Benchmarking Imputation Strategies for Missing Time-Series Data in Critical Care Using Real-World-Inspired Scenarios

Benchmarking Imputation Strategies for Missing Time-Series Data in Critical Care Using Real-World-Inspired Scenarios | 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 Benchmarking Imputation Strategies for Missing Time-Series Data in Critical Care Using Real-World-Inspired Scenarios Michael Poette, Sandrine Mouysset, Daniel Ruiz, Jean-Marc Alliot, and 2 more This is a preprint; it has not been peer reviewed by a journal. https://doi.org/ 10.21203/rs.3.rs-6973012/v1 This work is licensed under a CC BY 4.0 License Status: Published Journal Publication published 10 Feb, 2026 Read the published version in Scientific Reports → Version 1 posted 10 You are reading this latest preprint version Abstract Handling missing data remains a central issue in ICU time-series analysis, where data gaps often stem from non-random factors such as sensor disconnections or clinical workflows. In this study, we systematically benchmarked several imputation strategies using monitoring data from MIMIC-IV and designed masking scenarios that reflect common ICU patterns, including random dropouts, temporary monitoring interruptions, and sensor-specific failures. We compared simple statistical approaches (mean, LOCF, interpolation), classical machine learning techniques (MICE, MissForest), and deep learning models (SAITS, BRITS, US-GAN, GP-VAE). SAITS, based on Transformer architecture, achieved the best performance in most settings. However, linear interpolation—despite its simplicity—yielded robust estimates in short univariate gaps and occasionally performed comparably to neural models. Our findings suggest that while deep learning methods improve overall imputation accuracy, simpler and more interpretable approaches may be sufficient for many ICU applications. This work introduces a practical framework for evaluating time-series imputation strategies under realistic constraints, with a focus on clinical relevance and operational deployability. Physical sciences/Mathematics and computing/Computer science Biological sciences/Computational biology and bioinformatics/Data processing Biological sciences/Computational biology and bioinformatics/Machine learning Full Text Additional Declarations No competing interests reported. Cite Share Download PDF Status: Published Journal Publication published 10 Feb, 2026 Read the published version in Scientific Reports → Version 1 posted Editorial decision: Revision requested 16 Oct, 2025 Reviews received at journal 26 Sep, 2025 Reviewers agreed at journal 12 Sep, 2025 Reviews received at journal 10 Aug, 2025 Reviewers agreed at journal 21 Jul, 2025 Reviewers invited by journal 20 Jul, 2025 Editor invited by journal 09 Jul, 2025 Editor assigned by journal 30 Jun, 2025 Submission checks completed at journal 27 Jun, 2025 First submitted to journal 25 Jun, 2025 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-6973012","acceptedTermsAndConditions":true,"allowDirectSubmit":false,"archivedVersions":[],"articleType":"Article","associatedPublications":[],"authors":[{"id":479152788,"identity":"9a25779f-ab71-4447-bdcf-c7dde6784fc5","order_by":0,"name":"Michael Poette","email":"data:image/png;base64,iVBORw0KGgoAAAANSUhEUgAAAZAAAAAyAQMAAABI0h/eAAAABlBMVEX///8AAABVwtN+AAAACXBIWXMAAA7EAAAOxAGVKw4bAAAA/ElEQVRIie2QsWrDMBCGzxTqRbSrTGif4YzAacBvksVePFnNVjxkEASUxZDVj1EIdLYJOIu6eyz0BVS6ZDClaj10kpKxg75Bdxz6+E8C8Hj+M7dApuZmKveuy4FozRmJHyUDuJ6m7LyC7aXKPNwc33UFS3Z87WlwSnMZ1kxrwJVNWdR9LloF/EU9FhSyIpdEJVEDuBAWBYcyFp00ykCS2ZgdcknL5IrAiLYUHFYfovsCvm9IYlJ+FfY5AjqUMhCdAP5M/xScgUMxb4kb1VPeqJI9QFEwSfqnqEa7Yn7sTVfrlO+2Kh4gTe924WavT5VjsalQy9yheDwej8fBN70tUwms2MMRAAAAAElFTkSuQmCC","orcid":"","institution":"Centre