of statistical learning theory[M]. New York: Wiley, 1998., articleTitle=null, refAbstract=null)], funds=[Fund(id=1244321231437152875, tenantId=1146029695717560320, journalId=1244284848500682798, articleId=1244321222117409754, awardId=2021YFC0122403, language=CN, fundingSource=国家重点研发计划项目(2021YFC0122403), fundOrder=null, country=null)], companyList=[AuthorCompany(id=1244321226521428238, tenantId=1146029695717560320, journalId=1244284848500682798, articleId=1244321222117409754, xref=null, ext=[AuthorCompanyExt(id=1244321226525622543, tenantId=1146029695717560320, journalId=1244284848500682798, articleId=1244321222117409754, companyId=1244321226521428238, language=EN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=School of Engineering Medicine; Beijing Advanced Innovation Center for Biomedical Engineering, Beihang University, Beijing 100191, China), AuthorCompanyExt(id=1244321226534011152, tenantId=1146029695717560320, journalId=1244284848500682798, articleId=1244321222117409754, companyId=1244321226521428238, language=CN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=北京航空航天大学 医学科学与工程学院;北京生物医学工程高精尖创新中心,北京 100191)])], figs=[ArticleFig(id=1244321230245970427, tenantId=1146029695717560320, journalId=1244284848500682798, articleId=1244321222117409754, language=EN, label=Fig. 1, caption=Time-domain waveforms and spectrograms of breath sounds during the noise removal process (a) Before high-pass filtering, (b) After high-pass filtering, (c) After wavelet soft-threshold filtering, figureFileSmall=fh8/bwXz01TuE2FTH1q0CQ==, figureFileBig=gvyOVvYFLKuBWvMEA0KHAw==, tableContent=null), ArticleFig(id=1244321230359216643, tenantId=1146029695717560320, journalId=1244284848500682798, articleId=1244321222117409754, language=CN, label=图1, caption=呼吸音去噪过程中时域波形图和频谱图, figureFileSmall=fh8/bwXz01TuE2FTH1q0CQ==, figureFileBig=gvyOVvYFLKuBWvMEA0KHAw==, tableContent=null), ArticleFig(id=1244321230648623645, tenantId=1146029695717560320, journalId=1244284848500682798, articleId=1244321222117409754, language=EN, label=Tab. 1, caption=

Structure and network layer parameters of VGGish network

, figureFileSmall=null, figureFileBig=null, tableContent=
网络层类型参数
全连接层(输出层)特征图:128
全连接层特征图:4 096
全连接层特征图:4 096
最大池化层池化核:2×2
卷积层特征图:512 卷积核:3×3
卷积层特征图:512 卷积核:3×3
最大池化层池化核:2×2
卷积层特征图:256 卷积核:3×3
卷积层特征图:256 卷积核:3×3
最大池化层池化核:2×2
卷积层特征图:128 卷积核:3×3
最大池化层池化核:2×2
卷积层特征图:64 卷积核:3×3
输入对数梅尔频谱图96×64
), ArticleFig(id=1244321230795424298, tenantId=1146029695717560320, journalId=1244284848500682798, articleId=1244321222117409754, language=CN, label=表1, caption=

VGGish网络的结构与网络层参数

, figureFileSmall=null, figureFileBig=null, tableContent=
网络层类型参数
全连接层(输出层)特征图:128
全连接层特征图:4 096
全连接层特征图:4 096
最大池化层池化核:2×2
卷积层特征图:512 卷积核:3×3
卷积层特征图:512 卷积核:3×3
最大池化层池化核:2×2
卷积层特征图:256 卷积核:3×3
卷积层特征图:256 卷积核:3×3
最大池化层池化核:2×2
卷积层特征图:128 卷积核:3×3
最大池化层池化核:2×2
卷积层特征图:64 卷积核:3×3
输入对数梅尔频谱图96×64
), ArticleFig(id=1244321230900281913, tenantId=1146029695717560320, journalId=1244284848500682798, articleId=1244321222117409754, language=EN, label=Tab. 2, caption=

Classification result-evaluation index

, figureFileSmall=null, figureFileBig=null, tableContent=
评价指标需要吸痰无需吸痰总体准确率/%
精确率/%84.32±9.5295.50±2.6690.96±3.08
召回率/%93.10±3.7490.3±4.91
F1分数88.49±5.5192.83±2.88
), ArticleFig(id=1244321231034499651, tenantId=1146029695717560320, journalId=1244284848500682798, articleId=1244321222117409754, language=CN, label=表2, caption=

分类结果-评价指标

, figureFileSmall=null, figureFileBig=null, tableContent=
评价指标需要吸痰无需吸痰总体准确率/%
精确率/%84.32±9.5295.50±2.6690.96±3.08
召回率/%93.10±3.7490.3±4.91
F1分数88.49±5.5192.83±2.88
), ArticleFig(id=1244321231130968652, tenantId=1146029695717560320, journalId=1244284848500682798, articleId=1244321222117409754, language=EN, label=Tab. 3, caption=

