Abstract: Humans are social animals. Regardless of whether you are talking to others, working together on a project, or having a conversation, you are practicing social interactions. According to the content, social interactions can be divided into non-verbal interactions, verbal interactions, and mixed interactions. This process also includes verbal or non-verbal emotional exchange. Understanding the brain mechanism of social interaction could potentially provide guidelines for better social interactions. As early as 2002, Brothers has put forward “the social brain hypothesis” according to the complexity and diversity of social activities of primates. He believes that animals, including primates, have the neural mechanism to know and understand other individuals, and the mechanism can help individual to quickly deal with all kinds of information produced in the process of interacting with other individuals in the society, so as to achieve the purposes of communication with high quality. However, due to the limitation of the previous research technology, most of the experimental paradigms presented individual stimulus through computer screen, or let the participants interact with an imaginary partner, and then study the brain response and neural mechanism of the target subjects. Although researchers have explored a series of brain regions that constitute the human brain by studies of lesion subjects, brain imaging, and electrophysiological techniques, such as the amygdala, orbitofrontal cortex, ventromedial prefrontal cortex, the mirror neuron system and so on, it is still not clear that how to play and how to transfer the information between the brain in the dynamic and real-time social interaction. Functional near-infrared spectroscopy imaging is a convenient and friendly technology which provides an important technique support for studying the brain mechanism of social interaction in the natural setting. fNIRS is an emerging brain functional imaging technology, which uses blood volume and oxygenation as the carrier of information in the brain tissue, in order to understand local brain activity by measuring distribution and changes of blood volume and blood oxygen in the cerebral cortex. The present study describes the application of fNIRS in understanding the neural basis of three types of social interactions: verbal, nonverbal, and mixed social interactions. In the past decade, fNIRS hyperscanning were used by many researchers to measure changes in several cortical regions during social interaction, including frontal lobe, parietal lobe, and temporal lobe. However, largely because of different tasks used and different brain regions measured, their results are not consistent. We discuss the inconsistency of results and suggest the future research directions. First, future studies should improve the technology of functional near-infrared spectroscopy imaging itself and polish the data analysis and evaluation methods. Future studies need to develop advanced fNIRS devices to cover larger brain regions. We should construct a generalized predictive model to investigate the deep brain activity inferred from surface brain signals according to functional connectivity. Second, future research should refine and enrich the task paradigm used to study the brain mechanism of social interaction. And future research should use a more standardized control condition to control other variables. Finally, we need to focus on the application of technology integration in this field, providing multimodal evidence for the exploration the brain mechanisms of social interaction. By taking advantage of multi-modal functional brain imaging techniques, providing converging evidences, we could learn deeper of the neural basis of social interaction.

Key words: functional near-infrared spectroscopy, social brain, verbal interaction, nonverbal interaction, mixed interaction

摘要： 功能性近红外光谱技术是一种以大脑组织中的血容和血氧为信息载体，通过测量大脑皮层中血容和血氧的分布和变化情况来了解局部脑活动的技术手段。近年来该技术在双脑研究非言语互动、言语互动、混合互动等社会互动脑机制领域做出了重要贡献。为了更清晰地揭示自然情境中社会互动的神经机制，未来尚需进一步改进近红外光谱技术本身，开发有效的数据分析方法，细化互动任务类型，丰富完善实验范式，促进多种脑成像技术的融合。

关键词: 功能性近红外光谱技术, 社会脑, 言语互动, 非言语互动, 混合互动