An Efficient Methodology of Brain Abnormalities Detection using CNN Deep Learning Network

In this paper, automatic and a supervised MRI Brain abnormalities detection methodology is presented based on a raw brain MRI images using K-Means algorithm and CNN Deep Learning Network. The proposed methodology is based on three main parts. The first part is based on the construction of Deep learning network for BMRI abnormalities detection or classification as normal or abnormal brain using a training database. The second part is the brain component segmentation using K-means algorithm to extract the white and grey matter images. The third part is the brain component classification (grey and white matter) as normal or abnormal cases according to the deep learning network. According to the obtained results of brain abnormalities detection the proposed methodology successfully achieve the detection with 95% accuracy.