Study On The Neural Correlation Of Executive Functions Based On Resting-state FNIRS

Executive function (EF), referring to a set of higher order psychological processes that are involved in goal-oriented behavior, is an important component of cognitive function. EF plays vital roles in our every adaptation to the environment. Executive dysfunction not only leads to serious cognitive or social problems but also acts as the main characterization of psychiatric disorders. EF consists of a variety of cognitive components such as planning, working memory, short-term memory, inhibition, and switch. A large number of studies have consistently shown that the neural activity of prefrontal cortex (PFC) is involved in the complex cognitive functions included in EF. However, it should be noted that the existed neuroimaging studies mostly focused on the activation of the PFC elicited by EF intensive tasks. Limited existing studies have linked EF to the resting-state brain functional activity. It has been consistently shown that functional connectivity in the resting state between the sub-regions of our brains was more helpful to reveal the intrinsic neural mechanisms involved in cognitive processing without disturbance from external stimuli. In the present study, we firstly investigated the relations between the behavioral EF scores and the resting-state functional network topological properties in the PFC. Then we analyzed the prediction function of resting-state functional netw-rk topological properties for EF using least absolute shrinkage and selection operator (LASSO). The specific tasks of this paper are as follows:First, we acquired resting-stated data from ninety healthy young adults utilizing functional near infrared spectroscopy (fNIRS), and constructed complex brain functional networks in the PFC by graph theory. We found prominent small-world properties in our participants’resting-state brain functional networks.Second, we calculated three typical regional topological properties (i.e. nodal degree, nodal efficiency, and nodal betweenness centrality) and four global topological properties (i.e. clustering coefficient, characteristic path length, global efficiency, and local efficiency). Then the Pearson correlation analyses between the network topological properties and each of the five components of EF (i.e. planning, working memory, short-term memory, inhibition, and switch) were performed with the bootstrapping method. We found that only regional nodal but not global network properties showed significant correlations with behavioral EF scores. Also, different EF components were related to regional properties in different PFC areas.Third, the LASSO regression models were established with the PFC resting-state network topological properties parameters used as the explanatory variables, and scores of EF components used as the response variables. We analyzed the influence and prediction function of the resting-state network topological properties on the EF components. We found that the resting-state network topological properties have some explaining and predicting functions for EF behavior level.Our findings suggested that the PFC resting-state neural network plays an important role in individuals’performance in the executive function tasks. Further, the regional nodal properties of the resting-state functional network in the PFC can serve as neural markers for behavioral EF abilities. This study may contribute to promote us to understand the neural basis of EF, and provide the referential significance for the clinical diagnosis of executive dysfunction based on fNIRS.