Detection of Alzheimer's Disease using machine learning and deep learning

Alzheimer’s Disease (AD) is a progressive neurodegenerative disorder characterized by cognitive decline and structural brain changes that are often subtle in the early stages. Accurate and early diagnosis remains a major clinical challenge due to variability in patient conditions and reliance on manual interpretation of Magnetic Resonance Imaging (MRI) scans. This study proposes an automated and interpretable framework for Alzheimer’s Disease stage classification using Convolutional Neural Networks (CNN) integrated with Explainable Artificial Intelligence (XAI) techniques. Publicly available MRI datasets were collected and subjected to preprocessing steps including resizing, normalization, noise reduction, data augmentation, and labelling to ensure standardized and high-quality input for model training. The CNN model was trained to classify MRI scans into multiple disease stages, including Healthy Control (HC), Mild Cognitive Impairment (MCI), and Alzheimer’s Disease (AD). Hyperparameter tuning and validation strategies were applied to enhance generalization and robustness. Experimental results demonstrate that the proposed model achieves high classification accuracy on unseen data, indicating its potential for reliable early-stage detection. To improve clinical interpretability, the framework incorporates LIME (Local Interpretable Model-Agnostic Explanations) and Grad-CAM (Gradient-weighted Class Activation Mapping), which visually highlight the most influential brain regions contributing to model predictions. These explainability techniques enhance transparency, build clinician trust, and support informed medical decision-making. The integration of deep learning and explainable AI provides a scalable, reliable, and clinically supportive system for Alzheimer’s diagnosis, with strong potential for real-world healthcare deployment.