MultiScaleKANNet: a hybridCNN-KAN-Transformer architecture forradiographic bone-loss risk stratification fromknee X-rays

MultiScaleKANNet: a hybridCNN-KAN-Transformer architecture forradiographic bone-loss risk stratification fromknee X-rays | Research Square window.SnipcartSettings = { analytics: { enabled: false } }; (function() { var accessVector = localStorage.getItem('access_vector') || ''; window.dataLayer = window.dataLayer || []; if (accessVector) { window.dataLayer.push({ user: { profile: { profileInfo: { snid: accessVector } } } }); } })(); (function(w,d,s,l,i){w[l]=w[l]||[];w[l].push({'gtm.start':new Date().getTime(),event:'gtm.js'});var f=d.getElementsByTagName(s)[0],j=d.createElement(s),dl=l!='dataLayer'?'&l='+l:'';j.async=true;j.src='https://www.googletagmanager.com/gtm.js?id='+i+dl;f.parentNode.insertBefore(j,f);})(window,document,'script','dataLayer','GTM-K279D39R'); Browse Preprints In Review Journals COVID-19 Preprints AJE Video Bytes Research Tools Research Promotion AJE Professional Editing AJE Rubriq About Preprint Platform In Review Editorial Policies Our Team Advisory Board Help Center Sign In Submit a Preprint Cite Share Download PDF Article MultiScaleKANNet: a hybridCNN-KAN-Transformer architecture forradiographic bone-loss risk stratification fromknee X-rays Ahmed S. Shaban, Mohammed Tawfik, Islam Fathi, Ayman Myla This is a preprint; it has not been peer reviewed by a journal. https://doi.org/ 10.21203/rs.3.rs-9098416/v1 This work is licensed under a CC BY 4.0 License Status: Under Review Version 1 posted 12 You are reading this latest preprint version Abstract Osteoporosis is underdiagnosed because dual-energy X-ray absorptiometry (DXA) is costly and scarce. We present \textbf{MultiScaleKANNet}, a hybrid deep-learning architecture for radiographic bone-loss risk stratification from routine knee X-rays, combining convolutional feature learning, learnable nonlinear transformations via Kolmogorov--Arnold Network (KAN) layers, and Transformer-based multi-scale attention. We evaluated MultiScaleKANNet on 2,035 knee radiographs from four public Kaggle sources with three categories (Healthy, Osteopenia, Osteoporosis). The labels are \emph{proxy labels}---some derived from quantitative ultrasound T-scores rather than DXA---so results represent radiographic risk stratification, not clinical diagnosis. On a stratified held-out test set ($n{=}407$), the model achieved 97.30% accuracy (95% CI: 95.3--98.6%; Cohen's $\kappa{=}0.9584$; MCC${=}0.9585$). A source-held-out evaluation yielded 89.52% binary accuracy ($\kappa{=}0.7903$), suggesting in-distribution metrics may partly reflect dataset homogeneity. Ablation studies confirm synergistic gains from KAN layers (+2.46%), multi-scale processing (+4.17%), and Transformer attention (+4.91%), with 40% parameter reduction versus ResNet-18. This is a \emph{methodological feasibility study}; prospective DXA-confirmed validation is required. Health sciences/Diseases Health sciences/Health care Health sciences/Medical research Osteoporosis Kolmogorov–Arnold Network Transformer Knee X-ray Multi-scale feature extraction Deep learning Radiographic risk stratification Full Text Additional Declarations No competing interests reported. Cite Share Download PDF Status: Under Review Version 1 posted Editorial decision: Revision requested 13 Apr, 2026 Reviews received at journal 10 Apr, 2026 Reviewers agreed at journal 09 Apr, 2026 Reviews received at journal 08 Apr, 2026 Reviewers agreed at journal 07 Apr, 2026 Reviewers agreed at journal 05 Apr, 2026 Reviewers agreed at journal 02 Apr, 2026 Reviewers invited by journal 25 Mar, 2026 Editor assigned by journal 25 Mar, 2026 Editor invited by journal 24 Mar, 2026 Submission checks completed at journal 18 Mar, 2026 First submitted to journal 18 Mar, 2026 You are reading this latest preprint version Research Square lets you share your work early, gain feedback from the community, and start making changes to your manuscript prior to peer review in a journal. As a division of Research Square Company, we’re committed to making research communication faster, fairer, and more useful. We do this by developing innovative software and high quality services for the global research community. Our growing team is made up of researchers and industry professionals working together to solve the most critical problems facing