Inferring multimodal latent topics from electronic health records

• Published in: Nature communications Vol. 11; no. 1; pp. 2536 - 17
• Main Authors: Li, Yue, Nair, Pratheeksha, Lu, Xing Han, Wen, Zhi, Wang, Yuening, Dehaghi, Amir Ardalan Kalantari, Miao, Yan, Liu, Weiqi, Ordog, Tamas, Biernacka, Joanna M., Ryu, Euijung, Olson, Janet E., Frye, Mark A., Liu, Aihua, Guo, Liming, Marelli, Ariane, Ahuja, Yuri, Davila-Velderrain, Jose, Kellis, Manolis
• Format: Journal Article
• Language: English
• Published: London Nature Publishing Group UK 21.05.2020; Nature Publishing Group; Nature Portfolio
• Subjects: 631/114/1305; 639/705/1042; 692/700; Bayes Theorem; Bayesian analysis; Bipolar disorder; Cardiovascular diseases; Coronary artery disease; Data integration; Databases, Factual; Diagnostic systems; Disease; Electronic health records; Electronic Health Records - classification; Electronic Health Records - statistics & numerical data; Electronic medical records; Heart diseases; Humanities and Social Sciences; Humans; Informatics; Laboratory tests; Machine Learning; Mathematical models; Medical Informatics - methods; Missing data; Modelling; Models, Statistical; multidisciplinary; Phenotype; Precision medicine; Science; Science (multidisciplinary)
• ISSN: 2041-1723; 2041-1723
• DOI: 10.1038/s41467-020-16378-3

Electronic health records (EHR) are rich heterogeneous collections of patient health information, whose broad adoption provides clinicians and researchers unprecedented opportunities for health informatics, disease-risk prediction, actionable clinical recommendations, and precision medicine. However, EHRs present several modeling challenges, including highly sparse data matrices, noisy irregular clinical notes, arbitrary biases in billing code assignment, diagnosis-driven lab tests, and heterogeneous data types. To address these challenges, we present MixEHR, a multi-view Bayesian topic model. We demonstrate MixEHR on MIMIC-III, Mayo Clinic Bipolar Disorder, and Quebec Congenital Heart Disease EHR datasets. Qualitatively, MixEHR disease topics reveal meaningful combinations of clinical features across heterogeneous data types. Quantitatively, we observe superior prediction accuracy of diagnostic codes and lab test imputations compared to the state-of-art methods. We leverage the inferred patient topic mixtures to classify target diseases and predict mortality of patients in critical conditions. In all comparison, MixEHR confers competitive performance and reveals meaningful disease-related topics. Electronic Health Records (EHR) are subject to noise, biases and missing data. Here, the authors present MixEHR, a multi-view Bayesian framework related to collaborative filtering and latent topic models for EHR data integration and modeling.