Presenting a novel approach based on deep learning neural network and using brain images to diagnose Alzheimer's disease

A major symptom of Alzheimer's disease is memory impairment, which is the most prevalent form of dementia. The risk of Alzheimer's disease is significantly increased by brain injury and post-traumatic stress disorder. It is imperative to develop an accurate computerized diagnosis system as a result of the large volume of neural data, and the low number of samples available. This paper aims to develop an automatic disease diagnosis system using deep-learning neural networks. A combination of two powerful neural networks has been proposed to identify the Alzheimer's disease, including a developed form of the 16-layer VGA and AlexNet models. Using Python programming language, 170 brain images of war veterans were examined in the present study. This study selected 70% of database images for training and 30% for testing. To extract features, the first step of training used deep learning with convolutional networks, and the second stage used the learned features to classify health status. Several methods presented in previous studies have been analyzed and compared with the results obtained in this study. The results show that AlexNet, with the fully connected classification layer, results in higher accuracy. Compared to the existing methods, this method has higher diagnostic accuracy, which has resulted in an increase of more than 4% in many cases.