ECG and EEG Analysis of Brain-Heart Interactions During Seizure Episodes

Heart Rate Variability (HRV), EEG Entropy,
Brain-Heart Interaction, Seizure Disorders, Neurological Health,
Autonomic Nervous System

Seizure disorders, such as epilepsy, profoundly im-
pact brain and autonomic nervous system function, disrupting
physiological interactions between the brain and heart. This study
investigates these brain-heart interactions by analyzing Heart
Rate Variability (HRV) and Electroencephalography (EEG) en-
tropy during seizure and non-seizure states. Utilizing a dataset
from patients with intractable seizures, key HRV parameters
(e.g., SDNN, RMSSD, LF/HF ratio, etc.) and EEG entropy
measures (Approximate Entropy) were compared across seizure
episodes and baseline periods. Results indicate a significant
reduction in HRV metrics, such as SDNN and RMSSD, during
seizures, suggesting autonomic imbalance and heightened sympa-
thetic activity. Additionally, the ECG signal exhibited a marked
increase in the T/R amplitude ratio during seizures, further
reflecting the heightened cardiac response and autonomic dys-
regulation associated with these episodes. EEG entropy analysis
revealed a marked decrease in signal complexity during seizures,
indicating less dynamic brain activity. The combined findings of
reduced HRV increased T/R amplitude ratio and decreased EEG
entropy underscore the value of these biomarkers for assessing
Brain-heart interaction disruptions in seizure disorders. Our
results highlight the potential for HRV, T/R amplitude ratio, and
EEG entropy as a non-invasive clinical tool to monitor seizures
activity and evaluate neurological health, paving the way for
more refined diagnostic and therapeutic approaches.

This paper investigates the coupled dynamics of autonomic and cortical activity during epileptic seizures by concurrently analyzing heart rate variability (HRV) metrics and EEG entropy measures. Using the CHB-MIT pediatric seizure dataset, the authors extract standard time-domain (SDNN, RMSSD, pNN50), frequency-domain (LF, HF, LF/HF), and nonlinear (Poincaré SD1/SD2) HRV parameters from ECG, alongside approximate entropy of EEG signals. Their results demonstrate a significant reduction in HRV (e.g., a 16.7% drop in RMSSD) and EEG entropy during seizure episodes, coupled with an elevated T/R amplitude ratio-indicative of sympathetic overdrive and diminished neural complexity. The study's strengths lie in its multimodal approach and rigorous statistical analysis (paired t-tests, p<0.05), which together underscore the potential of these biomarkers for non-invasive seizure monitoring. However, its reliance on a single pediatric cohort and the absence of adult or chronic epilepsy subjects limit generalizability. Nevertheless, this work provides a valuable framework for integrating ECG and EEG features in wearable seizure-detection systems and informs future research on personalized, real-time neurological health assessment.