scientific publishing. Also discoverable on Platform About Our Team In Review Editorial Policies Advisory Board Help Center Resources Author Services Accessibility API Access RSS feed Manage Cookie Preferences © Research Square 2026 | ISSN 2693-5015 (online) Privacy Policy Terms of Service Do Not Sell My Personal Information {"props":{"pageProps":{"initialData":{"identity":"rs-9098416","acceptedTermsAndConditions":true,"allowDirectSubmit":false,"archivedVersions":[],"articleType":"Article","associatedPublications":[],"authors":[{"id":612179235,"identity":"83d90cfc-097d-4a10-a5eb-3a070584a361","order_by":0,"name":"Ahmed S. Shaban","email":"","orcid":"","institution":"Irbid National University","correspondingAuthor":false,"prefix":"","firstName":"Ahmed","middleName":"S.","lastName":"Shaban","suffix":""},{"id":612179236,"identity":"5198958d-628d-414e-85f6-4099599b2d31","order_by":1,"name":"Mohammed Tawfik","email":"data:image/png;base64,iVBORw0KGgoAAAANSUhEUgAAAZAAAAAyAQMAAABI0h/eAAAABlBMVEX///8AAABVwtN+AAAACXBIWXMAAA7EAAAOxAGVKw4bAAAAvUlEQVRIiWNgGAWjYBACCQY2NgaGAgk5EOfAA+K1GFgYg7UkkKClIrEBxCNKi+S0Y2kPPhhIpM8PO/wQaIudnG4DAS3S0mnHDWcYSORuvJ1mANSSbGx2gIAWOen0NmkekJbZCSAtBxK3Easl3XB2+gfitAAddgykJUFeOodIWyRnp6VJAv1iuEE6p+BAggERfpG4nWYm8aGiTl5+dvrmDx8q7OQIaoEDA7BKA2KVg4B8AymqR8EoGAWjYEQBAKsgP2EX4Tw2AAAAAElFTkSuQmCC","orcid":"","institution":"Sana'a University","correspondingAuthor":true,"prefix":"","firstName":"Mohammed","middleName":"","lastName":"Tawfik","suffix":""},{"id":612179237,"identity":"abff7feb-3799-485c-83cc-f6de58b09878","order_by":2,"name":"Islam Fathi","email":"","orcid":"","institution":"Ajloun National University","correspondingAuthor":false,"prefix":"","firstName":"Islam","middleName":"","lastName":"Fathi","suffix":""},{"id":612179238,"identity":"6473d566-5d4e-40f1-896a-d69866b0d5f2","order_by":3,"name":"Ayman Myla","email":"","orcid":"","institution":"Horus University","correspondingAuthor":false,"prefix":"","firstName":"Ayman","middleName":"","lastName":"Myla","suffix":""}],"badges":[],"createdAt":"2026-03-11 22:53:43","currentVersionCode":1,"declarations":"","doi":"10.21203/rs.3.rs-9098416/v1","doiUrl":"https://doi.org/10.21203/rs.3.rs-9098416/v1","draftVersion":[],"editorialEvents":[],"editorialNote":"","failedWorkflow":false,"files":[{"id":105566523,"identity":"ad522b9b-0692-4df4-96d2-25b4ea9809af","added_by":"auto","created_at":"2026-03-27 12:56:37","extension":"pdf","order_by":1,"title":"","display":"","copyAsset":false,"role":"manuscript-pdf","size":1077630,"visible":true,"origin":"","legend":"","description":"","filename":"Ahmedsv2.pdf","url":"https://assets-eu.researchsquare.com/files/rs-9098416/v1_covered_bc709043-f28d-4330-bacd-7dcc834e6cbc.pdf"}],"financialInterests":"No competing interests reported.","formattedTitle":"MultiScaleKANNet: a hybridCNN-KAN-Transformer architecture forradiographic bone-loss risk stratification fromknee X-rays","fulltext":[],"fulltextSource":"","fullText":"","funders":[],"hasAdminPriorityOnWorkflow":false,"hasManuscriptDocX":false,"hasOptedInToPreprint":true,"hasPassedJournalQc":"","hasAnyPriority":false,"hideJournal":false,"highlight":"","institution":"","isAcceptedByJournal":true,"isAuthorSuppliedPdf":true,"isDeskRejected":"","isHiddenFromSearch":false,"isInQc":false,"isInWorkflow":false,"isPdf":true,"isPdfUpToDate":true,"isWithdrawnOrRetracted":false,"journal":{"display":true,"email":"[email protected]","identity":"scientific-reports","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":false,"externalIdentity":"scirep","sideBox":"Learn more about [Scientific Reports](http://www.nature.com/srep/)","snPcode":"","submissionUrl":"","title":"Scientific Reports","twitterHandle":"","acdcEnabled":true,"dfaEnabled":true,"editorialSystem":"stoa","reportingPortfolio":"Scientific Reports","inReviewEnabled":true,"inReviewRevisionsEnabled":true},"keywords":"Osteoporosis, Kolmogorov–Arnold Network, Transformer, Knee X-ray, Multi-scale feature extraction, Deep learning, Radiographic risk stratification","lastPublishedDoi":"10.21203/rs.3.rs-9098416/v1","lastPublishedDoiUrl":"https://doi.org/10.21203/rs.3.rs-9098416/v1","license":{"name":"CC BY 4.0","url":"https://creativecommons.org/licenses/by/4.0/"},"manuscriptAbstract":"Osteoporosis is underdiagnosed because dual-energy X-ray absorptiometry (DXA) is costly and scarce. We present \\textbf{MultiScaleKANNet}, a