{"status":"ok","message-type":"work","message-version":"1.0.0","message":{"indexed":{"date-parts":[[2026,9,26]],"date-time":"2026-09-26T02:18:42Z","timestamp":1790389122204,"version":"4.1.0"},"reference-count":35,"publisher":"Public Library of Science (PLoS)","issue":"8","license":[{"start":{"date-parts":[[2024,8,16]],"date-time":"2024-08-16T00:00:00Z","timestamp":1723766400000},"content-version":"vor","delay-in-days":0,"URL":"http:\/\/creativecommons.org\/licenses\/by\/4.0\/"}],"funder":[{"DOI":"10.13039\/100020595","name":"National Science and Technology Council","doi-asserted-by":"publisher","award":["110-2314-B-037-068-MY3"],"award-info":[{"award-number":["110-2314-B-037-068-MY3"]}],"id":[{"id":"10.13039\/100020595","id-type":"DOI","asserted-by":"publisher"}]},{"DOI":"10.13039\/100020595","name":"National Science and Technology Council","doi-asserted-by":"publisher","award":["NSTC 112-2221-E-017-003-MY2"],"award-info":[{"award-number":["NSTC 112-2221-E-017-003-MY2"]}],"id":[{"id":"10.13039\/100020595","id-type":"DOI","asserted-by":"publisher"}]}],"content-domain":{"domain":["www.plosone.org"],"crossmark-restriction":false},"short-container-title":["PLoS ONE"],"abstract":"<jats:p>End-stage kidney disease (ESKD) presents a significant public health challenge, with hemodialysis (HD) remaining one of the most prevalent kidney replacement therapies. Ensuring the longevity and functionality of arteriovenous accesses is challenging for HD patients. Blood flow sound, which contains valuable information, has often been neglected in the past. However, machine learning offers a new approach, leveraging data non-invasively and learning autonomously to match the experience of healthcare professionas. This study aimed to devise a model for detecting arteriovenous grafts (AVGs) stenosis. A smartphone stethoscope was used to record the sound of AVG blood flow at the arterial and venous sides, with each recording lasting one minute. The sound recordings were transformed into mel spectrograms, and a 14-layer convolutional neural network (CNN) was employed to detect stenosis. The CNN comprised six convolution blocks with 3x3 kernel mapping, batch normalization, and rectified linear unit activation function. We applied contrastive learning to train the pre-training audio neural networks model with unlabeled data through self-supervised learning, followed by fine-tuning. In total, 27,406 dialysis session blood flow sounds were documented, including 180 stenosis blood flow sounds. Our proposed framework demonstrated a significant improvement (p&lt;0.05) over training from scratch and a popular pre-trained audio neural networks (PANNs) model, achieving an accuracy of 0.9279, precision of 0.8462, and recall of 0.8077, compared to previous values of 0.8649, 0.7391, and 0.6538. This study illustrates how contrastive learning with unlabeled blood flow sound data can enhance convolutional neural networks for detecting AVG stenosis in HD patients.<\/jats:p>","DOI":"10.1371\/journal.pone.0308385","type":"journal-article","created":{"date-parts":[[2024,8,16]],"date-time":"2024-08-16T17:30:45Z","timestamp":1723829445000},"page":"e0308385","update-policy":"https:\/\/doi.org\/10.1371\/journal.pone.corrections_policy","source":"Crossref","is-referenced-by-count":3,"title":["The application of blood flow sound contrastive learning to predict arteriovenous graft stenosis of patients with hemodialysis"],"prefix":"10.1371","volume":"19","author":[{"ORCID":"https:\/\/orcid.org\/0000-0001-6888-4138","authenticated-orcid":true,"given":"Hugo Y.