Differentiation of spindle associated hippocampal HFOs based on a correlation analysis.

Abstract

High Frequency Oscillations (HFOs) have been described as biomarkers of epileptogenic tissue; however their pathological/physiological classification poses a challenge to their predictive power. For the population of ripples co-occurring with sleep spindles, those ripples improving the antiparallel correlation of ripple-peaks and sleep spindle-troughs were classified as coupled-ripples and the rest as uncoupled-ripples. For the same population of ripples two reference groups called in-SOZ and non-SOZ were formed according to the ripples' location inside or outside the seizure onset zone (SOZ). Nine patients were analyzed and their formed groups were compared using three amplitude, three waveform and three frequency features. The coupled-ripples group showed similar feature values to the non-SOZ group. The correlation based classification approach shows potential to verify the SOZ and predict alterations in the memory consolidation process.