Deep Learning Approach to Diagnose Alzheimer’s Disease through Magnetic Resonance Images

Alzheimer’s disease (AD) is the disease that is neurodegenerative which causes 70 to 80% of cases of dementia worldwide. Even though over the past few years’ research on Alzheimer’s disease has risen, Although, the brain function and structure complexity enables diagnosis still a difficult task. In recent years, Machine Learning approaches have played an essential role in understanding the deep analysis of diseases detection with the help of drug delivery and image processing units in biomedical science. These techniques have enhanced the efficiency and ability to understand complex medical situations. This paper investigates a Convolutional Neural Networks (CNN) based Algorithm of Deep learning and Computer-Aided Diagnosis (CAD) for the Alzheimer’s disease (AD) early diagnosis by differentiating the affected brain and normal brain. Since detection of AD has been a difficult task. However, using CNN-based CAD, we successfully classify the MRI data of AD subjects with an accuracy rate of 97%. This research suggests that Deep Learning algorithms are the most successful methods for classifying clinical MRI subjects by the scale and shift homogeneous distill features extracted by a Convolutional Neural Network. In addition, this approach will help doctors and clinicians to understand and predict more complex systems in the biomedical field.