Comparison of Performance Using Neural Networks with Various Learning Configurations for Image Classification of Alzheimer’s Disease in Humans

Abstract

This study focuses on the classification of Alzheimer's disease images in humans using the Teachable Machine platform, a machine learning tool designed to classify images based on varying stages of Alzheimer's disease: Non-demented, Very Mild Demented, Mild Demented, and Moderate Demented. A dataset comprising 33,984 images was utilized, with 85% of the data assigned to the training set and 15% to the testing set. A supervised learning approach, employing neural networks, was used to develop a model capable of distinguishing between images representing the different stages of Alzheimer's disease. We propose the Teachable Machine program has an effective algorithm for classifying Alzheimer’s disease images in humans, achieving a best accuracy rate of 89.75% with a batch size of 16, 150 epochs and a learning rate of 0.001.