Assessing autonomic activity from the EKG related to seizure onset detection and localization

Methods are presented for the detection and analysis of heart period variation (HPV) in both the time domain and the frequency domain. Fractal analysis and rate-change detection procedures are used for the recognition of transients in the HPV random variable. The power spectra of HPV are used to quantify relative autonomic input to the heart, with power at high frequencies (sinus arrhythmia) sensitive to parasympathetic input and low frequency power due to a combination of sympathetic and parasympathetic activities. Spectral analysis of HPV is used to study the dynamic activity of the autonomic nervous system, under conditions observed clinically which are considered to be risk factors for sudden death in epileptics. Results and conclusions from a study of a population of epilepsy patients are also presented