Focal cortical dysplasia (FCD) is one of the main causes of epilepsy, and it is of great assistance if the FCD lesions can be localized before the magnetic resonance (MR) imaging guided resective surgery. However, visual detection of these features within the FCD lesional regions is time consuming. - Many automated FCD detection methods have been developed by feature computation, and single classifier based classification. However, the quantity of falsely recognized nonFCD regions as positives is too large that the classification results can be less useful in automated recognition of the FCD lesions. Based on the existing studies, we propose an unanimous voting of the multiple classifiers (UVMC) method to reduce the false positive classification results of the FCD lesions detection. The proposed UVMC method was experimented on 10 MR images of patients with FCD lesions, and 31 MR images of healthy controls. The proposed UVMC method achieved much less number of false positive voxels with improved trade-off between the precision and recall.