Understanding Face Perception using Minimum Spanning Tree Dynamic Brain Networks

Abstract Face perception in the brain is considered one of the most complex functional processes of the brain. Many functional connectivity studies were carried out in the past to understand it. Despite several efforts, it is not well understood. In the present study, we have attempted to study how face perception is different from objects through dynamic functional connectivity using unbiased MST graphs. The EEG data was collected for 15 healthy subjects while presenting face and object images. A detailed dynamic MST analysis was performed for faces as well as objects. MST graph theoretical properties were evaluated for every 100 ms of time-length of stimulus presentation in different EEG frequency bands (i.e., Delta, Theta, Alpha, Beta, and Gamma). Our results indicate that the change in the face processing network occurs dynamically over EEG frequency bands. More star-like MST trees were found for faces in the Delta, Theta, Alpha, and Beta frequency bands. The obtained MSTs also reflected a significant difference between face and object perception in the brain. Furthermore, our findings infer that there exists an efficiently organized, fully connected network for face processing in the brain, composed of many sub-networks responsible for high-speed information processing. Thus, the results show novel findings of faster and more efficient communication between nodes for faces as compared to objects. The MST dynamic brain network measures calculated for each band and time segment form a complete framework for the evaluation of face processing in the brain and thus provide a new insight into the existing face processing paradigm. Therefore, this study including temporal changes and the graphical importance of brain functional networks related to face-processing, can help in identifying pathological brain networks for neurological disorders.