Generic medical concept embedding and time decay for diverse patient outcome prediction tasks

Yupeng Li; Wei Dong; Boshu Ru; Adam Black; Xinyuan Zhang; Yuanfang Guan

doi:10.1016/j.isci.2022.104880

Generic medical concept embedding and time decay for diverse patient outcome prediction tasks

iScience. 2022 Aug 4;25(9):104880. doi: 10.1016/j.isci.2022.104880. eCollection 2022 Sep 16.

Authors

Yupeng Li¹, Wei Dong², Boshu Ru¹, Adam Black³, Xinyuan Zhang², Yuanfang Guan⁴

Affiliations

¹ Merck & Co., Inc., Rahway, NJ, USA.
² Ann Arbor Algorithms Inc., Ann Arbor, MI 48104, USA.
³ Odysseus Data Services, Cambridge, MA 02142, USA.
⁴ Department of Computational Medicine and Bioinformatics, University of Michigan, Ann Arbor, MI 48109, USA.

Abstract

Many fields, including Natural Language Processing (NLP), have recently witnessed the benefit of pre-training with large generic datasets to improve the accuracy of prediction tasks. However, there exist key differences between the longitudinal healthcare data (e.g., claims) and NLP tasks, which make the direct application of NLP pre-training methods to healthcare data inappropriate. In this article, we developed a pre-training scheme for longitudinal healthcare data that leverages the pairing of medical history and a future event. We then conducted systematic evaluations of various methods on ten patient-level prediction tasks encompassing adverse events, misdiagnosis, disease risks, and readmission. In addition to substantially reducing model size, our results show that a universal medical concept embedding pretrained with generic big data as well as carefully designed time decay modeling improves the accuracy of different downstream prediction tasks.

Keywords: Automation in bioinformatics; Bioinformatics; Health sciences; Medical informatics.