{"status":"ok","message-type":"work","message-version":"1.0.0","message":{"indexed":{"date-parts":[[2026,6,9]],"date-time":"2026-06-09T16:10:48Z","timestamp":1781021448717,"version":"3.54.1"},"reference-count":47,"publisher":"Frontiers Media SA","license":[{"start":{"date-parts":[[2023,9,13]],"date-time":"2023-09-13T00:00:00Z","timestamp":1694563200000},"content-version":"vor","delay-in-days":0,"URL":"https:\/\/creativecommons.org\/licenses\/by\/4.0\/"}],"funder":[{"DOI":"10.13039\/501100012295","name":"NIHR Newcastle Biomedical Research Centre","doi-asserted-by":"publisher","id":[{"id":"10.13039\/501100012295","id-type":"DOI","asserted-by":"publisher"}]}],"content-domain":{"domain":["frontiersin.org"],"crossmark-restriction":true},"short-container-title":["Front. Physiol."],"abstract":"<jats:p><jats:bold>Introduction:<\/jats:bold> A pilot study assessing a novel approach to identify patients with Systemic Sclerosis (SSc) using deep learning analysis of multi-site photoplethysmography (PPG) waveforms (\u201cDL-PPG\u201d).<\/jats:p><jats:p><jats:bold>Methods:<\/jats:bold> PPG recordings having baseline, unilateral arm pressure cuff occlusion and reactive hyperaemia flush phases from 6 body sites were studied in 51 Controls and 20 SSc patients. RGB scalogram images were obtained from the PPG, using the continuous wavelet transform (CWT). 2 different pre-trained convolutional neural networks (CNNs, namely, GoogLeNet and EfficientNetB0) were trained to classify the SSc and Control groups, evaluating their performance using 10-fold stratified cross validation (CV). Their classification performance (i.e., accuracy, sensitivity, and specificity, with 95% confidence intervals) was also compared to traditional machine learning (ML), i.e., Linear Discriminant Analysis (LDA) and K-Nearest Neighbour (KNN).<\/jats:p><jats:p><jats:bold>Results:<\/jats:bold> On a participant basis DL-PPG accuracy, sensitivity and specificity for GoogLeNet were 83.1 (72.3\u201390.9), 75.0 (50.9\u201391.3) and 86.3 (73.7\u201394.3)% respectively, and for EfficientNetB0 were 87.3 (77.2\u201394.0), 80.0 (56.3\u201394.3) and 90.1 (78.6\u201396.7)%. The corresponding results for ML classification using LDA were 66.2 (53.9\u201377.0), 65.0 (40.8\u201384.6) and 66.7 (52.1\u201379.2)% respectively, and for KNN were 76.1 (64.5\u201385.4), 40.0 (19.1\u201363.9), and 90.2 (78.6\u201396.7)% respectively.<\/jats:p><jats:p><jats:bold>Discussion:<\/jats:bold> This study shows the potential of DL-PPG classification using CNNs to detect SSc. EfficientNetB0 gave an overall improved performance compared to GoogLeNet, with both CNNs performing better than the traditional ML methods tested. Our automatic AI approach, using transfer learning, could offer significant benefits for SSc diagnostics in a variety of clinical settings where low-cost portable and easy-to-use diagnostics can be beneficial.<\/jats:p>","DOI":"10.3389\/fphys.2023.1242807","type":"journal-article","created":{"date-parts":[[2023,9,13]],"date-time":"2023-09-13T21:42:04Z","timestamp":1694641324000},"update-policy":"https:\/\/doi.org\/10.3389\/crossmark-policy","source":"Crossref","is-referenced-by-count":5,"title":["Deep learning classification of systemic sclerosis from multi-site photoplethysmography signals"],"prefix":"10.3389","volume":"14","author":[{"given":"Sadaf","family":"Iqbal","sequence":"first","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Jaume","family":"Bacardit","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Bridget","family":"Griffiths","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"John","family":"Allen","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]}],"member":"1965","published-online":{"date-parts":[[2023,9,13]]},"reference":[{"key":"B1","doi-asserted-by":"publisher","first-page":"104","DOI":"10.1109\/OJEMB.2021.3066097","article-title":"Deep learning classification of systemic sclerosis skin using the MobileNetV2 model","volume":"2","author":"Akay","year":"2021","journal-title":"IEEE Open J. Eng. Med. Biol."