决策中主观价值计算与整合的神经机制及其影响因素

The Neural Mechanism and Influencing Factors of the Subjective Value Computation and Integration during Decision-making

决策在人类社会发展历程中扮演着非常重要的角色,而学界对其神经机制的探讨才不过几十年的时间。价值决策理论认为,人们先计算和表征事物的价值,随后进行比较和决策。在人脑中负责主观价值计算的神经基础有腹内侧前额叶皮层、眶额皮层、杏仁核等,而负责价值整合的脑区有腹侧纹状体、腹内侧前额叶皮层、眶额皮层等。人脑将不同属性及成本按照曲线交互模式进行整合形成主观价值,并通过自我控制、注意、认知调节和情绪调节等手段,调节主观价值。未来需要强调模式分析、个体差异和网络分析对价值计算与整合的影响。

Decision-making plays a very important role in the history of human social development, and it was only a few decades ago when researchers started to explore its neural mechanism. There is a growing consensus in decision neuroscience that brain makes simple choices by first assigning a value to all of the options under consideration and then compares them, and finally choosing the biggest value option to guide decision- making. This understanding then was named the value-based decision theory. It is popular for decision researchers and was used to explain all kinds of human behaviors in the domains of decision-making about value-based decision theory. In our review, we focus on the subjective value computation and integration during inter-temporal choice and risky decision-making because there are numerous reviews about the value computation and integration of stimulus rewards including food, water, fruit juice, money, and erotic stimulus in all kinds of species. These studies emphasized that value computation is not separable with the region of ventral medial prefrontal cortex, and there are distributed neural representation to compute the subjective value along the gradient of posterior-anterior axis which is consistent with the view of evolutionary of human brain and individual development. In this review, we summarize that the neural basis of the subjective value computation is related to the ventromedial prefrontal cortex, orbitofrontal cortex, amygdala and so on, whereas the neural basis of the subjective value integration is related to the ventromedial prefrontal cortex, orbitofrontal cortex, dorsal lateral prefrontal cortex and so on in the human brain. Meanwhile, the computation related to time and risk have common neural pattern using multiple-voxel pattern analysis, and human brain can integrate distinct attributes and costs to form the subjective value using the model of curve interaction on the regions of VMPFC and OFC. Furthermore, we believe that the human brain uses distinct regions to compute the value of alternatives of which output signals are input of another region （VMPFC） to integrate and form the subjective value. We can modulate the subjective value through self-control, attention and cognitive and emotion regulation methods. Self-control changes the subjective value of rewards by executive control mechanism, which engages dorsal lateral prefrontal cortex modulates the value computation and integration that engages ventral medial prefrontal cortex. Attention is thought to play a key role in the computation of stimulus values at the time of choice, which suggests that attention manipulations be used to improve decision-making in domains where self-control lapses are pervasive. We believe that the neural mechanisms used in successful self-control can be activated by exogenous attention cues which modulate stimulus value signals, and attention-modulated relative value signals may serve as the input of a comparator system that is used to make a choice. The computational and neurobiological mechanisms of cognitive regulation during decision making use two distinct regulatory mechanisms including value modulation （changing the values assigned to stimuli） and behavioral control modulation （changing how value signals affect behavior） which is related to VMPFC and DLPFC. Future research should continue emphasizing multi-voxel pattern analysis, individual differences and network analysis on the effect of value computation and integration.