Unsupervised Cluster-Augmented Mortality Prediction in MIMIC-IV Using Tabular ICU Time-Series Aggregation and Gradient Boosted Learning

Unsupervised Cluster-Augmented Mortality Prediction in MIMIC-IV Using Tabular ICU Time-Series Aggregation and Gradient Boosted Learning | 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 Unsupervised Cluster-Augmented Mortality Prediction in MIMIC-IV Using Tabular ICU Time-Series Aggregation and Gradient Boosted Learning Quang Bui, Aly Dhedhi This is a preprint; it has not been peer reviewed by a journal. https://doi.org/ 10.21203/rs.3.rs-9612331/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 Predicting in-hospital mortality for ICU patients remains one of the harder problems in clinical machine learning, largely because the data are messy, imbalanced, and recorded at irregular intervals. We built a framework that tackles this by combining three components: statistical aggregation of raw laboratory time-series, unsupervised patient clustering to surface latent physiological subgroups, and gradient-boosted classification using LightGBM and XGBoost. All experiments used the MIMIC-IV critical care database. KMeans was applied to derive phenotype labels, which were fed as additional features into the supervised stage. We held evaluation to a strict standard: patient-level GroupKFold splitting to rule out data leakage, and a metric suite covering AUROC, AUPRC, Brierscore, calibration error, and threshold-specific operating characteristics. Across folds the best models reached AUROC ≈ 0.86, with tree-based methods showing well-calibrated probability outputs and the cluster labels contributing to interpretable risk stratification. The full implementation is publicly available as a Google Colab notebook. Computational Biology ICU mortality prediction MIMIC-IV gradient boosting LightGBM XGBoost KMeans clustering phenotype discovery clinical machine learning time-series aggregation AUROC calibration Full Text Additional Declarations The authors declare no competing interests. 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-9612331","acceptedTermsAndConditions":true,"allowDirectSubmit":true,"archivedVersions":[],"articleType":"Research Article","associatedPublications":[],"authors":[{"id":634406180,"identity":"1f5de0ac-9407-434b-a0ed-c55cebd6bb36","order_by":0,"name":"Quang Bui","email":"data:image/png;base64,iVBORw0KGgoAAAANSUhEUgAAAZAAAAAyAQMAAABI0h/eAAAABlBMVEX///8AAABVwtN+AAAACXBIWXMAAA7EAAAOxAGVKw4bAAAA2UlEQVRIiWNgGAWjYNCCChDBw3CABC1ngJiNJC2MbRAtxAG+G8nPHn6dV5e4Xb734IEfNQxy5v0L8GuRvJFmbiy77XDizja+hIM9xxiMZW48wK/F4EaCmbTktgOJG47xGBxmYGNInCFxgJCW9G/SknPqoFr+EaUlx0zyYwMzRAswHBJn8DcQ8MuZN2XSDMcOG284lmNwsLdPwlhCAr8OBr7j6dskf9TUyW44fMb4w49vNnIS/AQcBoo+ZmCMOEJdA7RCIoGwFsYfDAz2CBGCtoyCUTAKRsFIAwDV10nbdOAbGAAAAABJRU5ErkJggg==","orcid":"https://orcid.org/0009-0006-7702-116X","institution":"American International School Vienna","correspondingAuthor":true,"prefix":"","firstName":"Quang","middleName":"","lastName":"Bui","suffix":""},{"id":634406205,"identity":"a01b8f61-5213-4008-afa1-034433881ccb","order_by":1,"name":"Aly Dhedhi","email":"","orcid":"","institution":"Jumeirah College","correspondingAuthor":false,"prefix":"","firstName":"Aly","middleName":"","lastName":"Dhedhi","suffix":""}],"badges":[],"createdAt":"2026-05-04 22:23:32","currentVersionCode":1,"declarations":{"humanSubjects":false,"vertebrateSubjects":false,"conflictsOfInterestStatement":false,"humanSubjectEthicalGuidelines":false,"humanSubjectConsent":false,"humanSubjectClinicalTrial":false,"humanSubjectCaseReport":false,"vertebrateSubjectEthicalGuidelines":false},"doi":"10.21203/rs.3.rs-9612331/v1","doiUrl":"https://doi.org/10.21203/rs.3.rs-9612331/v1","draftVersion":[],"editorialEvents":[],"editorialNote":"","failedWorkflow":false,"files":[{"id":108559365,"identity":"8dac72c2-87dd-44ba-ad56-ffb2f61400bc","added_by":"auto","created_at":"2026-05-06 