Use of Complex Networks for the Automatic Detection and the Diagnosis of Alzheimer's Disease

Author(s)

Pineda, Aruane Mello ; Ramos, Fernando M. ; Betting, Luiz Eduardo ; Campanharo, Andriana S. L. O. ; Rojas, I ; Joya, G ; Catala, A

• Document type: Journal article
• Source: ADVANCES IN COMPUTATIONAL INTELLIGENCE, IWANN 2019, PT I; v. 11506, p. 12-pg., 2019-01-01.

Abstract

Alzheimer's disease (AD) is classified as a chronic neurological disorder of the brain and affects approximately 25 million elderly individuals worldwide. This disorder leads to a reduction in people's productivity and imposes restrictions on their daily lives. Studies of AD often rely on electroencephalogram (EEG) signals to provide information on the behavior of the brain. Recently, a map from a time series to a network has been proposed and that is based on the concept of transition probabilities; the series results in a so-called "quantile graph" (QG). Here, this map, which is also called the QG method, is applied for the automatic detection of healthy patients and patients with AD from recorded EEG signals. Our main goal is to illustrate how the differences in dynamics in the EEG signals are reflected in the topology of the corresponding QGs. Based on various network metrics, namely, the clustering coefficient, the mean jump length and the betweenness centrality, our results show that the QG method can be used as an effective tool for automated diagnosis of Alzheimer's disease. (AU)

• FAPESP's process: 16/17914-3 - Effect of antiepileptic drugs and of epileptiform discharges in the functional connectivity of patients with juvenile myoclonic epilepsy
• Grantee: Luiz Eduardo Gomes Garcia Betting
• Support Opportunities: Regular Research Grants
• FAPESP's process: 18/25358-9 - Use of complex networks for the automatic detection, diagnosis and classification of the Alzheimer's Disease
• Grantee: Andriana Susana Lopes de Oliveira Campanharo
• Support Opportunities: Regular Research Grants