Classification result-confusion matrix

, figureFileSmall=null, figureFileBig=null, tableContent=
真实类别预测类别合计
需要吸痰无需吸痰
需要吸痰26820288
无需吸痰41380421
合计309400709
), ArticleFig(id=1244321231269380696, tenantId=1146029695717560320, journalId=1244284848500682798, articleId=1244321222117409754, language=CN, label=表3, caption=

分类结果-混淆矩阵

, figureFileSmall=null, figureFileBig=null, tableContent=
真实类别预测类别合计
需要吸痰无需吸痰
需要吸痰26820288
无需吸痰41380421
合计309400709
)], attaches=null, journal=Journal(id=1244283803934437421, delFlag=0, nameCn=医用生物力学, nameEn=Journal of Medical Biomechanics, nameHistory1=null, nameHistory2=null, issn=1004-7220, eissn=null, cn=31-1624R, coden=null, periodic=1, 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=G8Fk4zRvuEegNfdcEgpIYA==, journalPrice=null, startedYear=null, abbrevIsoEn=Journal of Medical Biomechanics, journalRemark=null, publicationField=null, createdTime=1774589976534, updatedTime=1774590784682, createdBy=18614031015, updatedBy=13701087609, firstLetterCn=J, firstLetterEn=J, subjectCode=Engineering, subjectName=null, subjectCodeEn=Engineering, subjectNameEn=null, picCn=G8Fk4zRvuEegNfdcEgpIYA==, picEn=Eh34RQA3oTtmzcSmNDdO9Q==, jcr=null, cjcr=null, exts=[JournalExt(id=1244287193645167332, language=CN, name=医用生物力学, 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=1774590784703, updatedTime=1774590784703, createdBy=13701087609, updatedBy=13701087609, submissionGuidelinesUrl=, submissionAuthorUrl=http://www.mechanobiology.cn/yyswlx/author/login, submissionEditorUrl=http://www.mechanobiology.cn/yyswlx/editor/login, submissionReviewUrl=http://www.mechanobiology.cn/yyswlx/reviewer/login, submissionCeEditorUrl=, submissionAeEditorUrl=, option={"copyright":""}), JournalExt(id=1244287193737442021, language=EN, name=Journal of Medical Biomechanics, 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=1774590784725, updatedTime=1774590784725, createdBy=13701087609, updatedBy=13701087609, submissionGuidelinesUrl=, submissionAuthorUrl=http://www.mechanobiology.cn/yyswlx/author/login, submissionEditorUrl=http://www.mechanobiology.cn/yyswlx/editor/login, submissionReviewUrl=http://www.mechanobiology.cn/yyswlx/reviewer/login, submissionCeEditorUrl=, submissionAeEditorUrl=, option={"copyright":""})], databaseList=null, tenantJournalId=1244284848500682798, websiteList=[Website(id=1244289124061000646, webName=null, webTitle=null, webDomain=null, webCopyrigh=null, webIpcNo=null, seoTitle=null, seoKeywords=null, seoDescription=null, tenantJournalId=null, journalId=1244284848500682798, 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/yyswlx/CN, language=CN, createTime=1774591244950, createBy=18614031015, updateTime=1774591273991, updateBy=18614031015, name=医用生物力学-中文, tplId=1146099689490845704, title=医用生物力学, delFlag=0, indexPage=/home, props=[WebsiteProps(id=1244289419298063322, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1244289124061000646, code=articleTextType, value=kx, createTime=1774591315340, updateTime=1774591315340, creator=18614031015, updator=18614031015), WebsiteProps(id=1244289419268703191, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1244289124061000646, code=banner, value=null, createTime=1774591315333, updateTime=1774591315333, creator=18614031015, updator=18614031015), WebsiteProps(id=1244289419340006365, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1244289124061000646, code=grayFlag, value=0, createTime=1774591315350, updateTime=1774591315350, creator=18614031015, updator=18614031015), WebsiteProps(id=1244289419260314582, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1244289124061000646, code=logo, value=https://castjournals.cast.org.cn/joweb/yyswlx/CN/file/pic?fileId=pmnHPc+Tur67OsVt9JPkuA==, createTime=1774591315331, updateTime=1774591315331, creator=18614031015, updator=18614031015), WebsiteProps(id=1244289419440669663, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1244289124061000646, code=minRunFlag, value=0, createTime=1774591315374, updateTime=1774591315374, creator=18614031015, updator=18614031015), WebsiteProps(id=1244289419285480409, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1244289124061000646, code=picServerUrl, value=https://castjournals.cast.org.cn/joweb/yyswlx/CN/file/pic, createTime=1774591315337, updateTime=1774591315337, creator=18614031015, updator=18614031015), WebsiteProps(id=1244289419428086750, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1244289124061000646, code=silenceFlag, value=0, createTime=1774591315371, updateTime=1774591315371, creator=18614031015, updator=18614031015), WebsiteProps(id=1244289419277091800, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1244289124061000646, code=staticResourcePath, value=https://castjournals.cast.org.cn/joweb/cast_kjdb_cn_619/, createTime=1774591315335, updateTime=1774591315335, creator=18614031015, updator=18614031015), WebsiteProps(id=1244289419310646235, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1244289124061000646, code=themeColor, value=null, createTime=1774591315343, updateTime=1774591315343, creator=18614031015, updator=18614031015), WebsiteProps(id=1244289419327423452, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1244289124061000646, code=themeStyle, value=null, createTime=1774591315347, updateTime=1774591315347, creator=18614031015, updator=18614031015)]), Website(id=1244289124119720904, webName=null, webTitle=null, webDomain=null, webCopyrigh=null, webIpcNo=null, seoTitle=null, seoKeywords=null, seoDescription=null, tenantJournalId=null, journalId=1244284848500682798, 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/yyswlx/EN, language=EN, createTime=1774591244964, createBy=18614031015, updateTime=1774591288789, updateBy=18614031015, name=医用生物力学-英文, tplId=1146101810881728533, title=Journal of Medical Biomechanics, delFlag=0, indexPage=/home, props=[WebsiteProps(id=1244289447987098584, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1244289124119720904, code=articleTextType, value=kx, createTime=1774591322180, updateTime=1774591322180, creator=18614031015, updator=18614031015), WebsiteProps(id=1244289447966127061, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1244289124119720904, code=banner, value=null, createTime=1774591322175, updateTime=1774591322175, creator=18614031015, updator=18614031015), WebsiteProps(id=1244289448003875803, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1244289124119720904, code=grayFlag, value=0, createTime=1774591322184, updateTime=1774591322184, creator=18614031015, updator=18614031015), WebsiteProps(id=1244289447957738452, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1244289124119720904, code=logo, value=https://castjournals.cast.org.cn/joweb/yyswlx/EN/file/pic?fileId=pmnHPc+Tur67OsVt9JPkuA==, createTime=1774591322173, updateTime=1774591322173, creator=18614031015, updator=18614031015), WebsiteProps(id=1244289448020653021, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1244289124119720904, code=minRunFlag, value=0, createTime=1774591322188, updateTime=1774591322188, creator=18614031015, updator=18614031015), WebsiteProps(id=1244289447978709975, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1244289124119720904, code=picServerUrl, value=https://castjournals.cast.org.cn/joweb/yyswlx/EN/file/pic, createTime=1774591322178, updateTime=1774591322178, creator=18614031015, updator=18614031015), WebsiteProps(id=1244289448012264412, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1244289124119720904, code=silenceFlag, value=0, createTime=1774591322186, updateTime=1774591322186, creator=18614031015, updator=18614031015), WebsiteProps(id=1244289447974515670, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1244289124119720904, code=staticResourcePath, value=https://castjournals.cast.org.cn/joweb/cast_kjdb_en_623/, createTime=1774591322177, updateTime=1774591322177, creator=18614031015, updator=18614031015), WebsiteProps(id=1244289447991292889, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1244289124119720904, code=themeColor, value=null, createTime=1774591322181, updateTime=1774591322181, creator=18614031015, updator=18614031015), WebsiteProps(id=1244289447999681498, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1244289124119720904, code=themeStyle, value=null, createTime=1774591322183, updateTime=1774591322183, creator=18614031015, updator=18614031015)])], journalTitle=医用生物力学, weixinUrl=null, journalUrl=http://www.mechanobiology.cn/, iacademicId=null, status=1, seqNo=null, journalTitleEn=Journal of Medical Biomechanics, journalPhotoCn=G8Fk4zRvuEegNfdcEgpIYA==, journalPhotoEn=Eh34RQA3oTtmzcSmNDdO9Q==, journalFirstLetter=J, 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/yyswlx/CN/10.16156/j.1004-7220.2025.05.031, detailUrlEn=https://castjournals.cast.org.cn/joweb/yyswlx/EN/10.16156/j.1004-7220.2025.05.031, pdfUrlCn=https://castjournals.cast.org.cn/joweb/yyswlx/CN/PDF/10.16156/j.1004-7220.2025.05.031, pdfUrlEn=https://castjournals.cast.org.cn/joweb/yyswlx/EN/PDF/10.16156/j.1004-7220.2025.05.031, aliStartDate=null, aliEndDate=null, collectionFlag=false, citedCount=null, citedUrl=null, reference=null)
收藏切换
基于迁移学习与呼吸音特征识别的机械通气吸痰时机判别
收藏切换
PDF下载
汪帅 , 郭江真 * , 陶春静 *
医用生物力学 | 论著 2025,40(5): 1318-1324
收起
收藏切换
医用生物力学 | 论著 2025, 40(5): 1318-1324
基于迁移学习与呼吸音特征识别的机械通气吸痰时机判别
全屏
汪帅, 郭江真* , 陶春静*
作者信息
  • 北京航空航天大学 医学科学与工程学院;北京生物医学工程高精尖创新中心,北京 100191

