An Elegant Approach for Diagnosis of Parkinson's disease on MRI Brain Images by Means of a Neural Network

Early diagnosis of Parkinson's disease (PD) is of immense importance, since clinical symptoms do not occur until substantial parts of the substantia nigra(SN) neurons in the brain stem have been irreparably damaged. Furthermore, large parts of the population are affected by this disease and although PD is currently regarded as incurable, the symptoms can be alleviated by the administration of drugs. Neuroprotective drugs could shelter neurons of the SN when used at the beginning of the disease in the preclinical state. Therefore a technology to detect early SN damage is wanted for the identification of individuals at risk for PD. We are presenting an loom for MRI brain slices by means of feature extraction and unsupervised clustering. In which clustering is carried out by means of a self-organizing map (SOM). Then, each pixel is classified according to the identified classes. The number of classes is a priori unknown and the artificial neural network that implements the SOM is used to determine the main classes. The detection of the classes in the SOM is done by using a K-means segmentation. This processing is useful to potential diagnosis of Parkinson´s disease in brain-stem area Principle of this solution is showing morphological operations to detection of pathological defects.