Effective Connectivity of the Large-Scale Brain Networks for Procrastination: A Granger Causality Analysis
CHEN ZhiYi,SU Ti,ZHANG Rong,XU Ting,ZHOU Fan,YANG YaQi,FENG TingYong
DOI: https://doi.org/10.1360/ssv-2019-0076
2019-01-01
Scientia Sinica Vitae
Abstract:Procrastination is increasingly becoming one of the biggest maladies on a global scale; it refers to a situation in which one is more likely to postpone courses in decision-making and the remaining out-of-control factors at the expense of long-term profits. Previous studies indicated that procrastination is not only linked to the mental health and subjective well-being of a person but also predisposes numerous aspects of neuropsychiatric diseases. Furthermore, emerging studies have documented the neural substrate of procrastination as a functional connectivity of the large-scale brain functional networks, providing useful evidence to clarify the core components of procrastination. However, the manner in which these networks function to form connectivity patterns for determining procrastination still needs to be elucidated. In this regard, the Granger causality model (GCM) was utilized to investigate the neural patterns of procrastination in relation to an effective-connection-based scheme. Thus, on the basis of the Informax algorithm, neuroimages were used in the independent component (IC) analysis to extract and further identify a portion of ICs as networks of interest, including the cingulate control executive network (CON), subcortical network (SCN) and salience network (SAN). Furthermore, GCMs were used to estimate the effective connectivity of these large-scale brain function networks, and the connections were further correlated for determining the procrastination. Based on the machine learning algorithm, support vector regression that relies on the radial basis function was derived for predicting the effective connectivity patterns of procrastination. The results indicated that the effective connectivity of the GCM of CON-SCN was negatively linked to procrastination, and CON was identified as the Granger reason for this model. In addition, similar findings were observed with respect to the effective connectivity of the GCM of CON-SAN; CON was also recognized to be the Granger reason for this model. Finally, the calculated predictive accuracy of the connectivity patterns with respect to procrastination denoted the significant predictive power of these connectivities with a high accuracy ( R 2=0.66, MSE=0.120). In conclusion, this study may confirm the notion that the top-down regulation of CON in both SCN and SAN can be considered to denote the neural understructure of procrastination and provide evidence on the robustness of linking self-control with the reward-related processes and task-evoked emotions and procrastination.