通讯作者:

郭江真,副教授,E-mail:
陶春静,教授,E-mail:

*

为共同通信作者
Determination of Sputum Suction Timing in Mechanical Ventilation Based on Transfer Learning and Breath Sounds Recognition
Shuai WANG, Jiangzhen GUO , Chunjing TAO
Affiliations
  • School of Engineering Medicine; Beijing Advanced Innovation Center for Biomedical Engineering, Beihang University, Beijing 100191, China
出版时间: 2025-10-01 doi: 10.16156/j.1004-7220.2025.05.031
文章导航
收藏切换
目的

提出基于迁移学习的呼吸音特征识别与吸痰时机自主判断方法。

方法

使用电子听诊器采集临床机械通气患者在吸痰前、后的主气道呼吸音,将吸痰前的呼吸音标记为“需要吸痰”类别。对采集数据进行高通滤波和小波软阈值滤波去噪,提取对数梅尔频谱图,并利用在Audio Set数据集上预训练的VGGish模型提取对数梅尔频谱图的特征向量,最后通过支持向量机分类器对特征向量进行分类,判断是否需要吸痰。

结果

对需要吸痰的呼吸音识别的精确率、召回率和F1分数分别为86.73%、93.06%和89.78%。

结论

基于迁移学习的呼吸音识别方法能够有效判断吸痰时机,在临床应用中具有潜在价值。

机械通气  /  呼吸音  /  迁移学习  /  吸痰时机判断
Objective

To propose a transfer learning-based method for breath sound feature recognition and autonomous determination of sputum suction timing.

Methods

An electronic stethoscope was used to collect breath sounds from the main airways of clinically ventilated patients before and after sputum suction, with pre-suction breath sounds labeled as requiring suction. The collected data underwent high-pass filtering and wavelet soft-threshold denoising, followed by the extraction of log-Mel spectrograms. A VGGish model pretrained on the Audio Set dataset was then employed to extract feature vectors from these spectrograms, which were subsequently classified using a support vector machine to determine whether suction was required.

Results

The precision, recall and F1 score for recognition of breath sounds requiring sputum suction were 86.73%, 93.06% and 89.78%, respectively.

Conclusions

The proposed breath sound recognition method based on transfer learning effectively determines the timing of sputum suction and shows a significant clinical potential.

