EEG and Fractal Dimension for Epileptic Seizures Detection

Epilepsy is a common neurological disorder that can cause seizures and unconsciousness. Early detection and treatment are crucial for reducing the severity of the injury. Commonly used to diagnose epilepsy, electroencephalograms (EEG) take a lot of time, are susceptible to interference, and can reduce accuracy. This research provides a reliable method for diagnosing epilepsy in EEG data. Each EEG segment is converted into an image using the short-time Fourier transform (STFT). It generates an 8-bit binary image for each TFI. Then, the FDs of the EEG data are estimated and determined using a box-counting approach. The least square-support vector machine (LS-SVM) classifier is employed to identify patients with epilepsy. This study uses original EEG signal data from Bonn University to test the suggested technique, which was 99% accuracy 98% sensitivity, and 99% specificity. Showing the provided method offers a high-accuracy classification and supports neurologists in making an epileptic condition diagnosis and subsequently proposing the proper treatment.