Correcting for physiological ripples improves epileptic focus identification and outcome prediction

Area of Science:

• Neuroscience
• Epileptology
• Biomedical Engineering

Background:

• High-frequency oscillations (HFOs), including ripples (80-250 Hz) and fast ripples (250-500 Hz), are crucial in epilepsy evaluation.
• Differentiating physiological HFOs from pathological ones remains a challenge, hindering accurate epileptic focus identification and surgical outcome prediction.

Purpose of the Study:

• To evaluate if defining abnormal HFO rates by statistical comparison to region-specific physiological HFO rates improves epileptic focus identification and surgical outcome prediction.
• To assess the efficacy of novel thresholds based on normative HFO rates from the MNI Open iEEG Atlas.

Main Methods:

• HFOs were detected in 151 epilepsy patients undergoing stereo-EEG and resective surgery.
• HFO detection performance was compared using standard thresholds (HFO rate > 1/min; 50% of total HFOs) versus novel thresholds derived from the MNI Open iEEG Atlas (global, regional, and regional + 10% corrected thresholds).

Main Results:

• The regional + 10% threshold demonstrated superior performance for ripple detection, achieving 77.3% accuracy in focus identification and 69.5% accuracy in outcome prediction.
• This novel threshold significantly improved focus identification and outcome prediction compared to the 50% threshold, particularly in cortical regions where physiological ripples are common.
• Normalization methods did not enhance the performance of fast ripple detection.

Conclusions:

• Comparing abnormal HFO rates to region-specific physiological rates in healthy brain tissue enhances the clinical utility of ripples for epilepsy evaluation.
• This statistically-based approach improves both epileptic focus identification and the prediction of surgical outcomes, addressing a key limitation in current clinical practice.