ECG-Derived Multiscale Entropy as a Marker of Reduced Heart-Rate Complexity in Parkinson’s Disease

Abstract

Parkinson’s disease (PD) is associated with cardio- vascular autonomic dysfunction, but its effects on heart-rate variability (HRV) complexity remain incompletely understood. This study evaluated whether ECG-derived multiscale entropy (MSE) captures PD-related alterations in HRV complexity and transfers across cohorts. MSE was computed across 20 temporal scales from RR-interval-derived heart-rate series in a Chilean cohort of 72 participants (43 controls, 29 PD). PD participants showed lower short-scale MSE, mainly across scales 1–5. An MSE-only supervised pipeline was developed using Chile-only nested cross-validation and then applied unchanged to an inde- pendent Japanese wearable RR-interval cohort of 45 participants (21 controls, 24 PD), with no refitting, rescaling, feature selection, threshold optimization, or calibration using Japanese data. The Chilean pooled out-of-fold ROC AUC was 0.739 (95% CI 0.616– 0.857), and locked Japanese validation reached 0.810 (95% CI 0.667–0.933). At threshold 0.5, external sensitivity was 0.958 and specificity was 0.381. These findings suggest that short- scale HRV complexity carries PD-relevant autonomic information with partial cross-cohort transferability, supporting MSE as a complexity framework for wearable RR-interval studies.

Eduardo Berríos

Research assistant

Estudiante de Ingeniería Civil Biomédica. Interesado en el uso de herramientas tecnológicas e informáticas para aplicaciones dentro del campo de la salud.

Leo Medina

Principal Investigator

Leo teaches engineering courses at Usach, and his research interests are in the neural engineering and computational neuroscience fields. His work has contributed to understand how nerve fibers respond to electrical stimulation.