mechanical ventilation  /  breath sound  /  transfer learning  /  determination of sputum suction time
汪帅, 郭江真, 陶春静. 基于迁移学习与呼吸音特征识别的机械通气吸痰时机判别. 医用生物力学, 2025 , 40 (5) : 1318 -1324 . DOI: 10.16156/j.1004-7220.2025.05.031
Shuai WANG, Jiangzhen GUO, Chunjing TAO. Determination of Sputum Suction Timing in Mechanical Ventilation Based on Transfer Learning and Breath Sounds Recognition[J]. Journal of Medical Biomechanics, 2025 , 40 (5) : 1318 -1324 . DOI: 10.16156/j.1004-7220.2025.05.031
研究表明,接近50%的急性呼吸衰竭患者需使用呼吸机进行机械通气[1-2]。机械通气时通常需要通过气管插管等方法建立人工气道,然而人工气道的存在会使咳嗽反射减弱,影响上气道的加温和湿化功能,降低分泌物排出能力,从而引起痰液淤积,严重时阻塞气道甚至危及生命[3-4]。研究表明,实施有效吸痰有助于预防和控制呼吸机相关性肺损伤等并发症发生,缩短机械通气时间和住院时长,从而减少住院费用[5-7]。目前,临床对于吸痰需求的评估主要依赖医生的听诊和主观经验判断,难以实现痰液淤积程度的连续监测和客观评估,不利于准确把握吸痰时机从而影响吸痰效果[8]。智能识别机械通气患者的吸痰需求,不仅能提升医疗质量,减轻医护人员负担,还可为呼吸机的智能化升级提供技术支持。
通常医生对患者吸痰需求的判断主要通过主气道等部位的呼吸音听诊,说明呼吸音带有吸痰需求相关的特征。目前大多数研究集中在肺部呼吸音的特征提取和异常肺音(哮鸣音、爆裂音)的识别上,很少对主气道部位的呼吸音进行研究。肺部呼吸音的特征提取方法主要包括双谱特征[9]、韦尔奇功率谱特征[10]、希尔伯特黄变换的频域特征[11-12]、小波分析[13-14]、梅尔频率倒谱系数[15]。肺部呼吸音识别的早期研究多采用传统的机器学习算法,如K近邻算法、支持向量机(support vector machine,SVM)[16]、逻辑回归[17]和BP神经网络[10,18]。随着深度学习特征提取能力的发展,卷积神经网络[19-20]和循环神经网络[21]等深度学习方法开始应用于肺部呼吸音识别领域。
在音频信号处理领域,迁移学习已成为解决数据量不足的重要方法。迁移学习的核心思想是通过将一个领域(通常是大数据集)上学习到的知识迁移到另一个领域(通常是小数据集),从而提高模型在小样本数据上的表现。迁移学习通过复用在大规模数据集上训练得到的深度学习模型特征提取能力,有效缓解数据不足导致的过拟合问题。
本文选择Google公司于2017年基于Audio Set大型音频数据集[22]训练的VGGish模型作为预训练的特征提取器。该模型基于深度卷积神经网络架构,通过在大规模音频数据集上的训练,能够有效提取音频信号的高层次声学特征。VGGish模型通过其预训练获得的128维声学特征表示,能够有效捕获呼吸音的时频特性并抑制无关噪声成分。已有研究表明,该模型在小型音频数据集的迁移学习场景中展现出卓越性能[23-24],这为本研究的特征提取提供了可靠的理论和技术支持。
针对主气道部位呼吸音临床数据采集困难、样本量少的问题,本文采用迁移学习方法进行特征提取。首先,使用电子听诊器收集患者吸痰前(需要吸痰)和吸痰后(无需吸痰)主气道部位的呼吸音,并使用高通滤波和小波滤波去除呼吸音中的环境声和心音等噪声,然后提取呼吸音的对数梅尔频谱图,利用在Audio Set数据集上训练好的VGGish模型提取呼吸音对数梅尔频谱图上的高级声学特征,最终使用SVM对高级声学特征进行分类。
使用Littman 3200型电子听诊器(3M公司,美国)采集和传输呼吸音信号。该设备具有专业的环境噪声消除技术,可以降低85%的背景噪声;同时配备蓝牙技术,可以满足数据传输和收集的要求。
实验所用的呼吸音信号由北京大学第三医院呼吸与危重症医学科临床医生在重症监护病房采集,采集对象为有排痰需求的机械通气患者。本研究通过了北京航空航天大学生物与医学伦理委员会审查(BM20230118)。临床医生分别在患者排痰前后在主气道位置,即胸骨隆突处,使用电子听诊器收集呼吸音信号。在信号收集的过程中,经过临床医生判断有吸痰需求之后,分别收集吸痰前、后的呼吸音作为需要吸痰和无需吸痰的呼吸音表征。由于1个呼吸周期的时长为3~5 s,采集的呼吸音应至少包括1个完整的呼吸周期,从而包括完整的呼吸音特征,而每次采集的起始时刻位于呼吸周期中的不同时刻,故采集时长应最短应为10 s,每次采集的呼吸音时长10~30 s。
在采集到的呼吸音信号中,除了心音信号,还会不可避免地混入环境、设备噪声以及人体内部其他器官的噪声,这些噪声主要集中在低频段。考虑到呼吸音信号的频带为100~2 000 Hz,将频率低于100 Hz的信号视为噪声信号。因此,为了有效去除这些低频噪声,采用四阶巴特沃斯高通滤波器(截止频率100 Hz)进行滤波处理。
经过高通滤波之后,呼吸音中的低频噪声已被去除,但由于部分心音成分与呼吸音信号频带重叠,仍需要进一步处理以提取纯净的呼吸音信号。为此,采用小波软阈值法去除心音成分,从而得到纯净的肺音。在小波分解的过程中,选择coif小波基对呼吸音信号进行5层小波分解,得到d1~d5共5层小波分解的细节系数和第5层小波分解的近似系数a5。然后对所有细节系数施加软阈值去噪处理,最终通过小波重构得到去噪后的信号。
数据增强是指通过对原始数据进行信号处理或时频变换,生成具有相同信息的衍生样本。数据增强的作用是增加用于训练的样本数量,从而增强模型的泛化能力,同时也能有效地缓解过拟合。本研究采用在不同时间点对原始呼吸音进行切片的方法,从有限长度的音频中获得更多样本。
由于VGGish模型的输入为对数梅尔频谱图,故需要先提取出呼吸音的对数梅尔频谱图。为了和标准的VGGish模型输入对齐,提取对数梅尔频谱图分为以下4步:
(1)使用16 kHz对预处理后的呼吸音数据进行重采样。
