Efficient Computer-Aided Diagnosis of Alzheimer's Disease and Parkinson's Disease-A Survey

摘要

Dementia is a broad category of brain-related diseases that continues for a long term and severely affects thinking and daily functioning of a human being. Among different types of dementia the fatal type of brain problems are Alzheimer's Disease (AD) and Parkinson's Disease (PD). More than 70% of cases are reported as dementia is in the Alzheimer's category. In AD, the patient's brain gets severely damaged, especially the outer part of the brain like cerebral cortex, hippocampus, ventricles, etc. The AD patients have enlarged ventricles, shrinkage in hippocampus and cortex. PD is also a common dementia after AD. In PD, the patient's mid-brain gets damaged, i.e., substantia nigra. The proposed work presents an efficient automation for the detection of the AD and PD with Machine Learning Techniques (MLT). To detect the presence of PD and AD, two different types of brain image databases have to be selected: Positron Emission Tomography (PET) and Single Photon Emission Computed Tomography (SPECT) database images, both of them contain data for AD and PD patients in comparison with the healthy brain images. From the input image, different features have to be extracted like statistical moments, geometrical moments, texture features, etc. Then Region of Interest (ROI) has to be selected to differentiate disease-affected areas. The results have to be generated automatically by comparing input image with the trained samples in the database. The proposed system concentrates on applying the MLT for segregating the outer part of brain with central part of brain for diagnosing the AD and PD in comparison with the healthy brain data.