Explaining Dementia Using Graph Theory and Machine Learning

Abstract

Dementia is a central nervous system neurodegenerative disease. Dementia has been discussed around different types. Among all, Alzheimer’s disease (AD) and vascular dementias (VaD) are the commonly caused disease. There is a huge research gap that graph theory has rarely been followed to explain the cognitive systems functioning particularly diseases like dementia. Thus, theoretical models of graphs can be used to interrogate the cognitive systems and the likely presence of dementia to understand the reasonings behind and thus the treatment. In this paper, three studies have been discussed in which dementia is investigated through graph theory and machine learning by using theoretical foundations to support the evidence. The first study discusses the significance of graph theory techniques and its coined ideas. There are fundamental designed parameters; connectivity, diameter vertex centrality, betweenness centrality, clustering coefficient, degree distribution, cluster analysis and graph cores. In addition to this, these are featured to analyze magneto-encephalography data to find out functional network intensity in Alzheimer’s disease affected patients. The second study explores particularly the changed structure of Alzheimer’s disease. The third study coins the significance of machine leaning philosophy that paves the way for the black box and diagnosis.