(2)以25 ms帧长和10 ms帧移对重采样后的呼吸音数据进行短时傅里叶变换得到频谱图。
(3)将上一步得到的频谱图经过64阶的梅尔滤波器组计算得到梅尔频谱,再取对数得到对数梅尔频谱。
(4)以0.96 s帧长和0.96 s帧移重组帧,相邻两帧之间没有重叠区域,每一帧包含64个梅尔频带。对于时长为N s的呼吸音频,对应的对数梅尔频谱图为[N/0.96]×96×64,[N/0.96]代表向上取整数。
VGGish模型是专门针对音频特征提取而设计的深度神经网络[25]。该模型首先在大规模的Audio Set数据集[22]上进行预训练,进而能够学习并提取通用的音频表示。Audio Set数据集由Google公司于2017年开源,包含约200万条人工标注的音频剪辑片段,涵盖600余类声音类型,包括人声、动物叫声、乐器声以及各种日常环境声音等。基于该数据集训练得到的VGGish模型具有良好的通用性和泛化能力,尤其适合在呼吸音数据集上进行迁移学习。在网络结构方面,VGGish模型以VGG16为基础,共包含6个卷积层、3个全连接层以及4个最大池化层。结果表明,当输入的对数梅尔频谱图维度为N×96×64时,输出特征向量维度为N×128(见表1)。
SVM对小样本、非线性和高维度的问题效果较好,广泛应用于处理分类和回归任务[26-27]。SVM通过求解对偶优化问题实现分类[28],其目标函数为:
约束条件为:
式中:αi为拉格朗日乘子;Kxixj)为核函数;xixj为样本特征向量。常用的核函数主要包括线性核函数、多项式核函数和高斯核函数。由于高斯核函数具有较强的非线性映射能力,目前应用最为广泛,本文采用高斯核函数对原始特征进行映射:
式中:δ为代表训练样本在特征空间中作用范围的核函数参数,一般需要调参确定。
实验数据集包含16例机械通气患者的29组吸痰前/后呼吸音样本。这些数据的平均时长为17.0 s,总时长为492.8 s,采样频率为22.05 kHz,音频格式为wav文件。
为了保证输入信号维度的一致性,在呼吸音信号输入模型之前还需要对其进行切片处理。考虑到切片时长需要包括至少1个完整呼吸音周期,人体呼吸周期为3~5 s,每个切片的起始时刻可能位于呼吸周期中的不同位置,将切片时长定为10 s。在切片的同时对呼吸音进行数据增强,将切片间隔定为0.2 s,最终得到709个呼吸音切片,包括288个“需要吸痰”的呼吸音切片和421个“无需吸痰”的呼吸音切片。
针对呼吸音二分类问题的评价指标有召回率(Recall)、精确率(Precision)、准确率(Accuracy)和F1分数,相关公式计算如下:
式中:TP表示样本的真实值为正类,模型也识别为正类的样本数量;FP表示样本的真实值为负类,模型识别为正类的样本数量;TN表示样本的真实值为负类,模型也识别为负类的样本数量;FN表示样本的真实值为正类,模型识别为负类的样本数量。召回率代表对真实值为正类的样本识别情况,精确率代表识别为正类的样本中真实值也为正类的比例,准确率代表全部样本中识别正确样本所占的比例,F1分数是精确率和召回率的调和平均值,能够反映二者的总体状态。
先后使用高通滤波和小波滤波对原始的呼吸音信号进行预处理。结果显示,大部分心音等低频噪声在高通滤波之后得到有效滤除。高通滤波前100 Hz以下的低频区有强烈的噪声信号,可能对最后呼吸音信号的特征提取造成干扰;而在高通滤波之后,100 Hz以下的低频区信号强度明显减弱。同时,300 Hz以下的低频区域仍然有部分环境音、心音及其他低频噪声和呼吸音信号交叠在一起,500 Hz以上的高频区域则混杂了监护病房内的背景噪声。这些噪声有可能会影响呼吸音特征提取的质量,故需要进一步利用小波方法去除(见图1)。
使用网格搜索法确定SVM的惩罚因子C和高斯核函数参数γ。其中,惩罚因子C控制模型对分类误差的容忍度(平衡分类准确率与模型复杂度),高斯核参数γ决定单个样本对决策边界的影响范围(控制数据映射到高维空间后的分布特性)。设置C的搜索范围为[1,10],搜索步长为0.5;设置γ的搜索范围为[0,0.05],搜索步长为0.01;使用十折交叉验证法下模型的分类准确率作为评价方法,最终确定惩罚因子C=5.0,高斯核参数γ=0.01。
使用上述参数的SVM对呼吸音数据集进行分类,并使用十折交叉法计算评价指标。实验结果显示,二分类任务十折交叉验证的平均准确率为90.96%,对“需要吸痰”类别样本的平均召回率为93.10%,预测为“需要吸痰”类别的平均精确率为84.32%(见表2)。
在十折交叉验证的过程中,每个样本都会被当做测试集进行一次实验。通过统计分析每次实验的预测结果,得到了整体的混淆矩阵(见表3)。基于混淆矩阵,进一步计算和分析模型的各项性能参数,以全面评估其分类能力。实验结果表明,二分类任务的整体准确率为91.40%,对“需要吸痰”呼吸音识别的精确率、召回率和F1分数分别为86.73%、93.06%和89.78%。
本文对呼吸音数据进行分类,实验结果表明,在“需要吸痰”的样本中,模型的精确率为86.73%,召回率为93.06%,F1分数为89.78%。尽管模型能够较为准确地识别出需要吸痰的样本(较高的召回率),但在分类时仍有部分“无需吸痰”的样本被误判为“需要吸痰”,导致精确率有所下降。这种误分类的原因可能与数据集的不平衡(即“需要吸痰”的样本数量较少)、个体差异以及信号噪声干扰等因素有关。模型的总体准确率为91.40%,表明该模型在二分类任务中表现良好,能够有效区分需要吸痰和无需吸痰的呼吸音样本,具有较高的临床实用价值。与现有文献中报道的呼吸音研究不同,本研究的创新之处在于将呼吸音特征用于评估机械通气患者的痰液淤积情况,这一领域的研究尚处于起步阶段。以往的研究主要集中在异常呼吸音的识别上,而机械通气患者的吸痰需求判断仍依赖于人工听诊,存在较大的主观性,并且缺乏连续监测和客观量化的手段。因此,本文提出的基于迁移学习的呼吸音识别方法,能够从客观、自动化的角度为医生提供有力的辅助决策支持,尤其是在多任务场景中,对于提高患者的护理质量具有重要意义。本文所采用的迁移学习方法在提升模型分类性能方面发挥了关键作用。VGGish模型在Audio Set数据集上进行预训练,具有较强的音频特征提取能力,并能够有效地迁移到机械通气患者呼吸音数据的分类任务中。通过提取对数梅尔频谱图的特征向量,并结合SVM分类器的高效性,模型在数据处理上的优势得以充分体现。此外,本文采用的高通滤波和小波软阈值滤波去噪策略,有效减少了背景噪声的干扰,提高了呼吸音信号的识别精度。