-H.","family":"Lin","sequence":"first","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Tiffany","family":"Shien","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Juan-Wei","family":"Xu","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Yen-Jung","family":"Kuo","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Phang-Lang","family":"Chen","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Sheng-Wen","family":"Niu","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"I-Ching","family":"Kuo","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Hsuan-Fu","family":"Kuo","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Kai-Chieh","family":"Yang","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Yi-Ren","family":"Yeh","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]}],"member":"340","published-online":{"date-parts":[[2024,8,16]]},"reference":[{"key":"pone.0308385.ref001","unstructured":"USRDS Annual Data Report: Epidemiology of Kidney Disease in the United States, 2023."},{"key":"pone.0308385.ref002","article-title":"A comparison of the epidemiology of kidney replacement therapy between Europe and the United States: 2021 data of the ERA Registry and the USRDS","author":"V.S. Stel","year":"2024","journal-title":"Nephrol Dial Transplant"},{"key":"pone.0308385.ref003","doi-asserted-by":"crossref","first-page":"10","DOI":"10.1053\/j.ajkd.2018.01.034","article-title":"Medicare Costs Associated With Arteriovenous Fistulas Among US Hemodialysis Patients","volume":"72","author":"M. Thamer","year":"2018","journal-title":"American Journal of Kidney Diseases"},{"key":"pone.0308385.ref004","doi-asserted-by":"crossref","first-page":"402","DOI":"10.1016\/j.crad.2006.01.003","article-title":"Salvaging and maintaining non-maturing Brescia-Cimino haemodialysis fistulae by percutaneous intervention","volume":"61","author":"C. Sandhu","year":"2006","journal-title":"Clinical Radiology"},{"key":"pone.0308385.ref005","doi-asserted-by":"crossref","first-page":"2078","DOI":"10.1046\/j.1523-1755.1999.00409.x","article-title":"Vascular access for hemodialysis","volume":"55","author":"S. J. Schwab","year":"1999","journal-title":"Kidney International"},{"key":"pone.0308385.ref006","first-page":"S150","article-title":"NKF-KDOQI clinical practice guidelines for vascular access","volume":"30","author":"National Kidney Foundation-Dialysis Outcomes Quality Initiative","year":"1997","journal-title":"American Journal of Kidney Diseases"},{"key":"pone.0308385.ref007","doi-asserted-by":"crossref","first-page":"1833","DOI":"10.1056\/NEJMra0902710","article-title":"Hemodialysis","volume":"363","author":"J. Himmelfarb","year":"2010","journal-title":"The New England Journal of Medicine"},{"issue":"1","key":"pone.0308385.ref008","doi-asserted-by":"crossref","first-page":"3","DOI":"10.1177\/10775587221093043","article-title":"Synchronous Home-Based Telemedicine for Primary Care: A Review","volume":"80","author":"Z. Lindenfeld","year":"2023","journal-title":"Medical Care Research and Review"},{"issue":"1","key":"pone.0308385.ref009","doi-asserted-by":"crossref","first-page":"1009","DOI":"10.1186\/s13063-022-06946-x","article-title":"Effects of Internet of Things-based power cycling and neuromuscular training on pain and walking ability in elderly patients with KOA: protocol for a randomized controlled trial","volume":"23","author":"Y. Zhang","year":"2022","journal-title":"Trials"},{"key":"pone.0308385.ref010","doi-asserted-by":"crossref","first-page":"e195","DOI":"10.1016\/S2589-7500(20)30292-2","article-title":"Approval of artificial intelligence and machine learning-based medical devices in the USA and Europe (2015\u201320): a comparative analysis","volume":"3","author":"U. J. Muehlematter","year":"2021","journal-title":"The Lancet Digital Health"},{"key":"pone.0308385.ref011","doi-asserted-by":"crossref","first-page":"233","DOI":"10.1038\/nmeth.4642","article-title":"Statistics versus machine learning","volume":"15","author":"D. Bzdok","year":"2018","journal-title":"Nature Methods"},{"key":"pone.0308385.ref012","unstructured":"I. Goodfellow, Y. Bengio, A. Courville,\u201cDeep Learning\u201d, MIT Press, 2016."