Prediction of Alzheimer's disease before symptoms Hermes Toros Xavier, Vanessa Maria Caetano Soares

Abstract

A new study points to the possibility that a simple linguistic test developed using artificial intelligence is capable of predicting, with a high level of accuracy, which cognitively normal individuals will develop Alzheimer's disease.

In this predictive model, linguistic performance and its variables derive from the interpretation of a situation outlined in a chart, whose written responses are analyzed and compared with the clinical and neuropsychological variables of each individual. The average time to diagnosis of Alzheimer's was 7.59 years. The test accuracy was 70% Significant

These results suggest that language performance in the analysis of natural situations may reveal early signs of progression to Alzheimer's, before clinical impairment, which leads to diagnosis.