尽管本研究取得了较好的分类效果,但仍存在一些不足。例如,部分误判可能与数据样本的不平衡有关,未来的研究可以尝试通过扩充数据集或采用更复杂的分类模型来进一步优化识别精度。此外,模型的泛化能力仍需进一步验证,特别是在不同医院、不同呼吸音采集设备条件下的数据表现。本研究的主要贡献在于为机械通气患者的排痰需求提供了一种新的可持续的评估方式,并为呼吸机的智能化发展提供了初步的理论和技术支持。
本文提出了一种基于呼吸音信号的机械通气患者吸痰需求识别方法。该方法通过对呼吸音信号进行高通滤波和小波软阈值滤波去噪,并提取对数梅尔频谱图,利用VGGish预训练模型提取特征向量,结合SVM分类器进行吸痰需求的分类。实验结果表明,模型对“需要吸痰”呼吸音识别的整体精确率、召回率和F1分数分别为86.73%、93.06%和89.78%,验证了该方法的有效性。未来的工作将集中在收集更多的临床数据,以进一步优化和迭代模型,增强其泛化能力。同时,计划将该模型部署到呼吸机中,推动呼吸机的智能化发展,提升机械通气患者临床护理的质量和效率。
  • 国家重点研发计划项目(2021YFC0122403)
参考文献 引证文献
排序方式:
[1]
STEFAN MS, SHIEH MS, PEKOW PS, et al. Epidemiology and outcomes of acute respiratory failure in the United States, 2001 to 2009: A national survey[J]. J Hosp Med, 2013, 8(2): 76-82.
[2]
BEZZI M. Evolution of mortality over time in patients receiving mechanical ventilation[J]. Am J Respir Crit Care Med, 2013, 188(2): 220-230.
[3]
FISK AC. The effects of endotracheal suctioning in the pediatric population: An integrative review[J]. Dimens Crit Care Nurs, 2018, 37(1): 44-56.
[4]
GARDNER D, SHIRLAND L. Evidence-based guideline for suctioning the intubated neonate and infant[J]. Neonatal Netw, 2009, 28(5): 281-302.
[5]
METERSKY ML, KALIL AC. Management of ventilator-associated pneumonia: Guidelines[J]. Clin Chest Med, 2018, 39(4): 797-808.
[6]
FERNANDO SM, TRAN A, CHENG W, et al. Diagnosis of ventilator-associated pneumonia in critically ill adult patients—A systematic review and meta-analysis[J]. Intensive Care Med, 2020, 46(6): 1170-1179.
[7]
PAPAZIAN L, KLOMPAS M, LUYT C-E. Ventilator-associated pneumonia in adults: A narrative review[J]. Intensive Care Med, 2020, 46(5): 888-906.
[8]
严玉娇. 成人危重症患者气道管理最佳证据的临床应用[D]. 武汉: 长江大学, 2021.
[9]
LI S, YI L. Feature extraction of lung sounds based on bispectrum analysis[C]//Proceedings of the 2010 Third International Symposium on Information Processing. Qingdao: IEEE, 2010: 393-397.
[10]
姚小静, 王洪, 刘盛雄. 基于遗传BP神经网络的肺音分类识别算法研究[J]. 航天医学与医学工程, 2016, 29(1): 45-51.
[11]
LI Z, DU M. HHT based lung sound crackle detection and classification[C]//Proceedings of International Symposium on Intelligent Signal Processing & Communication Systems. Hong Kong: IEEE, 2006: 385-388.
[12]
LU L, MA S, FENG S, et al. Envelope extraction of anaesthesia breathing sound signal on Hilbert huang transform[J]. Procedia Eng, 2012(29): 2693-2698.
[13]
ABBASI S, DERAKHSHANFAR R, ABBASI A, et al. Classification of normal and abnormal lung sounds using neural network and support vector machines[C]//Proceedings of the 2013 21st Iranian Conference on Electrical Engineering (ICEE). Mashhad: IEEE, 2013: 1-4.
[14]
