Identifying the seizure onset zone with phase-amplitude coupling.

Area of Science:

• Neuroscience
• Epilepsy Research
• Signal Processing

Background:

• Accurate seizure onset zone (SOZ) identification is crucial for treating drug-resistant epilepsy (DRE).
• Phase-amplitude coupling (PAC) is a valuable tool for studying neural interactions but is underutilized for SOZ identification.

Purpose of the Study:

• To investigate the application of PAC methods for identifying the SOZ in DRE patients.
• To analyze the differences in PAC modulation index (MI) distribution features between SOZ and non-seizure onset zone (NSOZ) regions.

Main Methods:

• Computed modulation index (MI) from clinical electrocorticography (ECoG) recordings of DRE patients.
• Performed statistical analysis on temporally evolving MI distributions across multiple frequency bands.
• Integrated MI distribution features with machine learning for SOZ identification performance evaluation.

Main Results:

• Significant differences in MI distribution features were observed between SOZ and NSOZ regions.
• Machine learning models utilizing MI distribution features achieved high SOZ identification accuracy (90.69%).
• The study systematically evaluated the impact of frequency bands and time windows on identification performance.

Conclusions:

• Modulation index distribution features derived from ECoG recordings are effective biomarkers for SOZ identification.
• This PAC-based approach offers a promising avenue for improving the diagnosis and treatment of DRE.