Browsing by Author "Nezamabadi, Kasra"
Now showing 1 - 2 of 2
Results Per Page
Sort Options
Item Unsupervised ECG Analysis: A Review(IEEE Reviews in Biomedical Engineering, 2022-02-28) Nezamabadi, Kasra; Sardaripour, Neda; Haghi, Benyamin; Forouzanfar, MohamadElectrocardiography is the gold standard technique for detecting abnormal heart conditions. Automatic analysis of electrocardiogram (ECG) can help physicians in the interoperation of the large amount of data produced daily by cardiac monitors. As the successful application of supervised machine learning algorithms relies on unprecedented amounts of labeled training data, there is a growing need for unsupervised algorithms for ECG analysis. Unsupervised learning aims to partition ECG into distinct abnormality classes without cardiologist-supplied labelsa process referred to as ECG clustering. In addition to abnormality detection, ECG clustering can discover inter and intra-individual patterns that carry valuable information about the whole body and mind, such as emotions and mental disorders. ECG clustering can also resolve specific challenges facing supervised learning systems, such as the imbalanced data problem, and can enhance biometric systems. While several reviews exist on supervised ECG analysis, a comprehensive review of unsupervised ECG analysis techniques is still lacking. This study reviews recent ECG clustering techniques with the focus on machine learning and deep learning algorithms. We critically review and compare these techniques, discuss their applications and limitations, and provide future research directions. This review provides further insights into ECG clustering and presents the necessary information required to adopt the appropriate algorithm for a specific application.Item XplainScar: explainable artificial intelligence to identify and localize left ventricular scar in hypertrophic cardiomyopathy from 12-lead electrocardiogram(University of Delaware, 2023) Nezamabadi, KasraMyocardial scar in the left ventricular contributes significantly to sudden cardiac death in hypertrophic cardiomyopathy (HCM). Although late gadolinium-contrast enhanced magnetic resonance imaging is commonly used for detecting HCM scar, its high cost, limited availability, and susceptibility to artifacts from implanted devices make it unsuitable for ongoing scar progression and risk stratification monitoring. The 12-lead electrocardiogram (ECG) is a widely accessible alternative, but its utilization in identifying LV scar has been limited by the complexity and heterogeneity of HCM, even for human experts. To address this challenge, we propose XplainScar, an innovative and explainable machine learning framework that identifies LV scar from 12-lead ECGs. XplainScar employs three key strategies: (1) extracting simple yet comprehensive ECG features to enable explainable predictions of scar in different LV regions, (2) integrating a novel combination of unsupervised and self-supervised representation learning to reduce inter-patient variability in diverse HCM cohorts, and (3) incorporating an explanation framework that provides valuable insights into LV scar ECG markers, aiding risk stratification and patient management. ☐ We retrospectively studied 500 HCM patients (John Hopkins HCM registry) for model development and 248 patients (UCSF HCM registry) for validation. Our approach identifies LV scar in the JH-dataset with high precision (90%), sensitivity (95%), specificity (80%), F1-score (90%), and generalizes well to the held-out test UCSF-data (precision:88%, sensitivity:90%, specificity:78%, F1-score:89%). The top ECG features identified for basal scar are Q-amplitude, Q-slope, non-terminal-QRS-duration in aVR, and area-under-QRS and T-wave-energy in V1-V2. T-wave inversion in V4–V6, area-under-QRS in V3, and TP-slope in V3-V4 predicted apical scar. Features selected for mid-scar prediction combine those for the basal and apical scars. ☐ XplainScar represents a pioneering advancement in ECG-based identification of HCM scar. It demonstrates good performance, generalizes well to unseen data from a different center, and reveals ECG markers of LV-scar in HCM. Some of these markers agree with previous studies, yet some have never been previously explored, expanding existing knowledge. XplainScar also illuminates the intricate interplay between different modalities, including electronic health records and echocardiography, and their correlation with LV-scar, facilitating informed risk stratification and patient management.