DTKF: A Sepsis Early Prediction Framework Based on Large Language Model Semantic Augmentation and Dual-Source Knowledge Distillation

DTKF: A Sepsis Early Prediction Framework Based on Large Language Model Semantic Augmentation and Dual-Source Knowledge Distillation | 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 DTKF: A Sepsis Early Prediction Framework Based on Large Language Model Semantic Augmentation and Dual-Source Knowledge Distillation Yitian Zhang, Xinlei Ji, Huijie Zhang, Liang Zhou, Jiachen Guo, and 1 more This is a preprint; it has not been peer reviewed by a journal. https://doi.org/ 10.21203/rs.3.rs-9199143/v1 This work is licensed under a CC BY 4.0 License Status: Under Review Version 1 posted 9 You are reading this latest preprint version Abstract Background Sepsis is a life-threatening organ dysfunction caused by a dysregulated host response to infection. Early and accurate prediction is crucial for reducing mortality. However, clinical settings often face challenges such as data scarcity, class imbalance, and the underutilization of unstructured clinical notes. Existing scoring systems and traditional machine learning models primarily rely on structured vital signs, failing to capture the rich semantic risk factors embedded in textual data. Although Large Language Models (LLMs) excel in text understanding, their high computational costs and potential hallucinations hinder direct deployment in resource-constrained emergency settings. Methods To address these limitations, this study proposes a novel Dual-Teacher Knowledge Fusion (DTKF) framework. First, we introduce a semantic feature augmentation module that leverages DeepSeek to extract semantic risk probabilities from clinical notes and explicitly fuses them with structured vital signs to construct high-order hybrid features for the student model. Second, to prevent overfitting under small-sample conditions, we design a dual-source knowledge distillation strategy. This strategy integrates rule-based knowledge and data-driven semantic knowledge to construct hybrid teacher signals, regulating a lightweight logistic regression student model via soft-label supervision. Results Experiments on the MIMIC-IV dataset demonstrate that DTKF achieves an AUC of 0.796. This result not only significantly outperforms traditional unimodal baselines but also improves performance by 6.6% compared to direct prediction using LLMs. Conclusions This study effectively integrates rule-based and data-driven knowledge, enhances feature representation using textual information, and improves model performance and robustness through distillation techniques, providing a clinically valuable solution for sepsis early warning. Sepsis Early Prediction Semantic Feature Augmentation Dual-Source Knowledge Distillation Large Language Models Multi-modal Fusion Full Text Additional Declarations No competing interests reported. Cite Share Download PDF Status: Under Review Version 1 posted Reviewers agreed at journal 12 May, 2026 Reviews received at journal 11 May, 2026 Reviewers agreed at journal 02 May, 2026 Reviewers agreed at journal 14 Apr, 2026 Reviewers invited by journal 14 Apr, 2026 Editor invited by journal 01 Apr, 2026 Editor assigned by journal 27 Mar, 2026 Submission checks completed at journal 27 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-9199143","acceptedTermsAndConditions":true,"allowDirectSubmit":false,"archivedVersions":[],"articleType":"Research Article","associatedPublications":[],"authors":[{"id":624533593,"identity":"7ac20a97-50b7-402b-815f-126cd480af76","order_by":0,"name":"Yitian Zhang","email":"","orcid":"","institution":"Shanghai University of Engineering Science","correspondingAuthor":false,"prefix":"","firstName":"Yitian","middleName":"","lastName":"Zhang","suffix":""},{"id":624533594,"identity":"57293a30-cda4-45e6-ad42-5dd16ba8bf3c","order_by":1,"name":"Xinlei Ji","email":"","orcid":"","institution":"University of Shanghai for Science and Technology","correspondingAuthor":false,"prefix":"","firstName":"Xinlei","middleName":"","lastName":"Ji","suffix":""},{"id":624533595,"identity":"fae09f03-0710-4dde-95d5-44dd77143179","order_by":2,"name":"Huijie Zhang","email":"","orcid":"","institution":"Shanghai University of Engineering