02:26:46","extension":"pdf","order_by":1,"title":"","display":"","copyAsset":false,"role":"manuscript-pdf","size":356298,"visible":true,"origin":"","legend":"","description":"","filename":"FinalPaper.pdf","url":"https://assets-eu.researchsquare.com/files/rs-9612331/v1_covered_0ec8a0b2-127a-4daf-9ce5-ad344341b7dc.pdf"}],"financialInterests":"The authors declare no competing interests.","formattedTitle":"\u003cp\u003eUnsupervised Cluster-Augmented Mortality Prediction in MIMIC-IV Using Tabular ICU Time-Series Aggregation and Gradient Boosted Learning\u003c/p\u003e","fulltext":[],"fulltextSource":"","fullText":"","funders":[],"hasAdminPriorityOnWorkflow":false,"hasManuscriptDocX":false,"hasOptedInToPreprint":true,"hasPassedJournalQc":"","hasAnyPriority":true,"hideJournal":true,"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":"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":"ICU mortality prediction, MIMIC-IV, gradient boosting, LightGBM, XGBoost, KMeans clustering, phenotype discovery, clinical machine learning, time-series aggregation, AUROC, calibration","lastPublishedDoi":"10.21203/rs.3.rs-9612331/v1","lastPublishedDoiUrl":"https://doi.org/10.21203/rs.3.rs-9612331/v1","license":{"name":"CC BY 4.0","url":"https://creativecommons.org/licenses/by/4.0/"},"manuscriptAbstract":"\u003cp\u003ePredicting in-hospital mortality for ICU patients remains one of the harder problems in clinical machine learning, largely because the data are messy, imbalanced, and recorded at irregular intervals. We built a framework that tackles this by combining three components: statistical aggregation of raw laboratory time-series, unsupervised patient clustering to surface latent physiological subgroups, and gradient-boosted classification using LightGBM and XGBoost. All experiments used the MIMIC-IV critical care database. KMeans was applied to derive phenotype labels, which were fed as additional features into the supervised stage. We held evaluation to a strict standard: patient-level GroupKFold splitting to rule out data leakage, and a metric suite covering AUROC, AUPRC, Brierscore, calibration error, and threshold-specific operating characteristics. Across folds the best models reached AUROC ≈ 0.86, with tree-based methods showing well-calibrated probability outputs and the cluster labels contributing to interpretable risk stratification. The full implementation is publicly available as a Google Colab notebook.\u003c/p\u003e","manuscriptTitle":"Unsupervised Cluster-Augmented Mortality Prediction in MIMIC-IV Using Tabular ICU Time-Series Aggregation and Gradient Boosted Learning","msid":"","msnumber":"","nonDraftVersions":[{"code":1,"date":"2026-05-06 02:26:40","doi":"10.21203/rs.3.rs-9612331/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":"1a83e00a-b3b1-49ba-8dc5-9004a7dc5506","owner":[],"postedDate":"May 6th, 2026","published":true,"recentEditorialEvents":[],"rejectedJournal":[],"revision":"","amendment":"","status":"posted","subjectAreas":[{"id":67515336,"name":"Computational Biology"}],"tags":[],"updatedAt":"2026-05-06T02:26:40+00:00","versionOfRecord":[],"versionCreatedAt":"2026-05-06 02:26:40","video":"","vorDoi":"","vorDoiUrl":"","workflowStages":[]},"version":"v1","identity":"rs-9612331","journalConfig":"researchsquare"},"__N_SSP":true},"page":"/article/[identity]/[[...version]]","query":{"redirect":"/article/rs-9612331","identity":"rs-9612331","version":["v1"]},"buildId":"XKTyCvWXoU3ODBz1xrDgd","isFallback":false,"isExperimentalCompile":false,"dynamicIds":[84888],"gssp":true,"scriptLoader":[]}