DefinePK

Abstract

Significant progress has been made in diagnosing Alzheimer's disease (AD) via MRI image processing, with deep learning methods being essential. In this work, we use the ADNI dataset to investigate novel approaches for MRI image-based AD detection and classification. Convolutional Neural Networks (CNN), ResNet, and InceptionV3 models have been shown to be effective in earlier studies; CNN achieved an astounding 96.7% accuracy rate. To improve performance even more, we present the Xception model for categorization, which achieves an impressive 99% accuracy rate.
By utilizing Xception's capabilities, we are able to identify patterns associated with AD in MRI images with greater accuracy, improving diagnostic precision. Our results highlight the potential of using a variety of deep learning architectures for Alzheimer's identification, with Xception showing promise as a method for increasing MRI image classification accuracy. This study supports further attempts to create strong and trustworthy instruments for early intervention strategies for Alzheimer’s and dementia care.