Training and Performance of an Electrocardiogram-enabled Machine Learning Model for Detection of Advanced Chronic Liver Disease.

Abstract

Introduction: Building on prior results, we hypothesized that an electrocardiogram (ECG)-enabled machine learning (ML) model could be used to detect advanced CLD.

Methods: A cohort with CLD and 12-lead ECGs was matched with controls from electronic health records. A machine learning model was trained as a binary classifier.

Results: There 12,930 CLD patients and 64,577 controls in the cohort. The model's discriminative ability to classify CLD showed an AUC 0.858 (95% CI: 0.850-0.866) and at the chosen threshold, CLD ECGs had 12 times higher odds of being classified as CLD (DOR 12.33, 95% CI: 11.16-13.63).

Discussion: An ECG-enabled ML model affords great promise in identifying advanced CLD in low resource areas.