},{"key":"pone.0308385.ref013","doi-asserted-by":"crossref","unstructured":"K. He, H. Fan, Y. Wu, S. Xie, and Ross Girshick. \u201cMomentum contrast for unsupervised visual representation learning,\u201d IEEE Conference on Computer Vision and Pattern Recognition, 2020.","DOI":"10.1109\/CVPR42600.2020.00975"},{"key":"pone.0308385.ref014","unstructured":"T. Chen, S. Kornblith, M. Norouzi, and G. Hinton. \u201cA simple framework for contrastive learning of visual representations,\u201d International Conference on Machine Learning, 2020."},{"key":"pone.0308385.ref015","article-title":"Duplex ultrasound evaluation of hemodialysis access: a detailed protocol","author":"V. Teodorescu","year":"2012","journal-title":"International Journal of Nephrology"},{"key":"pone.0308385.ref016","doi-asserted-by":"crossref","first-page":"253","DOI":"10.1007\/s40477-014-0113-6","article-title":"Color Doppler ultrasound and arteriovenous fistulas for hemodialysis","volume":"17","author":"P. Zamboli","year":"2014","journal-title":"Journal of Ultrasound"},{"key":"pone.0308385.ref017","doi-asserted-by":"crossref","first-page":"779","DOI":"10.1016\/j.crad.2016.01.011","article-title":"Advanced flow MRI: emerging techniques and applications","volume":"71","author":"M. Markl","year":"2016","journal-title":"Clinical Radiology"},{"key":"pone.0308385.ref018","first-page":"173","article-title":"Primary patency rate of native AV fistula: long term follow up","volume":"5","author":"G. H. Kazemzadeh","year":"2012","journal-title":"International Journal of Clinical and Experimental Medicine"},{"key":"pone.0308385.ref019","doi-asserted-by":"crossref","first-page":"910","DOI":"10.1016\/S0272-6386(99)70425-7","article-title":"Salvage of the nonfunctioning arteriovenous fistula","volume":"33","author":"G. A. Beathard","year":"1999","journal-title":"American Journal of Kidney Diseases"},{"key":"pone.0308385.ref020","doi-asserted-by":"crossref","first-page":"2880","DOI":"10.1109\/TASLP.2020.3030497","article-title":"PANNs: Large-Scale Pretrained Audio Neural Networks for Audio Pattern Recognition","volume":"28","author":"Y. Zhang","year":"2020","journal-title":"IEEE\/ACM Transactions on Audio, Speech, and Language Processing"},{"key":"pone.0308385.ref021","doi-asserted-by":"crossref","unstructured":"X. Chen, S. Xie, K. He. \u201cAn Empirical Study of Training Self-Supervised Vision Transformers,\u201d International Conference on Computer Vision, 2021.","DOI":"10.1109\/ICCV48922.2021.00950"},{"key":"pone.0308385.ref022","doi-asserted-by":"crossref","unstructured":"M. Caron, H. Touvron, I. Misra, H. Jegou, J. Mairal, P. Bojanowski, et al. \u201cEmerging Properties in Self-Supervised Vision Transformers,\u201d International Conference on Computer Vision, 2021.","DOI":"10.1109\/ICCV48922.2021.00951"},{"key":"pone.0308385.ref023","doi-asserted-by":"crossref","unstructured":"S. Wang, S. Tripathy, A. Mesaros, \u201cSelf-Supervised Learning of Audio Representations using Angular Contrastive Loss,\u201d ICASSP, 2023.","DOI":"10.1109\/ICASSP49357.2023.10094706"},{"key":"pone.0308385.ref024","doi-asserted-by":"crossref","unstructured":"Bae et al., \u201cPatch-Mix Contrastive Learning with Audio Spectrogram Transformer on Respiratory Sound Classification,\u201d INTERSPEECH, 2023.","DOI":"10.21437\/Interspeech.2023-1426"},{"key":"pone.0308385.ref025","unstructured":"S. Kim, G. Lee, S. Bae, S.-Y. Yun. \u201cMixCo: Mix-up Contrastive Learning for Visual Representation,\u201d NeurIPS Workshop on Self-Supervised Learning, 2020."