Método para diagnóstico e identificação de pacientes em risco de desenvolver Alzheimer utilizando técnicas de inteligência artificial

Abstract

Over the years, advances in health research have led to discovering cures and treatments for numerous diseases. However, there are still several diseases in which the cure is unknown and the existing treatments are ineffective. Among them is Alzheimer’s Disease (AD), which is a chronic neurodegenerative disease and affects mainly people over 65 years old. The AD has no cure and early diagnosis is essential for new therapeutic strategies. Thus, the search for treatments that may delay the progression of the disease becomes essential, especially if the disease is discovered early. This way, it is necessary to identify patients who are at risk of developing the disease, preferably in a non-invasive manner, with a low cost for implementation and application and accessible to a large part of the population. Therefore, this work aims to study Alzheimer’s disease, its characteristics and the database of the Alzheimer’s Disease Neuroimaging Initiative (ADNI), which contains clinical, imaging, genetic and biospecimen biomarkers, thus obtaining the necessary knowledge to develop a method for classifying patients who tend to progress from a preclinical stage or from pre-dementia to AD, in order to identify individuals at risk of developing the disease. In this way, the Two-Step Classifier (TSC), a method of classifying patients at risk, is proposed. The TSC was modeled in the RapidMiner Studio tool and is architected in two steps. The first-step is responsible for the classification of patients at risk, while the second-step performs the prediction of patients’ cognitive status. TSC reaches a precision of 89.3% in identifying patients at risk of developing Alzheimer.