Synthetic MRI Pretraining for Medical Imaging Tasks

Synthetic MRI Pretraining for Medical Imaging Tasks | 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Article Synthetic MRI Pretraining for Medical Imaging Tasks Rosanna Turrisi, Giuseppe Patanè This is a preprint; it has not been peer reviewed by a journal. https://doi.org/ 10.21203/rs.3.rs-9199202/v1 This work is licensed under a CC BY 4.0 License Status: Under Revision Version 1 posted 12 You are reading this latest preprint version Abstract Deep learning (DL) models have reached remarkable achievements in medical imaging, but their performance heavily depends on the availability of large and diverse datasets. To overcome this, transfer learning has emerged as a widely adopted solution, where models pretrained on large datasets are fine-tuned for specific medical tasks. Due to the scarcity of large-scale medical imaging datasets, most existing models are pretrained on natural image datasets such as ImageNet, while recent studies have explored high complex models trained on a vast amount of diverse medical unlabelled datasets. In this work, we generate and leverage a dataset of 9,000 MRI scans to pretrain DL architectures, demonstrating effective model learning with minimal computational cost and robustness to pretraining dataset size. We evaluate our approach across multiple clinical tasks, including brain tumour classification and ten benchmark tasks from MedMNIST covering 2D and 3D modalities. Compared with ImageNet-pretrained, foundation, and self-supervised models, synthetic pretraining consistently improves feature representations and downstream performance. Our results show that models pretrained on synthetic medical data outperform both ImageNet-pretrained models and those trained from scratch. Overall, our approach outperforms competing methods, establishing a new state-of-the-art on the MedMNIST benchmark. Biological sciences/Computational biology and bioinformatics Health sciences/Health care Physical sciences/Mathematics and computing Health sciences/Medical research Deep Learning Synthetic MRI Pretrained Models Transfer Learning Medical Image Classification Full Text Additional Declarations No competing interests reported. Cite Share Download PDF Status: Under Revision Version 1 posted Editorial decision: Revision requested 16 Apr, 2026 Reviews received at journal 15 Apr, 2026 Reviews received at journal 09 Apr, 2026 Reviewers agreed at journal 09 Apr, 2026 Reviewers agreed at journal 07 Apr, 2026 Reviews received at journal 02 Apr, 2026 Reviewers agreed at journal 01 Apr, 2026 Reviewers invited by journal 31 Mar, 2026 Editor invited by journal 30 Mar, 2026 Editor assigned by journal 24 Mar, 2026 Submission checks completed at journal 24 Mar, 2026 First submitted to journal 23 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-9199202","acceptedTermsAndConditions":true,"allowDirectSubmit":false,"archivedVersions":[],"articleType":"Article","associatedPublications":[],"authors":[{"id":616773308,"identity":"d0e48b79-476c-48f7-9215-818b0a173a55","order_by":0,"name":"Rosanna Turrisi","email":"data:image/png;base64,iVBORw0KGgoAAAANSUhEUgAAAZAAAAAyAQMAAABI0h/eAAAABlBMVEX///8AAABVwtN+AAAACXBIWXMAAA7EAAAOxAGVKw4bAAAA10lEQVRIiWNgGAWjYNCDBAYJOSiDCMAD1WIM0UKMHh4ondhAyBrz9vaHnwt32DHYS+Q++/Dgj0V6/+zeBwwPf+DWInPmjLH0zDPJDDwS6cYzEtskcmfcOW6A12ESEjkM0rxtzEAtacxAV0nkbpBIw+8XCYn0x7952+ohWhL+SKQbENaSYAa05TBUC5tEAmEtPGfMrHnbjvPwnHkGdFibhOGMG2kMBxLS8Ghhb398m7etWo69PY2Z8cefOnn+GWmMD3/Y4NYCAzwovAOENYyCUTAKRsEowAcAY7JBLu2BIqAAAAAASUVORK5CYII=","orcid":"","institution":"National Research Council","correspondingAuthor":true,"prefix":"","firstName":"Rosanna","middleName":"","lastName":"Turrisi","suffix":""},{"id":616773309,"identity":"dfaab033-6540-491e-9923-b82210a52c26","order_by":1,"name":"Giuseppe Patanè","email":"","orcid":"","institution":"National Research Council","correspondingAuthor":false,"prefix":"","firstName":"Giuseppe","middleName":"","lastName":"Patanè","suffix":""}],"badges":[],"createdAt":"2026-03-23 10:28:05","currentVersionCode":1,"declarations":"","doi":"10.21203/rs.3.rs-9199202/v1","doiUrl":"https://doi.org/10.21203/rs.3.rs-9199202/v1","draftVersion":[],"editorialEvents":[],"editorialNote":"","failedWorkflow":false,"files":[{"id":106403541,"identity":"1070f2e8-d947-4b21-b35c-d93500141ae7","added_by":"auto","created_at":"2026-04-08 09:14:28","extension":"pdf","order_by":1,"title":"","display":"","copyAsset":false,"role":"manuscript-pdf","size":794311,"visible":true,"origin":"","legend":"","description":"","filename":"turrisipatane.pdf","url":"https://assets-eu.researchsquare.com/files/rs-9199202/v1_covered_0e79d785-8436-4c45-8608-9cc78aacd64f.pdf"}],"financialInterests":"No