Improving Alzheimer's Disease Diagnosis with Artificial Intelligence and Neuropsychological Tests

Abstract

Alzheimer's Disease (AD) is a type of progressive dementia that causes loss of memory and function. It affects over 5.8 million people sixty years of age and older in the United States alone. When an individual has AD, neurodegeneration ultimately leads to severe impairment and eventually death. Throughout the stages of the disease, symptoms can range from mild forgetfulness to loss of motor control and bodily functions. Unfortunately, the disease may go unnoticed during initial stages if not tested for and detected early. Signs of Alzheimer's may accidentally be overlooked by physicians, especially if they are subtle. Artificial intelligence (AI) and machine learning models (ML) can be used to improve the identification process of early signs of AD. Some neuropsychological impairments AD patients suffer from include memory loss, difficulty recalling words or phrases, changes in personality and behavior, and decline in bodily functions that affect day to day life. AI may be able to pick up certain details and data that human perception cannot. If trained with a proper data set, ML models would be helpful to physicians in this area of healthcare. In this systematic review, the different neuropsychological impairments visible in the early stages of AD will be discussed as they relate to ML, deep learning, and AI applications intended to aid the diagnosis of AD.