Ivana Despotovic1, Ief Segers1, Ljiljana Platisa1, Ewout Vansteenkiste1, Aleksandra Pizurica1, Karel Deblaere2, Wilfried Philips1
1Department of Telecommunications & Information Processing TELIN-IPI-IBBT, Ghent University, Ghent, Belgium; 2Department of Radiology, Ghent University Hospital, Ghent, Belgium
Focal cortical dysplasia (FCD) is a major cause of refractory epilepsy and is characterized with subtle brain lesions. Identification of FCD lesions using MRI scans is the first and most difficult step in medical treatments. Most of the existing FCD detection algorithms are based on cortical thickness measurement, which requires brain MRI segmentation. In this work, we propose an improved technique for 3D brain MRI segmentation based on graph cuts algorithm. The performance of our algorithm is validated on both phantom and real patient MRI. Experimental results show that our method is effective and successfully indicates the FCD lesions.