Alzheimer Disease Detection using Deep Learning
Retinderdeep Singh, Chander Prabha, Hari Mohan Dixit, Shalini Kumari
Abstract
Alzheimer's disease represents a significant global health challenge, with accurate diagnosis being a critical factor in effective treatment. MRI has emerged as a potent tool for early detection and monitoring, given its non-invasive nature and the high-quality images it provides. This study introduces an innovative method for detecting Alzheimer's disease, leveraging the fine-tuned EfficientNet-B5 model, which was trained using the Augmented Alzheimer's MRI Dataset V2. The proposed model using Deep CNN has shown acceptable performance. The model is fine-tuned to identify subtle patterns and anomalies within MRI scans linked to Alzheimer's disease. By employing the Augmented Alzheimer's MRI Dataset V2 for training and evaluation, the model's robust adaptability and heightened diagnostic precision are ensured. The proposed system achieved 96.64% accuracy. This outcome underscores both the clinical promise of the method proposed and the effectiveness of employing deep learning in the realm of medical image analysis. Importantly, this method has the potential to enhance early Alzheimer's diagnosis and management, ultimately leading to improved patient outcomes and an enhanced quality of life.