Early Alzheimer’s Detection with Deep Learning

Abstract

Alzheimer's disease (AD) is the most prevalent form of dementia in the world and a progressive neurodegenerative disorder. Early detection is crucial in improving patient outcomes and developing effective treatment strategies. This thesis presents a comprehensive study on early Alzheimer's disease detection utilising deep learning techniques. A new case of Alzheimer's disease develops in the United States every 66 seconds. To address this pressing issue, advanced deep learning techniques, particularly convolutional neural networks (CNN), have proven superior to existing machine learning methods. Consequently, this project aims to enhance the performance of CNN models by leveraging the ADNI dataset. This comprehensive report employs sophisticated image processing techniques to analyse axial, coronal, and sagittal plane magnetic resonance (MRI) brain images. Picture segmentation, a crucial step in brain MRI analysis, is employed to identify and highlight affected regions precisely. By focusing on the hippocampus and brain volume, it becomes possible to diagnose critical areas of impairment through brain MRI scans.