Diagnosis of Alzheimer's Disease from Brain Magnetic Resonance Imaging Images using Deep Learning Algorithms
R.C. Suganthe, R.S. Latha, M. Kalaiselvi Geetha, G. R. Sreekanth
Abstract
Alzheimer's disease is one amongst the progressive disorder that cruelly affects the brain cells. It causes the death of nerve cells and tissue loss in brain. It usually tends to start slowly and aggravates overtime. The symptoms of Alzheimer's disease vary from person to person depending on the severity of the unhealthiness. It exhibits behavioral symptoms such as communication impairments, memory loss, taking a longer time to complete usual activities, and change in attitude and behavior. If the problem worsens over time, then it cannot be cured. Hence it should be identified at the earlier stage itself and treat the patient to lead a normal life on their own. Deep learning algorithms exhibit marvelous performance over conventional machine learning algorithms in identifying the complex patterns in the large volumes of high-dimensional medical imaging data. Hence, recently significant attention has been paid to apply deep learning for medical diagnosis. In this research, Deep Convolution Neural Network (DCNN) and VGG-16 inspired CNN (VCNN) models have been built to classify the different stages of Alzheimer's Disease from the Magnetic Resonance Imaging(MRI) images. Experiments are carried out on an ADNI dataset and the results obtained show that the proposed models achieved excellent accuracy.