hybrid deep-learning architecture for radiographic bone-loss risk stratification from routine knee X-rays, combining convolutional feature learning, learnable nonlinear transformations via Kolmogorov--Arnold Network (KAN) layers, and Transformer-based multi-scale attention. We evaluated MultiScaleKANNet on 2,035 knee radiographs from four public Kaggle sources with three categories (Healthy, Osteopenia, Osteoporosis). The labels are \\emph{proxy labels}---some derived from quantitative ultrasound T-scores rather than DXA---so results represent radiographic risk stratification, not clinical diagnosis. On a stratified held-out test set ($n{=}407$), the model achieved 97.30\\% accuracy (95\\% CI: 95.3--98.6\\%; Cohen's $\\kappa{=}0.9584$; MCC${=}0.9585$). A source-held-out evaluation yielded 89.52\\% binary accuracy ($\\kappa{=}0.7903$), suggesting in-distribution metrics may partly reflect dataset homogeneity. Ablation studies confirm synergistic gains from KAN layers (+2.46\\%), multi-scale processing (+4.17\\%), and Transformer attention (+4.91\\%), with 40\\% parameter reduction versus ResNet-18. This is a \\emph{methodological feasibility study}; prospective DXA-confirmed validation is required.","manuscriptTitle":"MultiScaleKANNet: a hybridCNN-KAN-Transformer architecture forradiographic bone-loss risk stratification fromknee X-rays","msid":"","msnumber":"","nonDraftVersions":[{"code":1,"date":"2026-03-26 14:44:28","doi":"10.21203/rs.3.rs-9098416/v1","editorialEvents":[{"type":"communityComments","content":0},{"type":"decision","content":"Revision requested","date":"2026-04-13T07:56:05+00:00","index":"","fulltext":""},{"type":"editorInvitedReview","content":"","date":"2026-04-10T18:09:06+00:00","index":"hide","fulltext":""},{"type":"reviewerAgreed","content":"294861839246892183304955142613360660378","date":"2026-04-09T04:46:20+00:00","index":"hide","fulltext":""},{"type":"editorInvitedReview","content":"","date":"2026-04-09T02:17:05+00:00","index":"hide","fulltext":""},{"type":"reviewerAgreed","content":"65397863411811806949674959343512990385","date":"2026-04-07T08:43:37+00:00","index":"hide","fulltext":""},{"type":"reviewerAgreed","content":"324282953271666330320478638633033408390","date":"2026-04-06T02:03:47+00:00","index":"hide","fulltext":""},{"type":"reviewerAgreed","content":"13795557391609427460077479879253958","date":"2026-04-02T11:45:27+00:00","index":"hide","fulltext":""},{"type":"reviewersInvited","content":"","date":"2026-03-25T07:38:41+00:00","index":"","fulltext":""},{"type":"editorAssigned","content":"","date":"2026-03-25T07:36:44+00:00","index":"","fulltext":""},{"type":"editorInvited","content":"","date":"2026-03-24T07:10:19+00:00","index":"","fulltext":""},{"type":"checksComplete","content":"","date":"2026-03-18T23:50:31+00:00","index":"","fulltext":""},{"type":"submitted","content":"Scientific Reports","date":"2026-03-18T15:30:21+00:00","index":"","fulltext":""}],"status":"published","journal":{"display":true,"email":"[email protected]","identity":"scientific-reports","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":false,"externalIdentity":"scirep","sideBox":"Learn more about [Scientific Reports](http://www.nature.com/srep/)","snPcode":"","submissionUrl":"","title":"Scientific Reports","twitterHandle":"","acdcEnabled":true,"dfaEnabled":true,"editorialSystem":"stoa","reportingPortfolio":"Scientific Reports","inReviewEnabled":true,"inReviewRevisionsEnabled":true}}],"origin":"","ownerIdentity":"f64c7e3a-44e5-458a-aef8-552c80d0ad9d","owner":[],"postedDate":"March 26th, 2026","published":true,"recentEditorialEvents":[],"rejectedJournal":[],"revision":"","amendment":"","status":"under-review","subjectAreas":[{"id":65139953,"name":"Health sciences/Diseases"},{"id":65139954,"name":"Health sciences/Health care"},{"id":65139955,"name":"Health sciences/Medical research"}],"tags":[],"updatedAt":"2026-04-20T06:40:51+00:00","versionOfRecord":[],"versionCreatedAt":"2026-03-26 14:44:28","video":"","vorDoi":"","vorDoiUrl":"","workflowStages":[]},"version":"v1","identity":"rs-9098416","journalConfig":"researchsquare"},"__N_SSP":true},"page":"/article/[identity]/[[...version]]","query":{"redirect":"/article/rs-9098416","identity":"rs-9098416","version":["v1"]},"buildId":"XKTyCvWXoU3ODBz1xrDgd","isFallback":false,"isExperimentalCompile":false,"dynamicIds":[84888],"gssp":true,"scriptLoader":[]}