{"status":"ok","message-type":"work","message-version":"1.0.0","message":{"indexed":{"date-parts":[[2026,1,21]],"date-time":"2026-01-21T10:17:34Z","timestamp":1768990654841,"version":"3.49.0"},"reference-count":19,"publisher":"China Science Publishing & Media Ltd.","issue":"3","license":[{"start":{"date-parts":[[2021,5,12]],"date-time":"2021-05-12T00:00:00Z","timestamp":1620777600000},"content-version":"vor","delay-in-days":131,"URL":"https:\/\/creativecommons.org\/licenses\/by\/4.0\/"}],"content-domain":{"domain":["direct.mit.edu"],"crossmark-restriction":true},"short-container-title":[],"published-print":{"date-parts":[[2021,9,8]]},"abstract":"<jats:p>Computational prediction of in-hospital mortality in the setting of an intensive care unit can help clinical practitioners to guide care and make early decisions for interventions. As clinical data are complex and varied in their structure and components, continued innovation of modelling strategies is required to identify architectures that can best model outcomes. In this work, we trained a Heterogeneous Graph Model (HGM) on electronic health record (EHR) data and used the resulting embedding vector as additional information added to a Convolutional Neural Network (CNN) model for predicting in-hospital mortality. We show that the additional information provided by including time as a vector in the embedding captured the relationships between medical concepts, lab tests, and diagnoses, which enhanced predictive performance. We found that adding HGM to a CNN model increased the mortality prediction accuracy up to 4%. This framework served as a foundation for future experiments involving different EHR data types on important healthcare prediction tasks.<\/jats:p>","DOI":"10.1162\/dint_a_00097","type":"journal-article","created":{"date-parts":[[2021,5,12]],"date-time":"2021-05-12T16:12:07Z","timestamp":1620835927000},"page":"329-339","update-policy":"https:\/\/doi.org\/10.1162\/mitpressjournals.corrections.policy","source":"Crossref","is-referenced-by-count":24,"title":["Deep Learning with Heterogeneous Graph Embeddings for Mortality\n                    Prediction from Electronic Health Records"],"prefix":"10.3724","volume":"3","author":[{"given":"Tingyi","family":"Wanyan","sequence":"first","affiliation":[{"name":"Hasso Plattner Institute for Digital Health at Mount Sinai, Icahn School of Medicine at Mount Sinai, New York, New York 10065, USA"},{"name":"School of Informatics, Computing, and Engineering, Indiana University, Bloomington, IN 47405-7000, USA"}]},{"given":"Hossein","family":"Honarvar","sequence":"additional","affiliation":[{"name":"Hasso Plattner Institute for Digital Health at Mount Sinai, Icahn School of Medicine at Mount Sinai, New York, New York 10065, USA"}]},{"given":"Ariful","family":"Azad","sequence":"additional","affiliation":[{"name":"School of Informatics, Computing, and Engineering, Indiana University, Bloomington, IN 47405-7000, USA"}]},{"given":"Ying","family":"Ding","sequence":"additional","affiliation":[{"name":"Dell Medical School, University of Texas at Austin, Austin, Texas 78701-1996, USA"},{"name":"School of Informatics, University of Texas at Austin, Austin, Texas 78712-1139, USA"}]},{"given":"Benjamin S.","family":"Glicksberg","sequence":"additional","affiliation":[{"name":"Hasso Plattner Institute for Digital Health at Mount Sinai, Icahn School of Medicine at Mount Sinai, New York, New York 10065, USA"},{"name":"Department of Genetics and Genomic Sciences, Icahn School of Medicine at Mount Sinai, New York, New York 10065, USA"}]}],"member":"2026","published-online":{"date-parts":[[2021,9,8]]},"reference":[{"key":"2021102914232019500_ref1","first-page":"1","article-title":"Real-time mortality prediction in the Intensive Care\n                        Unit","volume-title":"AMIA Annual Symposium\n                    Proceedings","author":"Johnson","year":"2017"},{"key":"2021102914232019500_ref2","doi-asserted-by":"crossref","first-page":"49","DOI":"10.1145\/3093241.3093267","article-title":"Mortality prediction of ICU patients using Machine Leaning: A\n                        survey","volume-title":"Proceedings of the International\n                        Conference on Compute and Data Analysis","author":"Sharma","year":"2017"},{"issue":"4","key":"2021102914232019500_ref3","doi-asserted-by":"crossref","first-page":"334","DOI":"10.1016\/j.annemergmed.2018.11.036","article-title":"Development and evaluation of a machine learning model for\n                        the early identification of patients at risk for sepsis","volume":"73","author":"Delahanty","year":"2019","journal-title":"Annals of Emergency Medicine"},{"issue":"12","key":"2021102914232019500_ref4","doi-asserted-by":"crossref","first-page":"905","DOI":"10.1016\/S2213-2600(18)30300-X","article-title":"Machine learning for real-time prediction of complications in\n                        critical care: A retrospective study","volume":"6","author":"Meyer","year":"2018","journal-title":"The Lancet\n                        Respiratory Medicine"},{"key":"2021102914232019500_ref5","doi-asserted-by":"crossref","first-page":"R56","DOI":"10.1093\/hmg\/ddy114","article-title":"The next generation of precision medicine: Observational\n                        studies, electronic health records, biobanks and continuous\n                        monitoring","volume":"27","author":"Glicksberg","year":"2018","journal-title":"Human Molecular Genetics"},{"issue":"6","key":"2021102914232019500_ref6","doi-asserted-by":"crossref","first-page":"395","DOI":"10.1038\/nrg3208","article-title":"Mining electronic health records: