Abstract
Left ventricular (LV) scar is a major risk factor for sudden death and heart failure in hypertrophic cardiomyopathy (HCM). LV scar evolves over time and needs longitudinal assessment. Currently, LV scar detection relies on late gadolinium enhancement MRI, which is limited by high cost and artifacts from implanted cardiac devices. To address this, we developed XplainScar, an explainable machine learning model that identifies LV scar using 12-lead electrocardiogram (ECG) data. XplainScar was trained and validated on retrospective data from 748 HCM patients across two centers (500 from Johns Hopkins hospital for model development, and 248 from UCSF for validation). XplainScar employs a combination of unsupervised and self-supervised representation learning to effectively predict scar presence, and discover ECG features associated with LV scar. XplainScar rapidly analyzes ECG data (< 1 min for 10 patients) and demonstrates strong predictive performance on the held-out test set, achieving an F1-score of 89%, sensitivity of 90%, specificity of 78%, and precision of 88%. By providing an effective, cost-effective, and transparent alternative to MRI, XplainScar has the potential to assist with patient care, and reduce healthcare costs related to LV scar monitoring in HCM. XplainScar is available at https://github.com/KasraNezamabadi/XplainScar .