DeepAD-Insight: A Benchmark Evaluation of Deep Learning Architectures for Multiclass Alzheimer’s Disease Classification and Staging from Brain MRI Data

Alzheimer’s Disease (AD), a progressive neurode-generative condition, poses a significant public health challenge. Early and precise diagnosis is crucial for timely intervention and effective management. However, there remains a need for further exploration and integration of methodologies to enhance accuracy, interpretability, and generalizability across diverse datasets and clinical settings. This study investigates the efficacy of various deep learning models in classifying AD using brain MRI images. The research conducted a comprehensive analysis of popular deep learning and machine learning architectures. Methodology involved obtaining a Kaggle dataset of approximately 6400 brain MRI images categorized into four classes: ‘MildDemented’, ModerateDemented’, ‘NonDemented’, and ‘VeryMildDemented’. Models were trained on the training set and evaluated on the validation set, using several metrics. The CNN model achieved an impressive $\mathbf{9 8. 4 4 \%}$ accuracy, outperforming other architectures. Here findings contribute to ongoing efforts in developing accurate and reliable AD diagnosis methods, paving the way for improved early detection and intervention strategies.