Heart rate variability time-frequency analysis for newborn seizure detection

The identification of newborn seizures requires the processing of a number of physiological signals routinely recorded from patients, including the EEG and ECG, as well as EOG and respiration signals. Most existing studies have focused on using the EEG as the sole information source in automatic seizure detection. Some of these studies concluded that the information obtained from the EEG should be supplemented by other information obtained from other recorded physiological signals. This chapter documents an approach that uses the ECG as the basis for seizure detection and explores how such approach could be combined with the EEG based methodologies to achieve a robust automatic seizure detector.