Shortcut Learning in Deep Neural Networks for Knee Osteoarthritis Grading: Detection, Characterization, and Mitigation via Gradient Reversal and Geometric Augmentation

Shortcut Learning in Deep Neural Networks for Knee Osteoarthritis Grading: Detection, Characterization, and Mitigation via Gradient Reversal and Geometric Augmentation | 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 Research Article Shortcut Learning in Deep Neural Networks for Knee Osteoarthritis Grading: Detection, Characterization, and Mitigation via Gradient Reversal and Geometric Augmentation Shah Dhruv, Shubham Vaghasiya, Vivek Viroja This is a preprint; it has not been peer reviewed by a journal. https://doi.org/ 10.21203/rs.3.rs-9495057/v1 This work is licensed under a CC BY 4.0 License Status: Posted Version 1 posted You are reading this latest preprint version Abstract Deep learning models for automated knee osteoarthritis (OA) grading achieve high in-distribution accuracy but frequently exploit spurious image-acquisition cues — a phenomenon termed shortcut learning — rather than learning disease-relevant anatomical features. We present a systematic three-phase experimental framework to detect, characterise, and mitigate shortcut learning in Kellgren–Lawrence (KL) grade classification using a disentangled ResNet-18 architecture extended with a Gradient Reversal Layer (GRL). Phase 1 (baseline) achieves a macro one-vs-rest AUC of 0.989 while simultaneously exhibiting a domain sensitivity probe AUC of 0.887, corresponding to an estimated mutual information I(Z; D) of 1.04 nats between backbone features Z and scanner-site domain labels D. Phase 2 (fisheye geometric augmentation) disrupts spatial shortcuts, reducing disease AUC by 17.6 pp and domain sensitivity AUC by 26.6 pp. Phase 3 (hybrid GRL, λ = 1.0) recovers disease AUC to 0.847 while suppressing domain sensitivity AUC to 0.541 and reducing I(Z; D) to approximately 0.09 nats. Integrated Gradients attribution analysis — verified against the completeness and sensitivity axioms — confirms an anatomical shift in model focus from non-diagnostic image borders to the medial tibio-femoral joint space. An information-theoretic decomposition of feature variance shows that the GRL reduces domain-explained variance from 44% to 13% of total representation capacity. Collectively, these findings establish that benchmark AUC alone is insufficient for validating clinical AI, and provide a mathematically grounded framework for shortcut characterisation and adversarial mitigation in musculoskeletal diagnostics. shortcut learning knee osteoarthritis Kellgren–Lawrence grading gradient reversal layer domain generalization mutual information Integrated Gradients adversarial disentanglement Full Text Additional Declarations The authors declare potential competing interests as follows: Cite Share Download PDF Status: Posted Version 1 posted 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-9495057","acceptedTermsAndConditions":true,"allowDirectSubmit":true,"archivedVersions":[],"articleType":"Research Article","associatedPublications":[],"authors":[{"id":627686946,"identity":"1f9a9a83-a977-4f8a-bde8-5155fce7c8aa","order_by":0,"name":"Shah Dhruv","email":"data:image/png;base64,iVBORw0KGgoAAAANSUhEUgAAAZAAAAAyAQMAAABI0h/eAAAABlBMVEX///8AAABVwtN+AAAACXBIWXMAAA7EAAAOxAGVKw4bAAAA7UlEQVRIiWNgGAWjYDCCGyCi4p8cY3tD4gMgk4ePOC1nDhgz9xx4bADSwkaUFsa2A4ntMxKfSYAECGrhu937TJq37U5i74zktMqvOXYybAzMDx/dwKNF8s5xM2mec8+MZ/Y8S7stuy0Z6DA2Y+McPFoMbqSxSc4oY5bd2J6TdltyGzNQCw+bNGEtbMyM+w/kfyuW3FZPnBaJD22HFRs7EtIYP247TFiL5I00ZosPZ9KMGXsOJEszbjvOw8ZMwC98N9IYbyRU2ICj8uPPbdX2/OzNDx/j04ICmHnAJLHKQYDxBymqR8EoGAWjYMQAAAHOTsqEFeseAAAAAElFTkSuQmCC","orcid":"","institution":"Parul