Alzheimer’s Disease Classification from Cross-sectional Brain MRI using Deep Learning
S. Sreelakshmi, G. Malu, Elizabeth Sherly
Abstract
Alzheimer’s disease (AD) comes under the category of neurodegenerative disorder that is incurable. Early diagnosis of AD with proper treatment can help to prevent brain tissue damage. Imaging modalities like MRI plays a vital role in studying AD over the past few decades. Detection of AD using biomarkers with the aid of Deep learning techniques made a remarkable movement in the easy and accurate detection of this disease. The applications of deep learning in the area of biomedical image processing is exponentially growing. An automatic semantic segmentation technique is used here for AD classification. In this study, the Segnet architecture is used for automated segmentation of hippocampal atrophy of brain followed by classification yields an accuracy of 97%, which is a promising result compared with some of the state-of-the-art approaches.