Seizure detection approach using S-transform and singular value decomposition.

Area of Science:

• Biomedical Engineering
• Signal Processing
• Neurology

Background:

• Automatic seizure detection is crucial for epilepsy diagnosis and management.
• Existing methods may face challenges with accuracy and false detection rates.

Purpose of the Study:

• To present a novel method for automatic seizure detection using S-transform and singular value decomposition (SVD).
• To evaluate the performance of the proposed method on intracranial electroencephalogram (EEG) recordings.

Main Methods:

• EEG signals were processed using S-transform to obtain time-frequency representations.
• Singular value decomposition (SVD) was applied to submatrices of the time-frequency data.
• Features were extracted from singular values and classified using Bayesian linear discriminant analysis (BLDA), followed by postprocessing.

Main Results:

• The proposed method demonstrated high performance on 183.07 hours of intracranial EEG data from 20 patients.
• Achieved a sensitivity of 96.40% and a specificity of 99.01%.
• Reported a low false detection rate of 0.16 per hour.

Conclusions:

• The S-transform and SVD-based method offers a promising approach for accurate and reliable automatic seizure detection.
• The system's high sensitivity and specificity suggest its potential clinical utility in epilepsy monitoring.