Comparative study of machine learning, deep learning and transfer learning models for Alzheimer’s disease detection from MRI Images

Abstract

Alzheimer’s Disease is a progressive neurodegenerative disorder that significantly effects memory, congition and quality of life. Accurate and Pre-detection of AD is important for timely intervention and improved patient results. In this paper, we propose a comparative study of ML, DL and TL approaches for AD’s detection using MRI images. Three models we developed and evaluated: Support Vector Machine (SVM) classifier using handcrafted features with dimensionality reduction, a Convolution Neural Network (CNN) built from scratch and a Transfer Tearning model EfficientNetB0. The SVM model served as a baseline for classical machine learning performance, while the CNN demonstrated the capability of DL in automatic feature extraction. Experimental results highlight the effectiveness of DL and TL techniques for medical image classification tasks, emphasizing their potential in computer-aided diagnosis of Alzheimer’s Disease.