Compact Optimal Representation of Cortical Gyrus-Sulcus profile and subsequent Analysis for Epilepsy Severity and Forewarning

Abnormalities in the cortical folding Structure of a human brain indicate neurological disorders. Therefore, analysis and investigation of cortical gyrus pattern plays an important role in the diagnostic applications of magnetic resonance images. The study of 3D brain structure involves the collection and processing of a huge amount of data resulting in computational delay. To overcome this, a novel method of compact representation of cortical gyrus profile in 2D, is presented. Based on this compact representation, 2D local gyrification indices are obtained for the selected regions. Consequently, the epilepsy severity or the need for forewarning, if any, is determined. Experimental results show that our proposed compact representation method achieves a data compression ratio of 3 to 4 with a percentage error of about 2%.