Early Detection of Alzheimer’s Disease using Non-Linear Analysis of Electroencephalogram (EEG) by Artificial Neural Network (ANN)

• Published in: NeuroQuantology Vol. 20; no. 9; p. 4031
• Main Authors: Soni, Sweta, Mathur, Shikha, Saini, Manu
• Format: Journal Article
• Language: English
• Published: Bornova Izmir NeuroQuantology 01.01.2022
• Subjects: Artificial neural networks; Classifiers; Disease control; Electroencephalography; Fractal analysis; Fractal geometry; Fractals; Hyperventilation; Maneuvers; Memory tasks; Neural networks; Nonlinear analysis; Older people; Patients
• ISSN: 1303-5150
• DOI: 10.14704/nq.2022.20.9.NQ44461

The aim of present Electroencephalography (EEG) study was to investigate the role of non-linear analysis via Artificial Neural Network (ANN) for early detection of Amnesic Alzheimer’s disease patient and normal elderly subjects. Katz’s Fractal Dimension (FD) analysis was carried out from the EEG data of the samples (20 patients with Alzheimer’s disease and 20 healthy controls) for the EEG manoeuvres (Eye closed, Eye open, hyperventilation, Memory task). EEG was recorded at rest and Mini Mental State Examination (MMSE) was performed before EEG recording in patients with Alzheimer’s disease patient and normal elderly subjects. The EEG bands so evaluated were delta (0.2-3.9 Hz), theta (4.0-7.9 Hz), alpha (8.1- 12.9 Hz), beta (13.0 -30.0 Hz) and gamma (30.1-80 Hz). A significant decrease MMSE score was observed in patients when compared to controls (p=0.000). In the present study Artificial Neural Network (ANN) classifier was used for the diagnosis task. It has been observed that the highest classification accuracy was achieved as 97.94% (trained data set) in eye closed EEG maneuvers in Katz’s Fractal Dimension (FD) in the present study. These results suggest that ANN classifier can detect Amnesic Alzheimer’s patients from normal elderly subjects.