},{"key":"B2","doi-asserted-by":"publisher","first-page":"R91","DOI":"10.1088\/0967-3334\/35\/7\/R91","article-title":"Microvascular imaging: techniques and opportunities for clinical physiological measurements","volume":"35","author":"Allen","year":"2014","journal-title":"Physiol. Meas."},{"key":"B3","doi-asserted-by":"publisher","first-page":"074001","DOI":"10.1088\/1361-6579\/ab9b67","article-title":"Age-related changes in pulse risetime measured by multi-site photoplethysmography","volume":"41","author":"Allen","year":"2020","journal-title":"Physiol. Meas."},{"key":"B4","doi-asserted-by":"publisher","first-page":"1","DOI":"10.1088\/0967-3334\/28\/3\/R01","article-title":"Photoplethysmography and its application in clinical physiological measurement","volume":"28","author":"Allen","year":"2007","journal-title":"Physiol. Meas."},{"key":"B5","doi-asserted-by":"publisher","first-page":"100301","DOI":"10.1088\/1361-6579\/ac2d82","article-title":"Photoplethysmography (PPG): state-of-the-art methods and applications","volume":"42","author":"Allen","year":"2021","journal-title":"Physiol. Meas."},{"key":"B6","doi-asserted-by":"publisher","first-page":"413","DOI":"10.1024\/0301-1526\/a000661","article-title":"ESVM guidelines-the diagnosis and management of Raynaud's phenomenon","volume":"46","author":"Belch","year":"2017","journal-title":"Vasa"},{"key":"B7","doi-asserted-by":"publisher","first-page":"195","DOI":"10.15406\/ijbsbe.2018.04.00125","article-title":"A review on wearable photoplethysmography sensors and their potential future applications in health care","volume":"4","author":"Castaneda","year":"2018","journal-title":"Int. J. Biosens. Bioelectron."},{"key":"B8","doi-asserted-by":"publisher","DOI":"10.1088\/1361-6579\/acead2","article-title":"The 2023 wearable photoplethysmography roadmap","author":"Charlton","year":"2023","journal-title":"Physiol. Meas"},{"key":"B9","doi-asserted-by":"publisher","first-page":"S3","DOI":"10.55563\/clinexprheumatol\/izadb8","article-title":"One year in review 2021: systemic sclerosis","volume":"39","author":"Di Battista","year":"2021","journal-title":"Clin. Exp. Rheumatol."},{"key":"B10","doi-asserted-by":"publisher","first-page":"100379","DOI":"10.1016\/j.cosrev.2021.100379","article-title":"A survey on deep learning and its applications","volume":"40","author":"Dong","year":"2021","journal-title":"Comput. Sci. Rev."},{"key":"B11","doi-asserted-by":"publisher","first-page":"14","DOI":"10.2174\/157340312801215782","article-title":"On the analysis of fingertip photoplethysmogram signals","volume":"8","author":"Elgendi","year":"2012","journal-title":"Curr. Cardiol. Rev."},{"key":"B12","doi-asserted-by":"publisher","first-page":"445","DOI":"10.2147\/MDER.S51426","article-title":"Non-invasive imaging of microcirculation: A technology review","volume":"7","author":"Eriksson","year":"2014","journal-title":"Med. Devices Evid. Res."},{"key":"B13","doi-asserted-by":"publisher","first-page":"78","DOI":"10.1186\/1471-2288-12-78","article-title":"t-tests, non-parametric tests, and large studies\u2014a paradox of statistical practice?","volume":"12","author":"Fagerland","year":"2012","journal-title":"BMC Med. Res. Methodol."},{"key":"B14","doi-asserted-by":"publisher","first-page":"539","DOI":"10.2147\/IJGM.S70892","article-title":"The spectral analysis of photoplethysmography to evaluate an independent cardiovascular risk factor","volume":"7","author":"Gandhi","year":"2014","journal-title":"Int. J. General Med."},{"key":"B15","volume-title":"Deep learning","author":"Goodfellow","year":"2016"},{"key":"B16","doi-asserted-by":"publisher","first-page":"580","DOI":"10.7861\/clinmed.2020-0754","article-title":"Raynaud's phenomenon","volume":"20","author":"Haque","year":"2020","journal-title":"Clin. Med."},{"key":"B17","first-page":"2649","article-title":"Evaluation and classification new algorithms in Image Resizing","volume":"5","author":"Hashemi","year":"2015","journal-title":"Int. J. Mechatronics, Electr. Comput. Technol."