Hospitalier Universitaire de Toulouse","correspondingAuthor":true,"prefix":"","firstName":"Michael","middleName":"","lastName":"Poette","suffix":""},{"id":479152789,"identity":"8be28cc2-c059-438f-9b41-92e70ba0bc30","order_by":1,"name":"Sandrine Mouysset","email":"","orcid":"","institution":"French National Centre for Scientific Research","correspondingAuthor":false,"prefix":"","firstName":"Sandrine","middleName":"","lastName":"Mouysset","suffix":""},{"id":479152790,"identity":"7ea24c66-aa2f-42ce-af09-ff35fba7e765","order_by":2,"name":"Daniel Ruiz","email":"","orcid":"","institution":"French National Centre for Scientific Research","correspondingAuthor":false,"prefix":"","firstName":"Daniel","middleName":"","lastName":"Ruiz","suffix":""},{"id":479152792,"identity":"1fce84c6-15f8-46f9-a786-245ee8be31cd","order_by":3,"name":"Jean-Marc Alliot","email":"","orcid":"","institution":"Centre Hospitalier Universitaire de Toulouse","correspondingAuthor":false,"prefix":"","firstName":"Jean-Marc","middleName":"","lastName":"Alliot","suffix":""},{"id":479152794,"identity":"444de836-222e-420f-af07-cc9dff179860","order_by":4,"name":"Vincent Pey","email":"","orcid":"","institution":"Centre Hospitalier Universitaire de Toulouse","correspondingAuthor":false,"prefix":"","firstName":"Vincent","middleName":"","lastName":"Pey","suffix":""},{"id":479152796,"identity":"e31a5379-07bf-4838-9625-e1e124f4811f","order_by":5,"name":"Vincent Minville","email":"","orcid":"","institution":"Centre Hospitalier Universitaire de Toulouse","correspondingAuthor":false,"prefix":"","firstName":"Vincent","middleName":"","lastName":"Minville","suffix":""}],"badges":[],"createdAt":"2025-06-25 09:38:11","currentVersionCode":1,"declarations":"","doi":"10.21203/rs.3.rs-6973012/v1","doiUrl":"https://doi.org/10.21203/rs.3.rs-6973012/v1","draftVersion":[],"editorialEvents":[{"content":"https://doi.org/10.1038/s41598-026-39035-z","type":"published","date":"2026-02-10T15:58:57+00:00"}],"editorialNote":"","failedWorkflow":false,"files":[{"id":102785440,"identity":"ffa30a77-2897-4f22-9e94-616124960779","added_by":"auto","created_at":"2026-02-16 16:06:39","extension":"pdf","order_by":1,"title":"","display":"","copyAsset":false,"role":"manuscript-pdf","size":1427680,"visible":true,"origin":"","legend":"","description":"","filename":"manuscript.pdf","url":"https://assets-eu.researchsquare.com/files/rs-6973012/v1_covered_5c94fd19-7616-4309-93e3-cef2f08d4158.pdf"}],"financialInterests":"No competing interests reported.","formattedTitle":"Benchmarking Imputation Strategies for Missing Time-Series Data in Critical Care Using Real-World-Inspired Scenarios","fulltext":[],"fulltextSource":"","fullText":"","funders":[],"hasAdminPriorityOnWorkflow":false,"hasManuscriptDocX":false,"hasOptedInToPreprint":true,"hasPassedJournalQc":"","hasAnyPriority":true,"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":"","lastPublishedDoi":"10.21203/rs.3.rs-6973012/v1","lastPublishedDoiUrl":"https://doi.org/10.21203/rs.3.rs-6973012/v1","license":{"name":"CC BY 4.0","url":"https://creativecommons.org/licenses/by/4.0/"},"manuscriptAbstract":"Handling missing data remains a central issue in ICU time-series analysis, where data gaps often stem from non-random factors such as sensor disconnections or clinical workflows. In this study, we systematically benchmarked several imputation strategies using monitoring data from MIMIC-IV and designed masking scenarios that reflect common ICU patterns, including random dropouts, temporary monitoring interruptions, and sensor-specific failures.