Towards better research\n                        applications and clinical care","volume":"13","author":"Jensen","year":"2012","journal-title":"Nature Reviews\n                        Genetics"},{"key":"2021102914232019500_ref7","doi-asserted-by":"crossref","first-page":"145","DOI":"10.1142\/9789813235533_0014","article-title":"Automated disease cohort selection using word embeddings\n                        from Electronic Health Records","volume-title":"Pacific\n                        Symposium on Biocomputing 2018","author":"Glicksberg","year":"2018"},{"issue":"1","key":"2021102914232019500_ref8","doi-asserted-by":"crossref","first-page":"18","DOI":"10.1038\/s41746-018-0029-1","article-title":"Scalable and accurate deep learning with electronic health\n                        records","volume":"1","author":"Rajkomar","year":"2018","journal-title":"NPJ Digital Medicine"},{"issue":"5","key":"2021102914232019500_ref9","doi-asserted-by":"crossref","first-page":"1589","DOI":"10.1109\/JBHI.2017.2767063","article-title":"Deep EHR: A survey of recent advances in deep learning\n                        techniques for electronic health record (EHR) analysis","volume":"22","author":"Shickel","year":"2017","journal-title":"IEEE Journal of Biomedical and Health Informatics"},{"key":"2021102914232019500_ref10","doi-asserted-by":"crossref","first-page":"432","DOI":"10.1137\/1.9781611974348.49","article-title":"Risk prediction with electronic health records: A deep\n                        learning approach","volume-title":"Proceedings of the 2016 SIAM\n                        International Conference on Data Mining","author":"Cheng","year":"2016"},{"issue":"1","key":"2021102914232019500_ref11","doi-asserted-by":"crossref","first-page":"279","DOI":"10.1186\/s13054-019-2561-z","article-title":"A deep learning model for real-time mortality prediction in\n                        critically ill children","volume":"23","author":"Kim","year":"2019","journal-title":"Critical Care"},{"key":"2021102914232019500_ref12","doi-asserted-by":"crossref","first-page":"214","DOI":"10.1109\/CHASE.2017.80","article-title":"HCNN: Heterogeneous convolutional neural networks for\n                        comorbid risk prediction with electronic health records","volume-title":"2017 IEEE\/ACM International Conference on Connected Health:\n                        Applications, Systems and Engineering Technologies (CHASE)","author":"Zhang","year":"2017"},{"key":"2021102914232019500_ref13","volume-title":"Phe2vec: Automated disease phenotyping based on unsupervised\n                        embeddings from electronic health records","author":"De\n                                Freitas","year":"2020"},{"issue":"1","key":"2021102914232019500_ref14","first-page":"96","article-title":"Deep representation learning of electronic health records to\n                        unlock patient stratification at scale","volume":"3","author":"Landi","year":"2020","journal-title":"Digital\n                        Medicine"},{"issue":"1","key":"2021102914232019500_ref15","doi-asserted-by":"crossref","first-page":"1","DOI":"10.1038\/srep26094","article-title":"Deep patient: An unsupervised representation to predict the\n                        future of patients from the electronic health records","volume":"6","author":"Miotto","year":"2016","journal-title":"Scientific Reports"},{"key":"2021102914232019500_ref16","first-page":"4552","article-title":"Mime: Multilevel medical embedding of electronic health\n                        records for predictive healthcare","volume-title":"Proceedings\n                        of the 32nd International Conference on Neural Information Processing\n                        Systems (NIPS'18)","author":"Choi","year":"2018"},{"key":"2021102914232019500_ref17","volume-title":"Graph convolutional transformer: Learning the graphical structure of\n                        electronic health records","author":"Choi","year":"2019"},{"issue":"1","key":"2021102914232019500_ref18","doi-asserted-by":"crossref","first-page":"1","DOI":"10.1038\/sdata.2016.35","article-title":"MIMIC-III, a freely accessible critical care\n                        database","volume":"3","author":"Johnson","year":"2016","journal-title":"Scientific Data"},{"issue":"6","key":"2021102914232019500_ref19","doi-asserted-by":"crossref","first-page":"84","DOI":"10.1145\/3065386","article-title":"Imagenet classification with deep convolutional neural\n                        networks","volume":"60","author":"Krizhevsky","year":"2017","journal-title":"Communications of the ACM"}],"container-title":["Data Intelligence"],"original-title":[],"language":"en","link":[{"URL":"http:\/\/direct.mit.edu\/dint\/article-pdf\/3\/3\/329\/1969101\/dint_a_00097.pdf","content-type":"application\/pdf","content-version":"vor","intended-application":"syndication"},{"URL":"http:\/\/direct.mit.edu\/dint\/article-pdf\/3\/3\/329\/1969101\/dint_a_00097.pdf","content-type":"unspecified","content-version":"vor","intended-application":"similarity-checking"}],"deposited":{"date-parts":[[2025,3,14]],"date-time":"2025-03-14T07:42:16Z","timestamp":1741938136000},"score":1,"resource":{"primary":{"URL":"https:\/\/www.sciengine.com\/doi\/10.1162\/dint_a_00097"}},"subtitle":[],"short-title":[],"issued":{"date-parts":[[2021]]},"references-count":19,"journal-issue":{"issue":"3","published-print":{"date-parts":[[2021,9,8]]}},"URL":"https:\/\/doi.org\/10.1162\/dint_a_00097","relation":{},"ISSN":["2641-435X"],"issn-type":[{"value":"2641-435X","type":"electronic"}],"subject":[],"published-other":{"date-parts":[[2021]]},"published":{"date-parts":[[2021]]}}}