An Optimized Deep Learning Approach To Identfiy the Alzheimer's Stages Identification Based on Biomarkers Extraction

Alzheimer's is a progressive brain disorder. It affects human brain cell connections; as a result, it creates memory losses and other significant mental functional problems. The diagnosis of early stages of Alzheimer's and proper medication helps control the disease's progression since Alzheimer's is not a curable disease. The precise prediction of AD stages is an open challenge for researchers. Therefore, Researchers are developing lots of Alzheimer's disease (AD) detection techniques to predict the stages accurately. Most AD detection approaches contain several issues and can't reach the maximum accuracy rate. Therefore, this research introduces an efficient AD stages detection approach to solve the accuracy issues. This research customizes two aspects of the present systems to achieve the research objective. Initially, the appropriate adaptation of the AD stage's MRI brain image processing methodologies is enhancing the biomarker regions detection accuracy, and then the AD stage detection accuracy is improved by an optimized deep learning model. In this, the characteristics of the cuckoo search optimizer are incorporated in the Deep Belief Network (DBN) to optimize the AD stages detection functionalities of the AD detection algorithm, which helps to reduce the detection error. The AD detection approach's performance is evaluated using evaluation metrics such as accuracy, correlation coefficient, precision, Mean absolute error (MAE), Root Mean Square Error (RMSE), and Signal to Noise ratio SNR. The evaluation results prove that the accuracy of the AD stages detection approach is increased up to 0.66%, and the detection error is reduced up to 0.0345 % than the comparison approaches.