Science","correspondingAuthor":false,"prefix":"","firstName":"Huijie","middleName":"","lastName":"Zhang","suffix":""},{"id":624533596,"identity":"3e47541f-7e93-4baf-ae96-d8d81d707cfb","order_by":3,"name":"Liang Zhou","email":"","orcid":"","institution":"Shanghai University of Medicine and Health Sciences","correspondingAuthor":false,"prefix":"","firstName":"Liang","middleName":"","lastName":"Zhou","suffix":""},{"id":624533597,"identity":"e4879acb-a748-477a-811a-f3217a0608cd","order_by":4,"name":"Jiachen Guo","email":"","orcid":"","institution":"Shanghai University of Medicine and Health Sciences","correspondingAuthor":false,"prefix":"","firstName":"Jiachen","middleName":"","lastName":"Guo","suffix":""},{"id":624533598,"identity":"75fd5750-88f9-416a-a24c-378303b76d51","order_by":5,"name":"Lei Yin","email":"data:image/png;base64,iVBORw0KGgoAAAANSUhEUgAAAZAAAAAyAQMAAABI0h/eAAAABlBMVEX///8AAABVwtN+AAAACXBIWXMAAA7EAAAOxAGVKw4bAAAAsUlEQVRIiWNgGAWjYDACZgbGAx8YJEBMA6K1MBycwSAhQYIWIDjMw8BAghaD47wHDtv8sahjYG/eJsFQc4cILYf5Eg7ntgEdxnOsTILh2DNitPAYHM5tAGqRyDGTYGw4TKQWiz9ALfJvSNHCwAayhYdILZJALQd72yQk23jSii0SjhGhhe/8GcMHP/7U8fOzH95440MNEVoUDkAZbCAigbAGBgb5BmJUjYJRMApGwcgGAJ4ZM5jvuHwEAAAAAElFTkSuQmCC","orcid":"","institution":"Shanghai Institute of Technology","correspondingAuthor":true,"prefix":"","firstName":"Lei","middleName":"","lastName":"Yin","suffix":""}],"badges":[],"createdAt":"2026-03-23 10:25:09","currentVersionCode":1,"declarations":"","doi":"10.21203/rs.3.rs-9199143/v1","doiUrl":"https://doi.org/10.21203/rs.3.rs-9199143/v1","draftVersion":[],"editorialEvents":[],"editorialNote":"","failedWorkflow":false,"files":[{"id":107558279,"identity":"b7baf0c2-f2b3-4fb0-8815-f6a0ba0b1fde","added_by":"auto","created_at":"2026-04-22 15:26:13","extension":"pdf","order_by":1,"title":"","display":"","copyAsset":false,"role":"manuscript-pdf","size":783826,"visible":true,"origin":"","legend":"","description":"","filename":"ASepsisEarlyPredictionFrameworkBasedonLargeLanguageModelSemanticAugmentationandDualSourceKnowledgeDistillation.pdf","url":"https://assets-eu.researchsquare.com/files/rs-9199143/v1_covered_4facd8b1-8757-49af-9ead-39a7fe6fcc66.pdf"}],"financialInterests":"No competing interests reported.","formattedTitle":"DTKF: A Sepsis Early Prediction Framework Based on Large Language Model Semantic Augmentation and Dual-Source Knowledge Distillation","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":"bmc-medical-informatics-and-decision-making","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":false,"externalIdentity":"midm","sideBox":"Learn more about [BMC Medical Informatics and Decision Making](http://bmcmedinformdecismak.biomedcentral.com/)","snPcode":"","submissionUrl":"https://www.editorialmanager.com/midm/default.aspx","title":"BMC Medical Informatics and Decision Making","twitterHandle":"BMC_series","acdcEnabled":true,"dfaEnabled":false,"editorialSystem":"em","reportingPortfolio":"BMC Series","inReviewEnabled":true,"inReviewRevisionsEnabled":true},"keywords":"Sepsis Early Prediction, Semantic Feature Augmentation, Dual-Source Knowledge Distillation, Large Language Models, Multi-modal Fusion","lastPublishedDoi":"10.21203/rs.3.rs-9199143/v1","lastPublishedDoiUrl":"https://doi.org/10.21203/rs.3.rs-9199143/v1","license":{"name":"CC BY 4.0","url":"https://creativecommons.org/licenses/by/4.0/"},"manuscriptAbstract":"\u003ch2\u003eBackground\u003c/h2\u003e \u003cp\u003eSepsis is a life-threatening organ dysfunction caused by a dysregulated host response to infection. Early and accurate prediction is crucial for reducing mortality. However, clinical settings often face challenges such as data scarcity, class imbalance, and the underutilization of unstructured clinical notes. Existing scoring systems and traditional machine learning models primarily rely on structured vital signs, failing to capture the rich semantic risk factors embedded in textual data. Although Large Language Models (LLMs) excel in text understanding, their high computational costs and potential hallucinations hinder direct deployment in resource-constrained emergency settings.