A Cortically-Inspired Predictive Coding Framework for Polyp Segmentation

A Cortically-Inspired Predictive Coding Framework for Polyp Segmentation | 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 A Cortically-Inspired Predictive Coding Framework for Polyp Segmentation Abdul Joseph Fofanah, Alpha Alimamy Kamara, Lian Wen, David Chen, and 1 more This is a preprint; it has not been peer reviewed by a journal. https://doi.org/ 10.21203/rs.3.rs-9145958/v1 This work is licensed under a CC BY 4.0 License Status: Under Review Version 1 posted 10 You are reading this latest preprint version Abstract Polyp segmentation in colonoscopy remains clinically challenging because existing deep learning models operate as feedforward systems: features are extracted, attention weights are computed, and predictions are generated in a single irreversible pass with no opportunity for revision. We propose GRAFNet, a segmentation framework that introduces three biologically inspired computational mechanisms absent from medical image analysis: (i) learnable orientation filters with centre-surround normalisation that adapt to diagnostically relevant edge structures, (ii) parallel retinal pathways with lateral inhibition for competitive feature suppression across scales, and (iii) predictive coding with inference-time feedback, which iteratively refines predictions by reconciling high-level hypotheses with low-level observations. These mechanisms are integrated through a closed-loop architecture where cortical feedback generates top-down predictions, computes prediction errors, and updates feature representations during inference, which enables the network to re-examine ambiguous regions dynamically. Across five polyp segmentation benchmarks, GRAFNet achieves state-of-the-art performance (0.929 Dice on CVC-ClinicDB, 0.915 on Kvasir-SEG, and 0.946 on CVC-300), with substantial gains on flat lesions (+9.8% over the best competitor) and reduced misclassification of challenging anatomical structures such as haustral folds (-4.8%). Beyond empirical improvements, the framework also provides interpretable insights into how recurrent feedback and competitive interactions improve segmentation under ambiguity. Polyp Segmentation Deep Learning Multiscale Attention Retinal-Cortical Feedback Endoscopic Image Analysis Full Text Additional Declarations No competing interests reported. Cite Share Download PDF Status: Under Review Version 1 posted Editorial decision: Revision requested 27 Apr, 2026 Reviews received at journal 22 Apr, 2026 Reviews received at journal 08 Apr, 2026 Reviewers agreed at journal 05 Apr, 2026 Reviewers agreed at journal 03 Apr, 2026 Reviewers invited by journal 02 Apr, 2026 Editor assigned by journal 02 Apr, 2026 Editor invited by journal 29 Mar, 2026 Submission checks completed at journal 27 Mar, 2026 First submitted to journal 27 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-9145958","acceptedTermsAndConditions":true,"allowDirectSubmit":false,"archivedVersions":[],"articleType":"Research Article","associatedPublications":[],"authors":[{"id":617763695,"identity":"23fc7086-839b-442a-8769-d426be16005a","order_by":0,"name":"Abdul Joseph Fofanah","email":"data:image/png;base64,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","orcid":"","institution":"Griffith University","correspondingAuthor":true,"prefix":"","firstName":"Abdul","middleName":"Joseph","lastName":"Fofanah","suffix":""},{"id":617763698,"identity":"1f1bc622-45b1-4ad0-b54a-a6e04e5f514c","order_by":1,"name":"Alpha Alimamy Kamara","email":"","orcid":"","institution":"Central South University","correspondingAuthor":false,"prefix":"","firstName":"Alpha","middleName":"Alimamy","lastName":"Kamara","suffix":""},{"id":617763699,"identity":"0d1fd72a-8183-4db3-afa0-91c629514027","order_by":2,"name":"Lian Wen","email":"","orcid":"","institution":"Griffith University","correspondingAuthor":false,"prefix":"","firstName":"Lian","middleName":"","lastName":"Wen","suffix":""},{"id":617763700,"identity":"36c0775e-8719-437c-935e-e49b9610ed52","order_by":3,"name":"David Chen","email":"","orcid":"","institution":"Griffith University","correspondingAuthor":false,"prefix":"","firstName":"David","middleName":"","lastName":"Chen","suffix":""},{"id":617763701,"identity":"af9999bb-871e-4f6c-95a5-40dee2ae21df","order_by":4,"name":"Zhongyi Zhang","email":"","orcid":"","institution":"Griffith University","correspondingAuthor":false,"prefix":"","firstName":"Zhongyi","middleName":"","lastName":"Zhang","suffix":""}],"badges":[],"createdAt":"2026-03-17 08:23:04","currentVersionCode":1,"declarations":"","doi":"10.21203/rs.3.rs-9145958/v1","doiUrl":"https://doi.org/10.21203/rs.3.rs-9145958/v1","draftVersion":[],"editorialEvents":[],"editorialNote":"","failedWorkflow":false,"files":[{"id":106724454,"identity":"66c932f0-df0c-4579-b7d7-31f1f25d719e","added_by":"auto","created_at":"2026-04-12 18:28:12","extension":"pdf","order_by":1,"title":"","display":"","copyAsset":false,"role":"manuscript-pdf","size":3262037,"visible":true,"origin":"","legend":"","description":"","filename":"GRAFNetSpringer.pdf","url":"https://assets-eu.researchsquare.com/files/rs-9145958/v1_covered_322f0b76-5eb4-491f-b080-c9e84148c03b.pdf"}],"financialInterests":"No competing interests reported.","formattedTitle":"A Cortically-Inspired Predictive Coding Framework for Polyp