Epileptic Seizure Detection Using an Algorithm Based on Fractal Dimension

Epilepsy is one common neurological disorder pertaining to approximately 1% of the population. It is a complex disorder that is not well-explained at the biochemical and physiological levels and hence there is a need for investigating novel methods that can be extracted and used for differentiating epileptic EEG signals from the normal one. Efficient detection is still a challenging task for many neurological disorders and there is need for investigating more novel features. The present study is on design of system that detects the epileptic activity with efficiency. Parameters of EEG are calculated in Time Frequency Domain and the analysis of signals was performed by nonlinear quantifier Higuchis Fractal Dimension (HFD) analysis which reflects the complexity of underlying brain dynamics. It is found that the fractal dimension feature with brain complexity gives an overall accuracy of 95% in determining seizures and showed parietal (P4) region of brain reflecting the sensation of seizure. This system for analysing epileptic activity detection is designed using MATLAB state of art Signal Processing algorithms. Fractal dimension was calculated reflecting brain complexity. Genes responsible for epilepsy and prediction of protein structures involved were identified that were not available earlier in any database.