Published Paper Details:

Early and accurate diagnosis of Alzheimer's Disease (AD) and Mild Cognitive Impairment (MCI) is vital for effective patient care and timely intervention. Magnetic Resonance Imaging (MRI) serves as a powerful modality for detecting structural brain changes, but traditional manual analysis is time-consuming and subjective. This study presents a novel hybrid framework that integrates the deep Convolutional Neural Network VGG16 with the local feature extraction capability of the Scale-Invariant Feature Transform (SIFT) algorithm to classify AD and MCI from MRI scans. The approach employs a feature fusion strategy that combines global high-level features from VGG16 with fine-grained local features from SIFT, eliminating the need for complex preprocessing steps like manual segmentation. Performance evaluation using confusion matrix-derived metrics demonstrates the framework's strong discriminative power. The model achieved an accuracy of 97.60%, sensitivity of 98.00% and specificity of 97.20%, highlighting its efficiency. These results confirm the model's high accuracy, robustness, and practicality, making it a promising tool for integration into clinical decision-support systems to facilitate early and reliable diagnosis of Alzheimer's Disease.