Show simple item record

dc.contributor.authorBrisimi, Theodora S.en_US
dc.contributor.authorXu, Tingtingen_US
dc.contributor.authorWang, Taiyaoen_US
dc.contributor.authorDai, Wuyangen_US
dc.contributor.authorAdams, William G.en_US
dc.contributor.authorPaschalidis, Ioannis Ch.en_US
dc.date.accessioned2018-06-20T15:16:12Z
dc.date.available2018-06-20T15:16:12Z
dc.date.issued2018-04
dc.identifier.citationTheodora S. Brisimi, Tingting Xu, Taiyao Wang, Wuyang Dai, William G. Adams, Ioannis Ch. Paschalidis. 2018. "Predicting Chronic Disease Hospitalizations from Electronic Health Records: An Interpretable Classification Approach." Proceedings of the IEEE, v. 106, issue 4, pp. 690 - 707.
dc.identifier.issn0018-9219
dc.identifier.issn1558-2256
dc.identifier.urihttps://hdl.handle.net/2144/29593
dc.description.abstractUrban living in modern large cities has significant adverse effects on health, increasing the risk of several chronic diseases. We focus on the two leading clusters of chronic diseases, heart disease and diabetes, and develop data-driven methods to predict hospitalizations due to these conditions. We base these predictions on the patients' medical history, recent and more distant, as described in their Electronic Health Records (EHRs). We formulate the prediction problem as a binary classification problem and consider a variety of machine learning methods, including kernelized and sparse Support Vector Machines (SVMs), sparse logistic regression, and random forests. To strike a balance between accuracy and interpretability of the prediction, which is important in a medical setting, we propose two novel methods: K -LRT, a likelihood ratio test-based method, and a Joint Clustering and Classification (JCC) method which identifies hidden patient clusters and adapts classifiers to each cluster. We develop theoretical out-of-sample guarantees for the latter method. We validate our algorithms on large data sets from the Boston Medical Center, the largest safety-net hospital system in New England.en_US
dc.description.sponsorshipNational Science Foundation (NSF); CNS-1645681; CCF-1527292; IIS-1237022; IIS-1724990; W911NF-12-1-0390 - Army Research Office (ARO); 1UL1TR001430 - National Institutes of Health (NIH); Clinical & Translational Science Institute at Boston University; Boston University Digital Health Initiativeen_US
dc.format.extentp. 690 - 707en_US
dc.relation.ispartofProceedings of the IEEE
dc.subjectArtificial intelligence and image processingen_US
dc.subjectBiomedical engineeringen_US
dc.subjectElectrical and electronic engineeringen_US
dc.subjectDiseasesen_US
dc.subjectMedical servicesen_US
dc.subjectDiabetesen_US
dc.subjectPredictive modelsen_US
dc.subjectSmart citiesen_US
dc.subjectHospitalsen_US
dc.subjectClustering methodsen_US
dc.subjectElectronic healthcareen_US
dc.subjectMachine learningen_US
dc.titlePredicting chronic disease hospitalizations from electronic health records: an interpretable classification approachen_US
dc.typeArticleen_US
dc.identifier.doi10.1109/JPROC.2017.2789319
pubs.elements-sourcecrossrefen_US
pubs.notesEmbargo: No embargoen_US
pubs.organisational-groupBoston Universityen_US
pubs.organisational-groupBoston University, College of Engineeringen_US
pubs.organisational-groupBoston University, College of Engineering, Department of Electrical & Computer Engineeringen_US
pubs.publication-statusPublisheden_US
dc.identifier.orcid0000-0002-3343-2913 (Paschalidis, Ioannis Ch.)


This item appears in the following Collection(s)

Show simple item record