{"status":"ok","message-type":"work","message-version":"1.0.0","message":{"indexed":{"date-parts":[[2026,10,28]],"date-time":"2026-10-28T03:25:26Z","timestamp":1793157926177,"version":"4.5.0"},"reference-count":34,"publisher":"Frontiers Media SA","license":[{"start":{"date-parts":[[2026,5,18]],"date-time":"2026-05-18T00:00:00Z","timestamp":1779062400000},"content-version":"vor","delay-in-days":0,"URL":"https:\/\/creativecommons.org\/licenses\/by\/4.0\/"}],"content-domain":{"domain":["frontiersin.org"],"crossmark-restriction":true},"short-container-title":["Front. Artif. Intell."],"abstract":"<jats:sec>\n                    <jats:title>Objective<\/jats:title>\n                    <jats:p>Deep learning models for phonocardiogram (PCG) classification often achieve high accuracy but lack transparency in relation to clinically meaningful features. Existing explainable artificial intelligence (XAI) methods rarely ensure alignment between model attributions and cardiac physiology.<\/jats:p>\n                  <\/jats:sec>\n                  <jats:sec>\n                    <jats:title>Methods<\/jats:title>\n                    <jats:p>We propose Scale-Consistent Attribution (SCA), a training-time regularization method that incorporates domain knowledge into model learning. SCA aligns spectral attention with a soft clinical prior, distinguishing low-frequency heart sounds from high-frequency murmurs. The method is evaluated on the PhysioNet 2016 and CirCor DigiScope datasets.<\/jats:p>\n                  <\/jats:sec>\n                  <jats:sec>\n                    <jats:title>Results<\/jats:title>\n                    <jats:p>SCA remarkably improves the physiological plausibility of explanations. Specifically, it reduces the divergence between model attribution and clinical priors by an order of magnitude. Crucially, the model maintains competitive classification accuracy. Ablation studies with incorrect priors confirm that these benefits stem from accurate clinical medical knowledge. Qualitative and quantitive analysis further reveals that SCA shifts model focus towards clinically relevant frequency regions.<\/jats:p>\n                  <\/jats:sec>\n                  <jats:sec>\n                    <jats:title>Conclusion<\/jats:title>\n                    <jats:p>SCA offers a robust framework for trustworthy and explainable PCG classifiers, improving the clinical plausibility of explanations while maintaining strong predictive performance.<\/jats:p>\n                  <\/jats:sec>","DOI":"10.3389\/frai.2026.1780496","type":"journal-article","created":{"date-parts":[[2026,5,18]],"date-time":"2026-05-18T05:42:50Z","timestamp":1779082970000},"update-policy":"https:\/\/doi.org\/10.3389\/crossmark-policy","source":"Crossref","is-referenced-by-count":1,"title":["Prior-aligned frequency-domain explanations for heart sound classification: a scale-consistent attribution approach"],"prefix":"10.3389","volume":"9","author":[{"given":"Qiyang","family":"Sun","sequence":"first","affiliation":[{"name":"Group on Language Audio & Music, Department of Computing, Imperial College London","place":["London, United Kingdom"]}],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Yupei","family":"Li","sequence":"additional","affiliation":[{"name":"Group on Language Audio & Music, Department of Computing, Imperial College London","place":["London, United Kingdom"]},{"name":"Chair of Health Informatics (CHI), TUM University Hospital","place":["Munich, Germany"]}],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Aydin","family":"Javadov","sequence":"additional","affiliation":[{"name":"Mobiliar Lab for Analytics, Swiss Federal Institute of Technology","place":["Zurich, Switzerland"]}],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Xiaoliang","family":"Wu","sequence":"additional","affiliation":[{"name":"School of Electronics and Computer Science, University