Intelligent hybrid approach for computer-aided diagnosis of Mild Cognitive Impairment

Abstract

Mild Cognitive Impairment (MCI) is a paramount nosological entity. The concept was introduced to define the clinical state of decline or the loss of cognitive abilities which implies an initial stage to severe dementia disorders. However, diagnosis of such impairment is a challenging task due to difficulties in costs, time, as well as finding qualified experts on this topic. In this paper, a hybrid intelligent approach based on symbolic and sub-symbolic machine learning techniques is proposed. It allows to analyze the results of different cognitive tests to support decisions-making by health service staff regarding the mental state of patients. The results show that the proposed approach has a high degree of effectiveness in computer-aided diagnosis of Mild Cognitive Impairment.