},{"key":"B18","doi-asserted-by":"publisher","first-page":"112","DOI":"10.1016\/j.berh.2016.04.001","article-title":"Raynaud's phenomenon","volume":"30","author":"Hughes","year":"2016","journal-title":"Best Pract. Res. Clin. Rheumatology"},{"key":"B19","doi-asserted-by":"publisher","first-page":"619692","DOI":"10.3389\/fdgth.2020.619692","article-title":"Advancing PPG signal quality and know-how through knowledge translation\u2014From experts to student and researcher","volume":"2","author":"Huthart","year":"2020","journal-title":"Front. Digital Health"},{"key":"B20","volume-title":"Multi-site photoplethysmography waveform analysis using machine learning for the detection of systemic sclerosis","author":"Iqbal","year":"2023"},{"key":"B21","first-page":"863","article-title":"A review of photoplethysmography-based physiological measurement and estimation, Part 2: multi-input methods","author":"Johnson","year":""},{"key":"B22","doi-asserted-by":"publisher","first-page":"84","DOI":"10.1145\/3065386","article-title":"Imagenet classification with deep convolutional neural networks","volume":"60","author":"Krizhevsky","year":"2017","journal-title":"Commun. ACM"},{"key":"B23","volume-title":"Photoplethysmography: Technology, signal analysis and applications","author":"Kyriacou","year":"2021"},{"key":"B24","doi-asserted-by":"publisher","first-page":"436","DOI":"10.1038\/nature14539","article-title":"Deep learning","volume":"521","author":"LeCun","year":"2015","journal-title":"Nature"},{"key":"B25","doi-asserted-by":"publisher","first-page":"101","DOI":"10.3390\/bios8040101","article-title":"Photoplethysmography and deep learning: enhancing hypertension risk stratification","volume":"8","author":"Liang","year":"2018","journal-title":"Biosensors"},{"key":"B26","first-page":"111","article-title":"Wearable ear blood oxygen saturation and pulse measurement system based on PPG","author":"Ma","year":""},{"key":"B27","doi-asserted-by":"publisher","first-page":"044004","DOI":"10.1088\/1361-6579\/ab807c","article-title":"Novel instrumental markers of proximal scleroderma provided by imaging photoplethysmography","volume":"41","author":"Mamontov","year":"2020","journal-title":"Physiol. Meas."},{"key":"B28","doi-asserted-by":"publisher","first-page":"1855","DOI":"10.1093\/rheumatology\/keu196","article-title":"Novel photoplethysmography cardiovascular assessments in patients with Raynaud\u2019s phenomenon and systemic sclerosis: A pilot study","volume":"53","author":"McKay","year":"2014","journal-title":"Rheumatology"},{"key":"B29","volume-title":"The discipline of machine learning","author":"Mitchell","year":"2006"},{"key":"B30","doi-asserted-by":"publisher","first-page":"36","DOI":"10.5120\/16850-6712","article-title":"An approach for ECG feature extraction using daubechies 4 (DB4) wavelet","volume":"96","author":"Mohamed","year":"2014","journal-title":"Int. J. Comput. Appl."},{"key":"B31","unstructured":"NHSinform2023"},{"key":"B32","doi-asserted-by":"publisher","first-page":"18","DOI":"10.1093\/rheumatology\/key026","article-title":"The patient experience of Raynaud\u2019s phenomenon in systemic sclerosis","volume":"58","author":"Pauling","year":"2019","journal-title":"Rheumatology"},{"key":"B33","doi-asserted-by":"publisher","first-page":"62","DOI":"10.1038\/s41584-021-00745-5","article-title":"Evaluating early diagnostic criteria for SSc","volume":"18","author":"Phillips","year":"2022","journal-title":"Nat. Rev. Rheumatol."},{"key":"B34","doi-asserted-by":"publisher","first-page":"292","DOI":"10.3109\/03009742.2010.530293","article-title":"The combination of laser Doppler perfusion imaging and photoplethysmography is useful in the characterization of scleroderma and primary Raynaud's phenomenon","volume":"40","author":"Rosato","year":"2011","journal-title":"Scand. J. Rheumatology"},{"key":"B35","doi-asserted-by":"publisher","first-page":"457","DOI":"10.1097\/MAJ.0b013e3181eecfad","article-title":"The different photoplethysmographic patterns can help to distinguish patients with primary and sclerodermic Raynaud phenomenon","volume":"340","author":"Rosato","year":"2010","journal-title":"Am. J. Med. Sci."