\nWe compared simple statistical approaches (mean, LOCF, interpolation), classical machine learning techniques (MICE, MissForest), and deep learning models (SAITS, BRITS, US-GAN, GP-VAE). SAITS, based on Transformer architecture, achieved the best performance in most settings. However, linear interpolation—despite its simplicity—yielded robust estimates in short univariate gaps and occasionally performed comparably to neural models.\nOur findings suggest that while deep learning methods improve overall imputation accuracy, simpler and more interpretable approaches may be sufficient for many ICU applications. This work introduces a practical framework for evaluating time-series imputation strategies under realistic constraints, with a focus on clinical relevance and operational deployability.","manuscriptTitle":"Benchmarking Imputation Strategies for Missing Time-Series Data in Critical Care Using Real-World-Inspired Scenarios","msid":"","msnumber":"","nonDraftVersions":[{"code":1,"date":"2025-07-02 02:49:01","doi":"10.21203/rs.3.rs-6973012/v1","editorialEvents":[{"type":"communityComments","content":0},{"type":"decision","content":"Revision requested","date":"2025-10-16T09:44:00+00:00","index":"","fulltext":""},{"type":"editorInvitedReview","content":"","date":"2025-09-26T15:05:11+00:00","index":"hide","fulltext":""},{"type":"reviewerAgreed","content":"196467075898158711442155843089195382330","date":"2025-09-12T16:24:32+00:00","index":"hide","fulltext":""},{"type":"editorInvitedReview","content":"","date":"2025-08-11T01:35:56+00:00","index":"hide","fulltext":""},{"type":"reviewerAgreed","content":"256157444132960994098611533650724812171","date":"2025-07-21T15:57:38+00:00","index":"hide","fulltext":""},{"type":"reviewersInvited","content":"","date":"2025-07-21T01:35:26+00:00","index":"","fulltext":""},{"type":"editorInvited","content":"","date":"2025-07-09T11:20:43+00:00","index":"","fulltext":""},{"type":"editorAssigned","content":"","date":"2025-06-30T12:50:36+00:00","index":"","fulltext":""},{"type":"checksComplete","content":"","date":"2025-06-28T03:47:09+00:00","index":"","fulltext":""},{"type":"submitted","content":"Scientific Reports","date":"2025-06-25T09:24:30+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":"c8e03c51-5d50-4c34-b78f-50a353c7703a","owner":[],"postedDate":"July 2nd, 2025","published":true,"recentEditorialEvents":[],"rejectedJournal":[],"revision":"","amendment":"","status":"published-in-journal","subjectAreas":[{"id":50878218,"name":"Physical sciences/Mathematics and computing/Computer science"},{"id":50878219,"name":"Biological sciences/Computational biology and bioinformatics/Data processing"},{"id":50878220,"name":"Biological sciences/Computational biology and bioinformatics/Machine learning"}],"tags":[],"updatedAt":"2026-02-16T16:03:39+00:00","versionOfRecord":{"articleIdentity":"rs-6973012","link":"https://doi.org/10.1038/s41598-026-39035-z","journal":{"identity":"scientific-reports","isVorOnly":false,"title":"Scientific Reports"},"publishedOn":"2026-02-10 15:58:57","publishedOnDateReadable":"February 10th, 2026"},"versionCreatedAt":"2025-07-02 02:49:01","video":"","vorDoi":"10.1038/s41598-026-39035-z","vorDoiUrl":"https://doi.org/10.1038/s41598-026-39035-z","workflowStages":[]},"version":"v1","identity":"rs-6973012","journalConfig":"researchsquare"},"__N_SSP":true},"page":"/article/[identity]/[[...version]]","query":{"redirect":"/article/rs-6973012","identity":"rs-6973012","version":["v1"]},"buildId":"8U1c8b4HqxoKbykW_rLl7","isFallback":false,"isExperimentalCompile":false,"dynamicIds":[84888],"gssp":true,"scriptLoader":[]}