Early Alzheimer’s Disease Detection Using Semi-Supervised GAN Based on Deep Learning

Abstract

Alzheimer’s disease (AD) prediction accuracy is crucial for minimising memory loss and enhancing Alzheimer’s disease patients’ quality of life. Neuroimaging has been explored as a possible method for diagnosing Alzheimer’s disease for the past decade. The goal of this study is to create a deep learning- an alzheimer’s disease assessment from beginning to finish ahead of schedule on. The semi-supervised generative adversarial network is designed to detect the presence of Alzheimer’s disease in magnetic resonance imaging data automatically. A model mapping is established on the original picture and Before the semi-supervised Generative Adversarial Network classifier predicts the AD, the segmented result is used to efficiently partition the left and right side hippocampal volume, and The deep feature from the segmented area is derived with convolution computational intelligence morphological operations. The current study uses the alzheimer’s disease neuroimaging Initiative dataset to conduct the experiment. This method presents a revolutionary deep learning framework for detecting alzheimer’s disease that can be used to patient data from the adult situation to improve medicine and standard of living.