Paper Title
Automated Alzheimer’s Disease Stage Classification Using Tinyvgg16 Convolutional Neural Networks and MRI Analysis

Abstract
One of the most critical health challenges that the 21 st century is grappling with is the Alzheimer disease (AD) which occurs in over 55 million people across the globe and is marked by bacteriological deterioration of the nervous system. Current methods of diagnosis rely mainly on clinical judgment, subjective assessment, and neuroimaging tests which are often expensive and pose a barrier for early detection and intervention. In this study, an automatic AD staging system based on Convolutional Neural Network (CNN) architecture TinyVGG16 is proposed for magnetic resonance imaging (MRI) data. The offered approach grouped patients into 4 groups namely: No Impairment, Very Mild Impairment, Mild Impairment, and Moderate Impairment. TinyVGG16 Two convolutional blocks with ten hidden units each were trained on a large dataset of MRI images with a spatial resolution of 128 pixels each. It was found that the system performed very well in all the evaluation measures as well as in generalization on unknown samples and weak overfitting. These findings pose a clinical potential of this method in the detection and staging of early AD as an effective, objective, and cost-efficient tool. By offering reproducible and quantitative assessments, the approach addresses key gaps in the existing diagnostic practices, facilitating interventions for therapy earlier. Potential technological change This technology can revolutionise the field of AD diagnosis in three different ways: less subjectivity, cheaper, and faster, more precise treatment will slow down the stages of AD progression and positively impact patient outcomes.