of Southampton","place":["Southampton, United Kingdom"]}],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Bj\u00f6rn W.","family":"Schuller","sequence":"additional","affiliation":[{"name":"Group on Language Audio & Music, Department of Computing, Imperial College London","place":["London, United Kingdom"]},{"name":"Chair of Health Informatics (CHI), TUM University Hospital","place":["Munich, Germany"]},{"name":"Munich Data Science Institute (MDSI)","place":["Munich, Germany"]},{"name":"Munich Center for Machine Learning (MCML)","place":["Munich, Germany"]}],"role":[{"vocabulary":"crossref","role":"author"}]}],"member":"1965","published-online":{"date-parts":[[2026,5,18]]},"reference":[{"key":"B1","doi-asserted-by":"publisher","first-page":"e1311","DOI":"10.1002\/smm2.1311","article-title":"Advancements in wearable heart sounds devices for the monitoring of cardiovascular diseases","volume":"6","author":"Ahmad","year":"2025","journal-title":"SmartMat"},{"key":"B2","doi-asserted-by":"publisher","first-page":"123099","DOI":"10.1109\/ACCESS.2024.3451660","article-title":"A hybrid framework of transformer encoder and residential conventional for cardiovascular disease recognition using heart sounds","volume":"12","author":"Al-Tam","year":"2024","journal-title":"IEEE Access"},{"key":"B3","doi-asserted-by":"publisher","first-page":"121","DOI":"10.1016\/0141-5425(84)90054-2","article-title":"Spectral analysis of heart sounds: relationships between some physical characteristics and frequency spectra of first and second heart sounds in normals and hypertensives","volume":"6","author":"Arnott","year":"1984","journal-title":"J. Biomed. Eng"},{"key":"B4","doi-asserted-by":"publisher","first-page":"216","DOI":"10.1109\/CISP.2008.758","article-title":"\u201cIsolation of systolic heart murmurs using wavelet transform and energy index,\u201d","author":"Atanasov","year":"2008","journal-title":"2008 Congress on Image and Signal Processing"},{"key":"B5","doi-asserted-by":"publisher","first-page":"1","DOI":"10.1109\/ICASSP49660.2025.10889947","article-title":"\u201cValve token masked autoencoder for missing recordings on cardiac abnormality classification,\u201d","author":"Chang","year":"2025","journal-title":"ICASSP 2025\u20132025 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP)"},{"key":"B6","doi-asserted-by":"publisher","first-page":"609","DOI":"10.22489\/CinC.2016.179-154","article-title":"\u201cClassification of normal\/abnormal heart sound recordings: the physionet\/computing in cardiology challenge 2016,\u201d","author":"Clifford","year":"2016","journal-title":"2016 Computing in cardiology conference (CinC)"},{"key":"B7","doi-asserted-by":"publisher","first-page":"108152","DOI":"10.1016\/j.apacoust.2021.108152","article-title":"Heart sounds classification using convolutional neural network with 1d-local binary pattern and 1d-local ternary pattern features","volume":"180","author":"Er","year":"2021","journal-title":"Appl. Acoust"},{"key":"B8","doi-asserted-by":"publisher","first-page":"1","DOI":"10.22489\/CinC.2022.151","article-title":"\u201cAn LSTM-based listener for early detection of heart disease,\u201d","author":"Gemke","year":"2022","journal-title":"2022 Computing in Cardiology (CinC)"},{"key":"B9","doi-asserted-by":"publisher","DOI":"10.1109\/CVPR.2016.90","article-title":"\u201cDeep residual learning for image recognition,\u201d","author":"He","year":"2016","journal-title":"Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition"},{"key":"B10","doi-asserted-by":"publisher","DOI":"10.1109\/CVPR.2017.243","article-title":"\u201cDensely connected convolutional networks,\u201d","author":"Huang","year":"2017","journal-title":"Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition"},{"key":"B11","article-title":"\u201cRAxSS: retrieval-augmented sparse sampling for explainable variable-length medical time series