Design and evaluation of convolutional neural network for detection of Alzheimer’s disease using MRI data

Abstract

Alzheimer’s Disease (AD) is a neurological condition where the decline of brain cells causes acute memory loss and severe loss in cognitive functionalities. Various Neuroimaging techniques have been developed to diagnose AD; among those, Magnetic Resonance Imaging (MRI) is one of the most prominent ones. Recent progress in medical image analysis using deep learning especially has automated this task significantly. Although the state-of-the-art architectures have achieved human-level performance in classifying AD images from Normal Control (NC), they often require predefined Regions of interest as a basis for feature extraction. This condition not only requires specialized domain knowledge of the human brain but also makes the overall design complicated. In this study, we designed a 15 layer Neural network architecture that can facilitate AD diagnosis without being dependent on any such neurological assumption. The network was tested over ADNI-1, a benchmark MRI dataset for AD research, and found an accuracy of 92.41% (AUC = 0.93). This network was further augmented with the help of ensemble learning other well known pre trained models for more accurate and consistent results, resulting in an overall accuracy of 92.44% for the entire system.