B1146
Title: Clinical knowledge extraction via sparse embedding regression with EHR data
Authors: Chuan Hong - Duke University (United States) [presenting]
Abstract: Traditional data mining of EHR data often requires the use of patient-level data, which hinders the ability to share data across institutions. KESER is a knowledge extraction pipeline via sparse embedding regression, which efficiently summarizes patient-level longitudinal EHR data into hospital-specific embedding data and enables the extraction of clinical knowledge based only on summary-level data. KESER bypasses the need for patient-level data in individual analyses providing a significant advance in enabling multi-center studies using EHR data.