Classification of Alzheimer’s disease using MRI data based on Deep Learning Techniques

Abstract

Alzheimer’s Disease (AD) is a worldwide concern impacting millions of people, with no effective treatment known to date. Unlike cancer, which has seen improvement in preventing its progression, early detection remains critical in managing the burden of AD. This paper suggests a novel AD-DL approach for detecting early AD using Deep Learning (DL) Techniques. The dataset consists of pictures of brain magnetic resonance imaging (MRI) used to evaluate and validate the suggested model. The method includes stages for pre-processing, DL model training, and evaluation. Five DL models with autonomous feature extraction and binary classification are shown. The models are divided into two categories: without Data Augmentation (without-Aug), which includes CNN-without-AUG, and with Data Augmentation (with-Aug), which includes CNNs-with-Aug, CNNs-LSTM-with-Aug, CNNs-SVM-with-Aug, and Transfer learning using VGG16-SVM-with-Aug. The main goal is to build a model with the best detection accuracy, recall, precision, F1 score, training time, and testing time. The dataset is used to evaluate the recommended methodology, showing encouraging results. The experimental results show that CNN-LSTM is superior, with an accuracy percentage of 99.92%. The outcomes of this study lay the groundwork for future DL-based research in AD identification.

Keywords

• Alzheimer’s disease (AD)
• Magnetic resonance imaging (MRI)
• Deep learning (DL)
• CNNs-without-aug
• CNNs-with-aug
• CNNs-LSTM-with-aug