Using Machine Learning Methods for Detecting Alzheimer's Disease through Hippocampal Volume Analysis

Alzheimer's disease (AD) is a neurodegenerative disorder that gradually destroys memory and thinking skills. It is the most common cause of dementia and the incidence of the disease increases with age. As the elderly population increases, the incidence of the disease is expected to increase further in the coming years so developing new treatments and diagnostic methods is getting more important. The work presented in this paper assesses the utility of image processing on the Magnetic Resonance Imaging (MRI) scans to estimate the probability of early diagnosis of dementia in Alzheimer's Disease. We analyzed the data diagnosed by the Alzheimer's Disease Neuroimaging Initiative (ADNI) protocol. The analyzed data were T1-weighted magnetic resonance images of 159 patients with Alzheimer's disease, 217 patients with mild cognitive impairment and 109 cognitively healthy older people. Within the scope of the study, 3-dimensional modeling of the hippocampus, considered to be one of the first and most affected brain regions of dementia, was calculated by means of semi-automatic segmentation software. Then, a data set was formed based on age, gender, diagnosis, right and left hippocampal volume values. The diagnosis via hippocampal volume information was made by using machine learning techniques. By using this approach, we conclude that brain MRIs can be used to separate the patients with Alzheimer's Disease (AD), Mild Cognitive Impairment (MCI) and Cognitive Normal (CN); while most of the researches were only be able to separate AD with CN. Results revealed that our approach improves the performance of the computer-aided diagnosis of the Alzheimer's disease.