Multimodal Classification for Alzheimer's Disease Diagnosis and Progression Prediction using Deep Learning
Alzheimer's Disease (AD) is a neurodegenerative disorder that leads to cognitive decline and is the leading cause of dementia in the elderly. Accurate and early diagnosis is crucial to improve patient outcomes, and computer-aided diagnosis is becoming an essential tool for screening at-risk individuals. In this study, we propose a deep learning based classifier that can classify multi-modal imaging data. Our approach involves using a modified ResNet50 architecture, combined with late fusion of Magnetic Resonance Imaging (MRI) and Positron Emission Tomography (PET) images, to classify patients into one of three stages of AD. We also developed two separate models to identify the risk of AD and monitor disease progression. By addressing these challenges, our research aims to provide clinicians with advanced tools for accurate diagnosis and effective monitoring of disease progression, ultimately contributing to improving patient outcomes and advancing our understanding of this devastating disease.