运动员动作预期的信息整合计算建模及机理研究进展

Elite athletes need to integrate contextual prior information with kinematic information to make accurate action anticipation. This paper reviews research on action anticipation based on the information integration model and its neural mechanisms. Results indicate that athletes adjust the weight of each information sources contribution to action anticipation based on its reliability. The partially observable Markov decision process model provides a mathematical framework for estimating the weight athletes assign to different information sources during action anticipation. Furthermore, the brain processes involved in integrating these information sources during action anticipation require further exploration. We hypothesize that CNV amplitude, theta oscillations, pMTG and DLPFC activation are key neural signals.