UYSAL S, UYSAL H, BOLAT B, et al. Classification of normal and abnormal lung sounds using wavelet coefficients[C]//Proceedings of the 2014 22nd Signal Processing and Communications Applications Conference (SIU). Trabzon: IEEE, 2014: 2138-2141.
[15]
PALANIAPPAN R, SUNDARAJ K, SUNDARAJ S. A comparative study of the SVM and K-NN machine learning algorithms for the diagnosis of respiratory pathologies using pulmonary acoustic signals[J]. BMC Bioinformatics, 2014(15): 223.
[16]
刘国栋, 许静. 基于SVM方法的神经网络呼吸音识别算法[J]. 通信学报, 2014, 35(10): 5.
[17]
POREVA A, KARPLYUK Y, VAITYSHYN V. Machine learning techniques application for lung diseases diagnosis[C]//Proceedings of 2017 5th IEEE Workshop on Advances in Information, Electronic and Electrical Engineering (AIEEE). Kyiv: IEEE, 2017: 1754210.
[18]
张晓燕. 基于BP神经网络的肺音识别与诊断研究[J]. 电子测试, 2016(7): 3.
[19]
BARDOU D, ZHANG K, AHMAD S, et al. Lung sounds classification using convolutional neuralnetworks[J]. Artif Intell Med, 2018(88): 58-69.
[20]
CHEN Q, ZHANG W, XIANG T, et al. Automatic heart and lung sounds classification using convolutional neural networks[C]//Proceedings of Signal & Information Processing Association Summit & Conference (APSIPA). Jeju: IEEE, 2016: 1-4.
[21]
LIN L, XU W, HONG Q, et al. Classification between normal and adventitious lung sounds using deep neural network[C]//Proceedings of the 2016 10th International Symposium on Chinese Spoken Language Processing (ISCSLP). Tianjin: IEEE, 2016: 1-5.
[22]
GEMMEKE JF, ELLIS DPW, FREEDMAN D, et al. Audio set: An ontology and human-labeled dataset for audio events[C]//Proceedings of 2017-2017 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP). New Orleans: IEEE, 2017: 776-780.
[23]
张志超, 李晓燕. 基于VGGish网络的音频信息情感智能识别算法[J]. 电子设计工程, 2022, 30(4): 26-30.
[24]
廖辉强. 基于语音帧自动标注和领域知识迁移的语音 情感识别研究[D]. 广州: 华南理工大学, 2020.
[25]
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). Computational and Biological Learning Society, 2015: 1-14.
[26]
CINYOL F, BAYSAL U, KÖKSAL D, et al. Incorporating support vector machine to the classification of respiratory sounds by convolutional neural network[J]. Biomed Signal Process Control, 2023, 79(1): 104093.
[27]
孙艳丽, 杨娜, 张正涛, . 基于核主元分析和支持向量机的结构损伤识别研究[J]. 应用基础与工程科学学报, 2018, 26(4): 888-900.
[28]
VAPNIK VN. The nature of statistical learning theory[M]. New York: Wiley, 1998.
2025年第40卷第5期
PDF下载
69
33
引用本文
BibTeX
文章信息
doi: 10.16156/j.1004-7220.2025.05.031
  • 接收时间:2025-01-07
  • 首发时间:2026-03-27
  • 出版时间:2025-10-01
补充材料
相关文章
文章信息
作者
出版历史
  • 收稿日期:2025-01-07
  • 修回日期:2025-02-11
基金
国家重点研发计划项目(2021YFC0122403)
作者信息
    北京航空航天大学 医学科学与工程学院;北京生物医学工程高精尖创新中心,北京 100191

通讯作者:

郭江真,副教授,E-mail:
陶春静,教授,E-mail:

*

为共同通信作者
参考文献
分享链接
https://castjournals.cast.org.cn/joweb/yyswlx/CN/10.16156/j.1004-7220.2025.05.031
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