University","correspondingAuthor":true,"prefix":"","firstName":"Shah","middleName":"","lastName":"Dhruv","suffix":""},{"id":627686947,"identity":"45438062-eda6-4105-a22a-e5711ee13e94","order_by":1,"name":"Shubham Vaghasiya","email":"data:image/png;base64,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","orcid":"","institution":"Parul University","correspondingAuthor":true,"prefix":"","firstName":"Shubham","middleName":"","lastName":"Vaghasiya","suffix":""},{"id":627686948,"identity":"12ede13e-2f33-4d28-b01a-39d8f1058507","order_by":2,"name":"Vivek Viroja","email":"data:image/png;base64,iVBORw0KGgoAAAANSUhEUgAAAZAAAAAyAQMAAABI0h/eAAAABlBMVEX///8AAABVwtN+AAAACXBIWXMAAA7EAAAOxAGVKw4bAAABDklEQVRIiWNgGAWjYDACZgYGxgYGCSDrAPPDDz8k5ECCBx4QqYXNWLLHxhisJYGARUAtECDBw5aWCObg02JwnMfw4cw2i3x+xsMPDCR4DqfPDzv8EGiLnZxuAw4th3mMDTe2SVjObDhm8KDA4nDuxttpBkAtycZmB7BrkWxmS5N8cEbCwODAAQOQLbkbZyeAtBxI3IZbS/pPiJbjH4B+OZxuODv9A14t/MzMxxg3VIC0nDEAeT9BXjoHvy1ALYclZwC1SDacKQMFsuEG6ZyCAwkGuP3Cxn+w8WOPQZ0Bv8TxzaColJefnb75w4cKOzlcWhBAAqrCAEwbEFIOdmIDhJZvIEb1KBgFo2AUjCQAAFu9YtdSJnAsAAAAAElFTkSuQmCC","orcid":"","institution":"Parul University","correspondingAuthor":true,"prefix":"","firstName":"Vivek","middleName":"","lastName":"Viroja","suffix":""}],"badges":[],"createdAt":"2026-04-22 10:47:40","currentVersionCode":1,"declarations":{"humanSubjects":false,"vertebrateSubjects":false,"conflictsOfInterestStatement":true,"humanSubjectEthicalGuidelines":false,"humanSubjectConsent":false,"humanSubjectClinicalTrial":false,"humanSubjectCaseReport":false,"vertebrateSubjectEthicalGuidelines":false},"doi":"10.21203/rs.3.rs-9495057/v1","doiUrl":"https://doi.org/10.21203/rs.3.rs-9495057/v1","draftVersion":[],"editorialEvents":[],"editorialNote":"","failedWorkflow":false,"files":[{"id":107706525,"identity":"e6080f6c-f99a-4aad-8705-e7a7579fe90c","added_by":"auto","created_at":"2026-04-24 09:18:16","extension":"pdf","order_by":1,"title":"","display":"","copyAsset":false,"role":"manuscript-pdf","size":1218621,"visible":true,"origin":"","legend":"","description":"","filename":"ShortcutLearninginDeepNeuralNetworksforKneeOsteoarthritis.pdf","url":"https://assets-eu.researchsquare.com/files/rs-9495057/v1_covered_0dbc9d57-39d0-441d-8a48-2341935da812.pdf"}],"financialInterests":"The authors declare potential competing interests as follows: ","formattedTitle":"\u003cp\u003eShortcut Learning in Deep Neural Networks for Knee Osteoarthritis Grading: Detection, Characterization, and Mitigation via Gradient Reversal and Geometric Augmentation\u003c/p\u003e","fulltext":[],"fulltextSource":"","fullText":"","funders":[],"hasAdminPriorityOnWorkflow":false,"hasManuscriptDocX":false,"hasOptedInToPreprint":true,"hasPassedJournalQc":"","hasAnyPriority":true,"hideJournal":true,"highlight":"","institution":"Parul University","isAcceptedByJournal":false,"isAuthorSuppliedPdf":true,"isDeskRejected":"","isHiddenFromSearch":false,"isInQc":false,"isInWorkflow":false,"isPdf":true,"isPdfUpToDate":true,"isWithdrawnOrRetracted":false,"journal":{"display":true,"email":"[email protected]","identity":"researchsquare","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":true,"externalIdentity":"","sideBox":"","snPcode":"","submissionUrl":"/submission","title":"Research Square","twitterHandle":"researchsquare","acdcEnabled":true,"dfaEnabled":false,"editorialSystem":"","reportingPortfolio":"","inReviewEnabled":false,"inReviewRevisionsEnabled":true},"keywords":"shortcut learning, knee osteoarthritis, Kellgren–Lawrence grading, gradient reversal