},{"key":"B36","doi-asserted-by":"publisher","first-page":"2103","DOI":"10.1007\/s10067-018-4182-3","article-title":"The incidence, prevalence, and survival of systemic sclerosis in the UK Clinical Practice Research Datalink","volume":"37","author":"Royle","year":"2018","journal-title":"Clin. Rheumatol."},{"key":"B37","doi-asserted-by":"publisher","first-page":"9","DOI":"10.1016\/j.rvm.2016.03.001","article-title":"Raynaud phenomenon","volume":"4","author":"Silva","year":"2016","journal-title":"Rev. Vasc. Med."},{"key":"B38","doi-asserted-by":"publisher","first-page":"595","DOI":"10.1001\/archinte.158.6.595","article-title":"Outcomes in primary Raynaud phenomenon: A meta-analysis of the frequency, rates, and predictors of transition to secondary diseases","volume":"158","author":"Spencer-Green","year":"1998","journal-title":"Archives Intern. Med."},{"key":"B39","unstructured":"Sruk\n            SRUK\n          2023"},{"key":"B40","doi-asserted-by":"publisher","first-page":"581","DOI":"10.1002\/art.1780230510","article-title":"Preliminary criteria for the classification of systemic sclerosis (scleroderma). Subcommittee for scleroderma criteria of the American Rheumatism Association Diagnostic and Therapeutic Criteria Committee","volume":"23","author":"Subcommittee for Scleroderma Criteria of the American Rheumatism Association Diagnostic and Therapeutic Criteria Committee","year":"1980","journal-title":"Arthritis Rheum."},{"key":"B41","first-page":"1","article-title":"Going deeper with convolutions","author":"Szegedy","year":""},{"key":"B42","first-page":"2820","article-title":"Mnasnet: platform-aware neural architecture search for mobile","author":"Tan","year":""},{"key":"B43","first-page":"6105","article-title":"Efficientnet: rethinking model scaling for convolutional neural networks","author":"Tan","year":""},{"key":"B44","doi-asserted-by":"publisher","first-page":"682","DOI":"10.3390\/math11030682","article-title":"Recent advances in stochastic gradient descent in deep learning","volume":"11","author":"Tian","year":"2023","journal-title":"Mathematics"},{"key":"B45","doi-asserted-by":"publisher","first-page":"2737","DOI":"10.1002\/art.38098","article-title":"2013 classification criteria for systemic sclerosis: an American College of rheumatology\/European league against rheumatism collaborative initiative","volume":"65","author":"van den Hoogen","year":"2013","journal-title":"Arthritis Rheum."},{"key":"B46","doi-asserted-by":"publisher","first-page":"20170250","DOI":"10.1098\/rsta.2017.0250","article-title":"Quantitative feature analysis of continuous analytic wavelet transforms of electrocardiography and electromyography","volume":"376","author":"Wachowiak","year":"2018","journal-title":"Philosophical Trans. R. Soc. A Math. Phys. Eng. Sci."},{"key":"B47","doi-asserted-by":"publisher","first-page":"754","DOI":"10.1136\/ard.2006.062901","article-title":"Clinical risk assessment of organ manifestations in systemic sclerosis: A report from the EULAR scleroderma trials and research group database","volume":"66","author":"Walker","year":"2007","journal-title":"Ann. Rheum. Dis."}],"container-title":["Frontiers in Physiology"],"original-title":[],"link":[{"URL":"https:\/\/www.frontiersin.org\/articles\/10.3389\/fphys.2023.1242807\/full","content-type":"unspecified","content-version":"vor","intended-application":"similarity-checking"}],"deposited":{"date-parts":[[2023,9,13]],"date-time":"2023-09-13T21:42:08Z","timestamp":1694641328000},"score":1,"resource":{"primary":{"URL":"https:\/\/www.frontiersin.org\/articles\/10.3389\/fphys.2023.1242807\/full"}},"subtitle":[],"short-title":[],"issued":{"date-parts":[[2023,9,13]]},"references-count":47,"alternative-id":["10.3389\/fphys.2023.1242807"],"URL":"https:\/\/doi.org\/10.3389\/fphys.2023.1242807","relation":{},"ISSN":["1664-042X"],"issn-type":[{"value":"1664-042X","type":"electronic"}],"subject":[],"published":{"date-parts":[[2023,9,13]]},"article-number":"1242807"}}