Stratifying Individuals into Non-Alcoholic Fatty Liver Disease Risk Levels using Time Series Machine Learning Models
Non-alcoholic fatty liver disease (NAFLD) affects 25% of the general population worldwide. While most NAFLD patients are asymptomatic, NAFLD may progress to fibrosis and later into cirrhosis, cardiovascular disease and diabetes. We stratified patients' risks for NAFLD and advanced fibrosis from 2007 to 2017 by modeling the fibrosis risk in 5,579 individuals (from main Israeli hospital). Time-series machine learning models (Hidden Markov Models and Group-Based Trajectory Models) profiled fibrosis risk by modeling patients’ latent medical status (resulted in three groups). Few works on NAFLD fibrosis have monitored individuals after the first visit. Our results, however, show the substantial value in considering all visits and discovering the latent medical state of each individual and its risk level. We show that individuals in the high-risk group had more abnormal lab test values and an increasing prevalence of chronic conditions (e.g. hypertension and diabetes). Classification models with the group information showed that they significantly improved risk fibrosis prediction. These findings suggest that longitudinal risk assessment (optionally accessed via EMR/EHR) enables early identification of specific individual groups exhibiting distinct medical trajectories based on their routine visits. Then, it may be used to make population-specific medical recommendations to avoid the progression of chronic disease and its complications. Incorporating unique machine learning methods into the analytics of IS papers is considered highly advisable and our work underscores the value of such data science components for IS researchers and managers.
Short Bio: Ofir Ben-Assuli is an associate dean for research and an associate professor in the faculty of Business Administration at the Ono Academic College, Israel. He is currently the chair of the Israel Association for Information Systems (ILAIS), A Chapter of the Association for Information Systems. He completed his Ph.D. degree in Management Information Systems, Tel Aviv University in 2011; MBA (Specialization in Information Systems and Finance), The Hebrew University in 2005 and a B.A. in Economics and Computer Science, The Ben-Gurion University in 2002. His main research interest is in the area of Decision-Making, Healthcare Information Technology and Medical Informatics. His publications have appeared in MIS Quarterly, European Journal of Information Systems, Decision Sciences, Decision Support Systems, among others. He has been awarded several grants for his research including U.S.-Israel Binational Science Foundation (BSF), German-Israeli Foundation (GIF) and Ministry of Health - Chief Scientist Office among others.