\u003c/p\u003e\u003ch2\u003eMethods\u003c/h2\u003e \u003cp\u003eTo address these limitations, this study proposes a novel Dual-Teacher Knowledge Fusion (DTKF) framework. First, we introduce a semantic feature augmentation module that leverages DeepSeek to extract semantic risk probabilities from clinical notes and explicitly fuses them with structured vital signs to construct high-order hybrid features for the student model. Second, to prevent overfitting under small-sample conditions, we design a dual-source knowledge distillation strategy. This strategy integrates rule-based knowledge and data-driven semantic knowledge to construct hybrid teacher signals, regulating a lightweight logistic regression student model via soft-label supervision.\u003c/p\u003e\u003ch2\u003eResults\u003c/h2\u003e \u003cp\u003eExperiments on the MIMIC-IV dataset demonstrate that DTKF achieves an AUC of 0.796. This result not only significantly outperforms traditional unimodal baselines but also improves performance by 6.6% compared to direct prediction using LLMs.\u003c/p\u003e\u003ch2\u003eConclusions\u003c/h2\u003e \u003cp\u003eThis study effectively integrates rule-based and data-driven knowledge, enhances feature representation using textual information, and improves model performance and robustness through distillation techniques, providing a clinically valuable solution for sepsis early warning.\u003c/p\u003e","manuscriptTitle":"DTKF: A Sepsis Early Prediction Framework Based on Large Language Model Semantic Augmentation and Dual-Source Knowledge Distillation","msid":"","msnumber":"","nonDraftVersions":[{"code":1,"date":"2026-04-22 15:24:55","doi":"10.21203/rs.3.rs-9199143/v1","editorialEvents":[{"type":"communityComments","content":0},{"type":"reviewerAgreed","content":"192227861071643094406132838021901935198","date":"2026-05-12T14:53:17+00:00","index":"hide","fulltext":""},{"type":"editorInvitedReview","content":"","date":"2026-05-11T23:23:43+00:00","index":"hide","fulltext":""},{"type":"reviewerAgreed","content":"61520438826847619732223782550480602849","date":"2026-05-03T01:15:59+00:00","index":"hide","fulltext":""},{"type":"reviewerAgreed","content":"132257185480058340854788560186978355677","date":"2026-04-14T18:47:33+00:00","index":"hide","fulltext":""},{"type":"reviewersInvited","content":"","date":"2026-04-14T18:39:54+00:00","index":"","fulltext":""},{"type":"editorInvited","content":"","date":"2026-04-01T12:31:33+00:00","index":"","fulltext":""},{"type":"editorAssigned","content":"","date":"2026-03-27T12:25:44+00:00","index":"","fulltext":""},{"type":"checksComplete","content":"","date":"2026-03-27T12:25:22+00:00","index":"","fulltext":""},{"type":"submitted","content":"BMC Medical Informatics and Decision Making","date":"2026-03-23T10:09:27+00:00","index":"","fulltext":""}],"status":"published","journal":{"display":true,"email":"[email protected]","identity":"bmc-medical-informatics-and-decision-making","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":false,"externalIdentity":"midm","sideBox":"Learn more about [BMC Medical Informatics and Decision Making](http://bmcmedinformdecismak.biomedcentral.com/)","snPcode":"","submissionUrl":"https://www.editorialmanager.com/midm/default.aspx","title":"BMC Medical Informatics and Decision Making","twitterHandle":"BMC_series","acdcEnabled":true,"dfaEnabled":false,"editorialSystem":"em","reportingPortfolio":"BMC Series","inReviewEnabled":true,"inReviewRevisionsEnabled":true}}],"origin":"","ownerIdentity":"be1e8de3-3406-4e1d-8926-c7a2c340f705","owner":[],"postedDate":"April 22nd, 2026","published":true,"recentEditorialEvents":[{"type":"reviewerAgreed","content":"192227861071643094406132838021901935198","date":"2026-05-12T14:53:17+00:00","index":51,"fulltext":""},{"type":"editorInvitedReview","content":"","date":"2026-05-11T23:23:43+00:00","index":50,"fulltext":""},{"type":"reviewerAgreed","content":"61520438826847619732223782550480602849","date":"2026-05-03T01:15:59+00:00","index":36,"fulltext":""}],"rejectedJournal":[],"revision":"","amendment":"","status":"under-review","subjectAreas":[],"tags":[],"updatedAt":"2026-04-22T15:24:55+00:00","versionOfRecord":[],"versionCreatedAt":"2026-04-22 15:24:55","video":"","vorDoi":"","vorDoiUrl":"","workflowStages":[]},"version":"v1","identity":"rs-9199143","journalConfig":"researchsquare"},"__N_SSP":true},"page":"/article/[identity]/[[...version]]","query":{"redirect":"/article/rs-9199143","identity":"rs-9199143","version":["v1"]},"buildId":"XKTyCvWXoU3ODBz1xrDgd","isFallback":false,"isExperimentalCompile":false,"dynamicIds":[84888],"gssp":true,"scriptLoader":[]}