Comparative Study of Two Classification Methods for the Detection of Alzheimer's Disease

Background: In the last few years, the number of patients suffering from Alzheimer disease has rapidly increased. This illness is a brain disease which begins at the hippocampus. Then, it spreads to the rest of the brain. It attacks especially people over the age of 65. Thus, it is necessary to develop methods to facilitate its early detection. Methods: As part of this thesis, we tried to compare two methods of classification; the method of k Nearest Neighbor (KNN) and Support Vector Machine (SVM) to make the diagnosis of Alzheimer. The first step consists in segmenting the images into distinct blocks. After that, a features extraction is applied only on the part containing the hippocampus of the brain. Result: The following step is the classification. It allows us to know the patient's condition. Conclusion: We assess the two classifiers on the Oasis image base. The obtained results show that the KNN method is more efficient than the SVM approach.