Crucial events identify early stage of cardiac autonomic neuropathy progression from ECG signals.

Abstract

Cardiac autonomic neuropathy (CAN) is a condition characterized by neuropathic damage resulting in aberrant regulation of heart rate, and often manifests as changes in the ECG signals characterized by specific features of complexity, such as crucial events. This research explored the relationship between CAN progression and complexity measures involving crucial events, which can be determined using the modified diffusion entropy analysis (MDEA). MDEA measures the scaling index (0.5 < δ < 1) of the diffusion trajectory made of the crucial events (defined using the method of stripes). ECGs from the CAN dataset were recorded for 20 minutes, and CAN was classified based on established criteria into three groups: normal (n=40), early (n=42), and definite (n=7) stages. Fifteen-minute segments of the ECG time series were preprocessed and denoised and multiscale modified diffusion entropy analysis (MSMDEA) was applied to quantify the scaling index δ. Significant differences between disease progression were detected by comparing the MSMDEA scaling index (δ) across 20 temporal scaling factors using post hoc analysis (p<0.05), whereas the original unscaled signal yielded no significant detection of the disease progression. Crucial events detection indicates that the normal ECG signal is closer to the highest critical complexity (δ=1 or μ= 2), associated with a healthy cardiac autonomic function. Hence, crucial event analysis can be an adjunct to precision cardiology to assess cardiac health conditions, specifically CAN, and their progression from early to severe stages.