A meta-analysis and roadmap of Alzheimer's diseases prediction by machine learning algorithms

Abstract

Precise detection of Alzheimer's disease provides a significant role in patient care, particularly at the quick stages of the disorder. Risk identification helps patients take protective steps long before permanent brain injury occurs. While several current experiments have been performing based on computerized diagnosed techniques for Alzheimer's disease, several techniques of system diagnosis or detections are constrained by congenital scrutiny. Alzheimer's disease should be diagnosed but should not be predicted from the earlier stages. Machine Learning and Deep Learning has been a popular technique used to detection of Alzheimer's disease. In this research article, we briefly study several related Alzheimer's disease literature and discuss how Machine Learning and Deep Learning could assist the researcher in detecting the disease in an efficient way at its earlier disease stages. Purpose Develop and verify a deep learning system that predicts the final diagnosis of Alzheimer's disease (AD), moderate cognitive impairment (MCI), or neither at fluorine 18 (18F) fluoride oxy glucose (FDG) PET of the brain and compares it to radiologic readers' performance. Materials and Procedures The Alzheimer's Disease Neuro imaging Initiative's perspective 18F-FDG PET brain images.