
A study investigated a machine learning model leveraging Wavelet Transform-based electrocardiogram (ECG) features (ewECG) to screen for asymptomatic left ventricular dysfunction (SBHF) in type 2 diabetes mellitus (T2DM) patients. Among 178 participants, ewECG features demonstrated superior accuracy (AUC 0.81) in identifying SBHF compared to NT-proBNP (AUC 0.56) and ARIC HF risk score (AUC 0.67). The model effectively screened for diastolic dysfunction, reduced global longitudinal strain, and left ventricular hypertrophy. These findings suggest the potential of ewECG-based machine learning as a non-invasive screening tool for SBHF in T2DM patients, aiding in early detection and management.
Like
Save
Share