Automatic Focal Cortical Dysplasiav(FCD) detection by Magnetic Resonance Image (MRI)

Nowadays, approximately 50 million people are suffering from epilepsy all over the world, of whom 30% have Focal Cortical Dysplasia (FCD), a malformation that occurs during brain cortical development. In clinical treatments, FCD lesions often have to be removed by resective surgery. Magnetic Resonance Imaging (MRI) is the most important clinical tool for identifying FCD lesions, and has allowed the diagnostic detection of FCD lesions in an increasing number of patients, leading to increased rates of successful resective surgery. However, detection of FCD lesions is still a challenging task because of various factors such as extremely subtle FCD malformations, complex convolutions of human cerebral cortex and partial volume effect due to imaging. Previous works develop MRI features of FCD lesions to highlight FCD regions. However, these MRI features also exist in Healthy Controls. We developed a new MRI features of FCD lesions, and use a multi-feature based method to perform automatic FCD detection. As a results, we improve the similarity index than the previous method. Sensitivity and specificity are also improved by proposed work. The proposed work can be a useful clinical tool to assist FCD detection.

Jaar van publicatie:2013

Toegankelijkheid:Open