Mild Cognitive Impairment Classification using Hjorth Descriptor Based on EEG Signal

Abstract

Electroencephalogram (EEG) has an important role for detection, classification, diagnosis and treatment of brain disorders. One indication of a brain disorder that can be diagnosed through EEG examination is Mild Cognitive Impairment (MCI). MCI can be a symptom of Alzheimer's disease (AD) at a higher level. In this paper, we apply time domain based EEG signal processing to classify these signals in MCI patients with normal controlled subjects. Hjorth Descriptor is used to obtained the signal features, namely complexity, mobility and activity. 10 EEG data consisting of 5 MCI patients and 5 normal subjects were analyzed. From the results of testing, the Hjorth parameters in normal subjects tend to have a greater value than the MCI subject.