Large-scale brain network modelling using graph-theory approach in neuroscience

From decade, neuroscientist trying to understand the brain functional dynamics of healthy individuals and the patients suffering with neuro-psychiatric disorders. In this study, we propose a large-scale brain network modelling using graph theory approach to understand the brain dynamics of healthy individual's by exploring the ageing. We used resting-state functional magnetic resonance imaging (rsfMRI) data. We selected two groups of subject's 1) young healthy subjects (mean age: 23 ±6) and 2) old healthy subject (mean age: 40±5). The rsfMRI data pre-processed for ac-pc baseline correction, realignment, registration, segmentation, normalization and band pass filtering. The preprocessed data were parcellated using 'Dosenbach-160' atlas and connectivity matrix computed using Pearson correlation. Large-scale brain network computed by network segregation (i.e. clustering coefficient), integration (i.e. participation coefficient), efficiency and small-worldness. Individual nodal graph measures were computed through integrated nodal clustering coefficient and participation coefficient. Finally, statistical analysis were carried out using two-sample t-test with FDR correction. We found the older healthy individuals have lower clustering coefficient, small-worldness and higher participation coefficient. Our finding suggests graph theory connectivity measure is a potential technique to understand the neural dynamics of ageing, cognitive processes in healthy individual and may be a potential methods to study the alter brain functions in patients with neuro-psychiatric disorders.