classification,\u201d","author":"Javadov","year":"2025","journal-title":"NeurIPS 2025 Workshop on Learning from Time Series for Health"},{"key":"B12","doi-asserted-by":"publisher","first-page":"170619","DOI":"10.1109\/ACCESS.2025.3612392","article-title":"A comparative analysis of machine learning and deep learning approaches for phonocardiogram classification using dataset integration","volume":"13","author":"Kalimuthu","year":"2025","journal-title":"IEEE Access"},{"key":"B13","article-title":"\u201cOne wave to explain them all: a unifying perspective on feature attribution,\u201d","author":"Kasmi","year":"2025","journal-title":"Forty-Second International Conference on Machine Learning"},{"key":"B14","doi-asserted-by":"publisher","first-page":"165","DOI":"10.1007\/978-981-99-0369-6_11","article-title":"\u201cDeep autoencoder neural networks for heart sound classification,\u201d","author":"Kaushal","year":"2023","journal-title":"Artificial Intelligence in Medical Virology"},{"key":"B15","doi-asserted-by":"publisher","first-page":"470","DOI":"10.4093\/dmj.2023.0036","article-title":"The role of echocardiography in evaluating cardiovascular diseases in patients with diabetes mellitus","volume":"47","author":"Lee","year":"2023","journal-title":"Diab. Metabol. J"},{"key":"B16","doi-asserted-by":"publisher","first-page":"632","DOI":"10.3233\/SHTI251454","article-title":"\u201cExplainable ai for healthcare,\u201d","author":"Li","year":"2025","journal-title":"Handbook on Smart Health"},{"key":"B17","doi-asserted-by":"publisher","first-page":"2343","DOI":"10.1016\/j.jacc.2023.11.003","article-title":"A heart-healthy and stroke-free world: using data to inform global action","volume":"82","author":"Mensah","year":"2023","journal-title":"J. Am. College Cardiol"},{"key":"B18","doi-asserted-by":"publisher","first-page":"183","DOI":"10.2147\/JMDH.S193904","article-title":"The first 200 years of cardiac auscultation and future perspectives","volume":"12","author":"Montinari","year":"2019","journal-title":"J. Multidisc. Healthcare"},{"key":"B19","doi-asserted-by":"publisher","first-page":"2524","DOI":"10.1109\/JBHI.2021.3137048","article-title":"The circor digiscope dataset: from murmur detection to murmur classification","volume":"26","author":"Oliveira","year":"2021","journal-title":"IEEE J. Biomed. Health Inform"},{"key":"B20","doi-asserted-by":"publisher","first-page":"43","DOI":"10.1007\/s13755-024-00302-w","article-title":"Rapid detection and interpretation of heart murmurs using phonocardiograms, transfer learning and explainable artificial intelligence","volume":"12","author":"\u00d6zcan","year":"2024","journal-title":"Health Inf. Sci. Syst"},{"key":"B21","doi-asserted-by":"publisher","first-page":"1","DOI":"10.22489\/CinC.2022.001","article-title":"\u201cClassification of phonocardiograms using residual convolutional neural network and MLP,\u201d","author":"Peng","year":"2022","journal-title":"2022 Computing in Cardiology (CinC)"},{"key":"B22","doi-asserted-by":"publisher","first-page":"100322","DOI":"10.1016\/j.mlwa.2022.100322","article-title":"Deep attention-based neural networks for explainable heart sound classification","volume":"9","author":"Ren","year":"2022","journal-title":"Mach. Learn. Applic"},{"key":"B23","doi-asserted-by":"publisher","DOI":"10.1007\/978-3-031-51458-6_7","article-title":"\u201cAssessment of the heart,\u201d","author":"Riley","year":"2024","journal-title":"History and Physical for the Pediatric Dental Patient: Establishing a Systematic Approach for Procedural Sedation"},{"key":"B24","doi-asserted-by":"publisher","first-page":"1","DOI":"10.1109\/APCI65531.2025.11137034","article-title":"\u201cDetection of abnormal heart sounds using mfcc and deep neural networks,\u201d","author":"Sabu","year":"2025","journal-title":"2025 International Conference on Advancements in Power, Communication and Intelligent Systems (APCI)"},{"key":"B25","doi-asserted-by":"publisher","DOI":"10.1109\/CVPR.2018.00474","article-title":"\u201cMobilenetv2: Inverted residuals and linear bottlenecks,\u201d","author":"Sandler","year":"2018","journal-title":"Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition"},{"key":"B26","doi-asserted-by":"publisher","first-page":"1465","DOI":"10.1016\/j.hlc.2023.09.011","article-title":"Determinants and outcomes of invasive coronary angiography in unselected patients presenting with chest pain to emergency departments in western australian teaching hospitals","volume":"32","author":"Sanfilippo","year":"2023","journal-title":"Heart, Lung Circ"},{"key":"B27","doi-asserted-by":"publisher","first-page":"110774","DOI":"10.1016\/j.compbiomed.2025.110774","article-title":"Spectronet-lstm: an interpretable deep learning approach to cardiac anomaly detection through heartbeat sound analysis","volume":"196","author":"Sharma","year":"2025","journal-title":"Comput. Biol. Med"},{"key":"B28","first-page":"3145","article-title":"\u201cLearning important features through propagating activation differences,\u201d","volume-title":"International Conference on Machine Learning","author":"Shrikumar","year":"2017"},{"key":"B29","doi-asserted-by":"publisher","first-page":"39563","DOI":"10.1007\/s11042-023-16930-5","article-title":"An automated diagnosis model for classifying cardiac abnormality utilizing deep neural networks","volume":"83","author":"Singh","year":"2024","journal-title":"Multimed. Tools Appl"},{"key":"B30","doi-asserted-by":"publisher","first-page":"1","DOI":"10.1145\/3709367","article-title":"Explainable artificial intelligence for medical applications: a review","volume":"6","author":"Sun","year":"2025","journal-title":"ACM Trans. Comput. Healthcare"},{"key":"B31","doi-asserted-by":"publisher","first-page":"382","DOI":"10.3390\/books978-3-7258-5598-8","article-title":"Advances in cardiovascular imaging: a platform to share recent research findings through utilization of advanced imaging technologies","volume":"12","author":"Sun","year":"2025","journal-title":"J. Cardiov. Dev. Dis"},{"key":"B32","doi-asserted-by":"publisher","first-page":"861","DOI":"10.1109\/ICASSP48485.2024.10446246","article-title":"\u201cIntelligent cardiac auscultation for murmur detection via parallel-attentive models with uncertainty estimation,\u201d","author":"Zhang","year":"2024","journal-title":"ICASSP 2024\u20132024 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP)"},{"key":"B33","doi-asserted-by":"publisher","first-page":"0182","DOI":"10.34133\/hds.0182","article-title":"Deep learning in heart sound analysis: from techniques to clinical applications","volume":"4","author":"Zhao","year":"2024","journal-title":"Health Data Sci"},{"key":"B34","doi-asserted-by":"publisher","first-page":"3222","DOI":"10.3390\/electronics13163222","article-title":"Review of phonocardiogram signal analysis: insights from the physionet\/cinc challenge 2016 database","volume":"13","author":"Zhu","year":"2024","journal-title":"Electronics"}],"container-title":["Frontiers in Artificial Intelligence"],"original-title":[],"link":[{"URL":"https:\/\/www.frontiersin.org\/articles\/10.3389\/frai.2026.1780496\/full","content-type":"unspecified","content-version":"vor","intended-application":"similarity-checking"}],"deposited":{"date-parts":[[2026,5,18]],"date-time":"2026-05-18T05:42:51Z","timestamp":1779082971000},"score":1,"resource":{"primary":{"URL":"https:\/\/www.frontiersin.org\/articles\/10.3389\/frai.2026.1780496\/full"}},"subtitle":[],"short-title":[],"issued":{"date-parts":[[2026,5,18]]},"references-count":34,"alternative-id":["10.3389\/frai.2026.1780496"],"URL":"https:\/\/doi.org\/10.3389\/frai.2026.1780496","relation":{},"ISSN":["2624-8212"],"issn-type":[{"value":"2624-8212","type":"electronic"}],"subject":[],"published":{"date-parts":[[2026,5,18]]},"article-number":"1780496"}}