Convolutional Neural Network Deep Learning Approach for Alzheimer Disease Classification

Numerous studies have explored the potential of deep learning (DL) in the early diagnosis of Alzheimer disease (AD). DL models, such as convolutional neural networks (CNNs), have demonstrated their potential in accurately detecting and classifying AD based on various data sources, including brain imaging scans and clinical data. Magnetic resonance imaging (MRI) is a brain imaging tool that helps in AD detecting at an early stage and this can help to slow its progression. In this study, we developed and compared two CNN models from scratch for the classification of AD stages. The first model consisted of four convolutional layers, while the second model used five convolutional layers. By examining the results of both models, we aimed to identify the impact of the depth of the CNN architecture on the classification performance. To address the issue of imbalanced datasets, we employed data augmentation techniques. The results of our experiments demonstrate the effectiveness of the proposed models with accuracies of the models of 94.79 and 98.30 respectively. Additionally, we conducted a comprehensive comparison between our approaches and state-of-the-art models in the literature that have been used to classify the same dataset. We evaluated the models using various performance metrics, including accuracy, sensitivity, specificity, and F-measure.