},{"issue":"163","key":"pone.0308385.ref026","article-title":"Deep learning analysis of blood flow sounds to detect arteriovenous fistula stenosis","volume":"6","author":"Zhou","year":"2023","journal-title":"npj Digital Medicine"},{"key":"pone.0308385.ref027","doi-asserted-by":"crossref","unstructured":"K. Koutini, J. Schlueter, H. Eghbal-zadeh, G. Widmer, \u201cEfficient Training of Audio Transformers with Patchout\u201d, Interspeech, 2022.","DOI":"10.21437\/Interspeech.2022-227"},{"key":"pone.0308385.ref028","unstructured":"N. Zeghidour, O. Teboul, F. d. C. Quitry, M. Tagliasacchi, \u201cLEAF: A Learnable Frontend for Audio Classification\u201d, International Conference on Learning Representations, 2021."},{"key":"pone.0308385.ref029","unstructured":"J. Devlin, M.-W. Chang, K. Lee, K. Toutanova, \u201cBERT: Pre-training of Deep Bidirectional Transformers for Language Understanding\u201d, Conference of the North American Chapter of the Association for Computational Linguistics, 2019."},{"key":"pone.0308385.ref030","unstructured":"A. v. d. Oord, Y. Li, O. Vinyals, \u201cRepresentation Learning with Contrastive Predictive Coding\u201d, Conference on Neural Information Processing Systems(NeurIPS), 2018."},{"issue":"6","key":"pone.0308385.ref031","doi-asserted-by":"crossref","first-page":"904","DOI":"10.1177\/11297298211015095","article-title":"Forearm brachial artery to brachial vein looped arteriovenous grafts: A single center, single surgeon observational study","volume":"23","author":"Zhou J.Liu","year":"2022","journal-title":"The Journal of Vascular Access"},{"issue":"10","key":"pone.0308385.ref032","doi-asserted-by":"crossref","first-page":"2245","DOI":"10.1161\/01.ATV.17.10.2245","article-title":"Increased blood flow induces regression of intimal hyperplasia","volume":"17","author":"E.J. Mattsson","year":"1997","journal-title":"Arteriosclerosis, Thrombosis, and Vascular Biology"},{"issue":"1","key":"pone.0308385.ref033","doi-asserted-by":"crossref","first-page":"53","DOI":"10.1038\/nrm2596","article-title":"Mechanotransduction in vascular physiology and atherogenesis","volume":"10","author":"C. Hahn","year":"2009","journal-title":"Nature Reviews Molecular Cell Biology"},{"key":"pone.0308385.ref034","doi-asserted-by":"crossref","first-page":"187","DOI":"10.1007\/978-1-4615-5959-7_16","article-title":"Compensatory enlargement, remodeling, and restenosis","volume":"430","author":"G KerenG","year":"1997","journal-title":"Advances in Experimental Medicine and Biology"},{"issue":"4","key":"pone.0308385.ref035","doi-asserted-by":"crossref","first-page":"609","DOI":"10.1161\/01.ATV.13.4.609","article-title":"Cell proliferation in human arteriovenous fistulas used for hemodialysis","volume":"13","author":"M. Rekhter","year":"1993","journal-title":"Arteriosclerosis, Thrombosis, and Vascular Biology"}],"container-title":["PLOS ONE"],"original-title":[],"language":"en","link":[{"URL":"https:\/\/dx.plos.org\/10.1371\/journal.pone.0308385","content-type":"unspecified","content-version":"vor","intended-application":"similarity-checking"}],"deposited":{"date-parts":[[2024,8,16]],"date-time":"2024-08-16T17:31:05Z","timestamp":1723829465000},"score":1,"resource":{"primary":{"URL":"https:\/\/dx.plos.org\/10.1371\/journal.pone.0308385"}},"subtitle":[],"editor":[{"given":"Mohammad Amin","family":"Fraiwan","sequence":"first","affiliation":[],"role":[{"vocabulary":"crossref","role":"editor"}]}],"short-title":[],"issued":{"date-parts":[[2024,8,16]]},"references-count":35,"journal-issue":{"issue":"8","published-online":{"date-parts":[[2024,8,16]]}},"URL":"https:\/\/doi.org\/10.1371\/journal.pone.0308385","relation":{},"ISSN":["1932-6203"],"issn-type":[{"value":"1932-6203","type":"electronic"}],"subject":[],"published":{"date-parts":[[2024,8,16]]}}}