competing interests reported.","formattedTitle":"Synthetic MRI Pretraining for Medical Imaging Tasks","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":"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":"Deep Learning, Synthetic MRI, Pretrained Models, Transfer Learning, Medical Image Classification","lastPublishedDoi":"10.21203/rs.3.rs-9199202/v1","lastPublishedDoiUrl":"https://doi.org/10.21203/rs.3.rs-9199202/v1","license":{"name":"CC BY 4.0","url":"https://creativecommons.org/licenses/by/4.0/"},"manuscriptAbstract":"Deep learning (DL) models have reached remarkable achievements in medical imaging, but their performance heavily depends on the availability of large and diverse datasets. To overcome this, transfer learning has emerged as a widely adopted solution, where models pretrained on large datasets are fine-tuned for specific medical tasks. Due to the scarcity of large-scale medical imaging datasets, most existing models are pretrained on natural image datasets such as ImageNet, while recent studies have explored high complex models trained on a vast amount of diverse medical unlabelled datasets. In this work, we generate and leverage a dataset of 9,000 MRI scans to pretrain DL architectures, demonstrating effective model learning with minimal computational cost and robustness to pretraining dataset size. We evaluate our approach across multiple clinical tasks, including brain tumour classification and ten benchmark tasks from MedMNIST covering 2D and 3D modalities. Compared with ImageNet-pretrained, foundation, and self-supervised models, synthetic pretraining consistently improves feature representations and downstream performance. Our results show that models pretrained on synthetic medical data outperform both ImageNet-pretrained models and those trained from scratch. Overall, our approach outperforms competing methods, establishing a new state-of-the-art on the MedMNIST benchmark.","manuscriptTitle":"Synthetic MRI Pretraining for Medical Imaging Tasks","msid":"","msnumber":"","nonDraftVersions":[{"code":1,"date":"2026-04-06 20:42:06","doi":"10.21203/rs.3.rs-9199202/v1","editorialEvents":[{"type":"communityComments","content":0},{"type":"decision","content":"Revision requested","date":"2026-04-17T03:21:01+00:00","index":"","fulltext":""},{"type":"editorInvitedReview","content":"","date":"2026-04-15T21:26:18+00:00","index":"hide","fulltext":""},{"type":"editorInvitedReview","content":"","date":"2026-04-09T06:30:34+00:00","index":"hide","fulltext":""},{"type":"reviewerAgreed","content":"259251580947532640674160644841766186311","date":"2026-04-09T06:22:46+00:00","index":"hide","fulltext":""},{"type":"reviewerAgreed","content":"42340353622939765942579642531981017677","date":"2026-04-07T13:11:55+00:00","index":"hide","fulltext":""},{"type":"editorInvitedReview","content":"","date":"2026-04-02T14:57:33+00:00","index":"hide","fulltext":""},{"type":"reviewerAgreed","content":"249278359052706080500721277923072853996","date":"2026-04-01T08:06:40+00:00","index":"hide","fulltext":""},{"type":"reviewersInvited","content":"","date":"2026-03-31T16:44:00+00:00","index":"","fulltext":""},{"type":"editorInvited","content":"","date":"2026-03-30T16:12:52+00:00","index":"","fulltext":""},{"type":"editorAssigned","content":"","date":"2026-03-24T14:13:18+00:00","index":"","fulltext":""},{"type":"checksComplete","content":"","date":"2026-03-24T14:12:34+00:00","index":"","fulltext":""},{"type":"submitted","content":"Scientific Reports","date":"2026-03-23T10:12:45+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":"55a92902-1aec-43ef-9321-cb29e116a6e3","owner":[],"postedDate":"April 6th, 2026","published":true,"recentEditorialEvents":[],"rejectedJournal":[],"revision":"","amendment":"","status":"in-revision","subjectAreas":[{"id":65635623,"name":"Biological sciences/Computational biology and bioinformatics"},{"id":65635624,"name":"Health sciences/Health care"},{"id":65635625,"name":"Physical sciences/Mathematics and computing"},{"id":65635626,"name":"Health sciences/Medical research"}],"tags":[],"updatedAt":"2026-04-17T03:24:26+00:00","versionOfRecord":[],"versionCreatedAt":"2026-04-06 20:42:06","video":"","vorDoi":"","vorDoiUrl":"","workflowStages":[]},"version":"v1","identity":"rs-9199202","journalConfig":"researchsquare"},"__N_SSP":true},"page":"/article/[identity]/[[...version]]","query":{"redirect":"/article/rs-9199202","identity":"rs-9199202","version":["v1"]},"buildId":"XKTyCvWXoU3ODBz1xrDgd","isFallback":false,"isExperimentalCompile":false,"dynamicIds":[84888],"gssp":true,"scriptLoader":[]}