Early detection of Alzheimer's Disease and Dementia Using Deep Convolutional Neural Networks
G. Lakshmi Praveena, G. P. Ramesh
Abstract
Alzheimer's disease (AD) is a progressive disease of the nervous system of the brain that weakens the brain functions which leads the patient to bedridden. Overall dementia cases are approximately 75% of elderly people above 65 years of age worldwide. Early detection AD cases constitute nearly 2-5%. Detecting Alzheimer's disease earlier is a difficult and challenging task, which requires human experts and MRI reports. An alternative approach for early detection such as a convolution neural network has been proposed in this paper with more reliable and cost-efficient. From the 3D MRI image report, Alzheimer's Disease and Dementia are detected and also the AD stages are diagnosed using CNN. The CNNs datasheet on sMRI of the brain is loaded in the online database. The image Classification task is analysed and evaluated using the ADNet. This analysis utilizes Magnetic Resonance (MR) brain images and Convolutional Neural Network (CNN) architecture with a deep learning pipeline. The challenging task is to classify Magnetic Resonance (MR) brain images based on the AD stage into Mild dementia (MD), Very Mild Dementia (VMD)., Non-dementia (ND), and Moderate Dementia (MoD). The results are outperformed with a high accuracy of 99.94 %.