Segmentation","fulltext":[],"fulltextSource":"","fullText":"","funders":[],"hasAdminPriorityOnWorkflow":false,"hasManuscriptDocX":false,"hasOptedInToPreprint":true,"hasPassedJournalQc":"","hasAnyPriority":false,"hideJournal":false,"highlight":"","institution":"","isAcceptedByJournal":false,"isAuthorSuppliedPdf":true,"isDeskRejected":"","isHiddenFromSearch":false,"isInQc":false,"isInWorkflow":false,"isPdf":true,"isPdfUpToDate":true,"isWithdrawnOrRetracted":false,"journal":{"display":true,"email":"[email protected]","identity":"discover-artificial-intelligence","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":false,"externalIdentity":"diai","sideBox":"Learn more about [Discover Artificial Intelligence](https://www.springer.com/44163)","snPcode":"","submissionUrl":"","title":"Discover Artificial Intelligence","twitterHandle":"","acdcEnabled":true,"dfaEnabled":true,"editorialSystem":"stoa","reportingPortfolio":"Discover Series","inReviewEnabled":true,"inReviewRevisionsEnabled":true},"keywords":"Polyp Segmentation, Deep Learning, Multiscale Attention, Retinal-Cortical Feedback, Endoscopic Image Analysis","lastPublishedDoi":"10.21203/rs.3.rs-9145958/v1","lastPublishedDoiUrl":"https://doi.org/10.21203/rs.3.rs-9145958/v1","license":{"name":"CC BY 4.0","url":"https://creativecommons.org/licenses/by/4.0/"},"manuscriptAbstract":"Polyp segmentation in colonoscopy remains clinically challenging because existing deep learning models operate as feedforward systems: features are extracted, attention weights are computed, and predictions are generated in a single irreversible pass with no opportunity for revision. We propose GRAFNet, a segmentation framework that introduces three biologically inspired computational mechanisms absent from medical image analysis: (i) learnable orientation filters with centre-surround normalisation that adapt to diagnostically relevant edge structures, (ii) parallel retinal pathways with lateral inhibition for competitive feature suppression across scales, and (iii) predictive coding with inference-time feedback, which iteratively refines predictions by reconciling high-level hypotheses with low-level observations. These mechanisms are integrated through a closed-loop architecture where cortical feedback generates top-down predictions, computes prediction errors, and updates feature representations during inference, which enables the network to re-examine ambiguous regions dynamically. Across five polyp segmentation benchmarks, GRAFNet achieves state-of-the-art performance (0.929 Dice on CVC-ClinicDB, 0.915 on Kvasir-SEG, and 0.946 on CVC-300), with substantial gains on flat lesions (+9.8% over the best competitor) and reduced misclassification of challenging anatomical structures such as haustral folds (-4.8%). Beyond empirical improvements, the framework also provides interpretable insights into how recurrent feedback and competitive interactions improve segmentation under ambiguity.","manuscriptTitle":"A Cortically-Inspired Predictive Coding Framework for Polyp Segmentation","msid":"","msnumber":"","nonDraftVersions":[{"code":1,"date":"2026-04-08 22:14:06","doi":"10.21203/rs.3.rs-9145958/v1","editorialEvents":[{"type":"communityComments","content":0},{"type":"decision","content":"Revision requested","date":"2026-04-27T07:03:34+00:00","index":"","fulltext":""},{"type":"editorInvitedReview","content":"","date":"2026-04-22T09:47:26+00:00","index":"hide","fulltext":""},{"type":"editorInvitedReview","content":"","date":"2026-04-08T12:09:03+00:00","index":"hide","fulltext":""},{"type":"reviewerAgreed","content":"333527940252009697450471634625509316450","date":"2026-04-05T10:32:43+00:00","index":"hide","fulltext":""},{"type":"reviewerAgreed","content":"122713667658246301482559828908206074153","date":"2026-04-03T04:02:49+00:00","index":"hide","fulltext":""},{"type":"reviewersInvited","content":"","date":"2026-04-03T03:59:53+00:00","index":"","fulltext":""},{"type":"editorAssigned","content":"","date":"2026-04-03T03:02:23+00:00","index":"","fulltext":""},{"type":"editorInvited","content":"","date":"2026-03-29T09:30:46+00:00","index":"","fulltext":""},{"type":"checksComplete","content":"","date":"2026-03-28T03:31:17+00:00","index":"","fulltext":""},{"type":"submitted","content":"Discover Artificial Intelligence","date":"2026-03-28T03:28:23+00:00","index":"","fulltext":""}],"status":"published","journal":{"display":true,"email":"[email protected]","identity":"discover-artificial-intelligence","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":false,"externalIdentity":"diai","sideBox":"Learn more about [Discover Artificial Intelligence](https://www.springer.com/44163)","snPcode":"","submissionUrl":"","title":"Discover Artificial Intelligence","twitterHandle":"","acdcEnabled":true,"dfaEnabled":true,"editorialSystem":"stoa","reportingPortfolio":"Discover Series","inReviewEnabled":true,"inReviewRevisionsEnabled":true}}],"origin":"","ownerIdentity":"99e46034-bd12-4593-bdaf-519917738d64","owner":[],"postedDate":"April 8th, 2026","published":true,"recentEditorialEvents":[],"rejectedJournal":[],"revision":"","amendment":"","status":"under-review","subjectAreas":[],"tags":[],"updatedAt":"2026-05-07T13:24:20+00:00","versionOfRecord":[],"versionCreatedAt":"2026-04-08 22:14:06","video":"","vorDoi":"","vorDoiUrl":"","workflowStages":[]},"version":"v1","identity":"rs-9145958","journalConfig":"researchsquare"},"__N_SSP":true},"page":"/article/[identity]/[[...version]]","query":{"redirect":"/article/rs-9145958","identity":"rs-9145958","version":["v1"]},"buildId":"XKTyCvWXoU3ODBz1xrDgd","isFallback":false,"isExperimentalCompile":false,"dynamicIds":[84888],"gssp":true,"scriptLoader":[]}