Discovering Novel intracranial EEG Biomarkers of Seizure Generating Tissue through Time-Frequency Analysis

A research team developed an automated system to identify biomarkers of seizure-generating tissue. Traditional visual EEG analysis was subjective and ineffective. The new system detects transient high-power intracranial EEG waveforms in the range of 1-500 Hz and characterizes them using eight features. Research on 14 seizure-free post-surgical patients from UCLA identified 9 waveform categories strongly associated with resected tissue. These waveforms, including beta, gamma, and ripple bands, were often invisible in standard EEG recordings. The developed classification system achieved better results than six compared algorithms (p<0.001) and was tested on 87 seizure-free patients from Detroit Medical Center. The new approach provides a more objective method for identifying seizure-generating tissue and improves automated EEG analysis capabilities.