layer, domain generalization, mutual information, Integrated Gradients, adversarial disentanglement","lastPublishedDoi":"10.21203/rs.3.rs-9495057/v1","lastPublishedDoiUrl":"https://doi.org/10.21203/rs.3.rs-9495057/v1","license":{"name":"CC BY 4.0","url":"https://creativecommons.org/licenses/by/4.0/"},"manuscriptAbstract":"\u003cp\u003eDeep learning models for automated knee osteoarthritis (OA) grading achieve high in-distribution accuracy but frequently exploit spurious image-acquisition cues — a phenomenon termed shortcut learning — rather than learning disease-relevant anatomical features. We present a systematic three-phase experimental framework to detect, characterise, and mitigate shortcut learning in Kellgren–Lawrence (KL) grade classification using a disentangled ResNet-18 architecture extended with a Gradient Reversal Layer (GRL). Phase 1 (baseline) achieves a macro one-vs-rest AUC of 0.989 while simultaneously exhibiting a domain sensitivity probe AUC of 0.887, corresponding to an estimated mutual information I(Z; D) of 1.04 nats between backbone features Z and scanner-site domain labels D. Phase 2 (fisheye geometric augmentation) disrupts spatial shortcuts, reducing disease AUC by 17.6 pp and domain sensitivity AUC by 26.6 pp. Phase 3 (hybrid GRL, λ = 1.0) recovers disease AUC to 0.847 while suppressing domain sensitivity AUC to 0.541 and reducing I(Z; D) to approximately 0.09 nats. Integrated Gradients attribution analysis — verified against the completeness and sensitivity axioms — confirms an anatomical shift in model focus from non-diagnostic image borders to the medial tibio-femoral joint space. An information-theoretic decomposition of feature variance shows that the GRL reduces domain-explained variance from 44% to 13% of total representation capacity. Collectively, these findings establish that benchmark AUC alone is insufficient for validating clinical AI, and provide a mathematically grounded framework for shortcut characterisation and adversarial mitigation in musculoskeletal diagnostics.\u003c/p\u003e","manuscriptTitle":"Shortcut Learning in Deep Neural Networks for Knee Osteoarthritis Grading: Detection, Characterization, and Mitigation via Gradient Reversal and Geometric Augmentation","msid":"","msnumber":"","nonDraftVersions":[{"code":1,"date":"2026-04-23 02:41:11","doi":"10.21203/rs.3.rs-9495057/v1","editorialEvents":[{"type":"communityComments","content":0}],"status":"published","journal":{"display":true,"email":"[email protected]","identity":"researchsquare","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":true,"externalIdentity":"","sideBox":"","snPcode":"","submissionUrl":"/submission","title":"Research Square","twitterHandle":"researchsquare","acdcEnabled":true,"dfaEnabled":false,"editorialSystem":"","reportingPortfolio":"","inReviewEnabled":false,"inReviewRevisionsEnabled":true}}],"origin":"","ownerIdentity":"0c2015ff-0f48-4c06-8da2-417edf979a58","owner":[],"postedDate":"April 23rd, 2026","published":true,"recentEditorialEvents":[],"rejectedJournal":[],"revision":"","amendment":"","status":"posted","subjectAreas":[],"tags":[],"updatedAt":"2026-04-23T02:41:11+00:00","versionOfRecord":[],"versionCreatedAt":"2026-04-23 02:41:11","video":"","vorDoi":"","vorDoiUrl":"","workflowStages":[]},"version":"v1","identity":"rs-9495057","journalConfig":"researchsquare"},"__N_SSP":true},"page":"/article/[identity]/[[...version]]","query":{"redirect":"/article/rs-9495057","identity":"rs-9495057","version":["v1"]},"buildId":"XKTyCvWXoU3ODBz1xrDgd","isFallback":false,"isExperimentalCompile":false,"dynamicIds":[84888],"gssp":true,"scriptLoader":[]}