A Novel Method for Diagnosing the Severity of Alzheimer Disease Using Deep Neural Network

Alzheimer's disease (AD) is progressive dementia that causes loss of communication between nerve cells in adults. According to forecasts by the World Alzheimer's Association by 2050, the number of people age 65 and older with Alzheimer’s dementia may grow to a projected 13.8 million, barring the development of medical breakthroughs to prevent, slow or cure Alzheimer’s disease. Doctors use a variety of clinical methods to classify Alzheimer's disease one of which is Clinical dementia rating (CDR). The main purpose of this research was to introduce a design of Deep Neural Network architecture for classifying the severity of Alzheimer cases based on the features that were extracted and transform into integer data from the MRI scans that gathered from the open access series of imaging studies (OASIS) by the Washington University Alzheimer’s Disease research center. Our initial experimental results show that the model developed here can reliably detect the severity of Alzheimer’s with accuracy of 75.0%. In terms of Sensitivity, Precision, and F1-score the obtained results 75.0%, 75.3%, and 75.9%. Regarding to obtained results from the experiments and evaluation based on metrics we can demonstrate that the proposed model can be employed to assist professionals in validating their diagnosis, also can be employed via cloud as CAD system for health centers around the globe

Keywords:

Deep Learning , Alzheimer , Clinical Dementia Rating (CDR) , Deep Neural Network , Classification