Cell . . . The neurobiological basis of decision-making judgment.

In the face of uncertainty, any organism that can make statistical calculations gains an adaptive advantage.generally, we use confidence to evaluate the probability of making a correct choice.accurate estimation of the confidence of a choice can bring benefits to biological behaviors, such as deciding whether to fight or run in the face of danger, such as which beverage to choose when entering a coffee shop or how to manage stock investment, etc.as human beings, we have been assessing our confidence in perception, memory and decision-making, which is part of a process of self-monitoring and reflection called metacognition, which is very personal and subjective.therefore, confidence judgments are considered to be a core aspect of metacognition, and the realization of this behavior may also require the cooperation of neural structures unique to advanced primates [1].although many organisms use probability calculation to guide their behavior, the neurobiological structure involved in it is still unclear.recently, Adam kepecs research group of Cold Spring Harbor Laboratory in the United States published a paper entitled behavior - and modality general representation of confidence in orbitofrontal cortex in cell. It was found that nerve element signals in orbital frontal cortex determine the sensory pattern of correct probability judgment in rodents, so as to support animal's metacognitive ability.then, what are the characteristics of the neurobiological basis for confidence judgment? First of all, neural activity needs to reflect the calculation of confidence, such as the statistical probability that a behavior or proposition is correct.secondly, neural activity should be associated with final behavioral decisions on the basis of confidence.however, the two criteria of neural activity and confidence, and then to the occurrence of guided behavior, are only based on computational considerations, but the realization of metacognitive ability needs other factors.the premise of metacognitive ability realization is an abstract and centralized confidence judgment, that is to say, the behavior guided by the participation of different information sources and different types of sensory organs needs to be generalized.therefore, the authors hope to establish a behavioral task model to test the activity of a single neuron in the orbitofrontal cortex, which is a key area for value-based decision-making.the model needs to satisfy the following conditions: 1) statistical calculation of correct probability; 2) prediction of confidence oriented selection; 3) independence of information sources; 4) generalization of multiple confidence oriented behaviors.based on the above four criteria, the authors developed a behavioral task in which neural coding characteristics of individual neurons in the orbitofrontal cortex could be evaluated by measuring decision selection and confidence reports based on two sensory patterns (Fig. 1).in Fig. 1, the authors trained rats to perform auditory and olfactory decision-making, and urged them to invest decision-making time according to the proportion of decision-making confidence.after making a decision, the mice obtain uncertain and delayed rewards by putting in time, thus providing a report of post decision confidence (Fig. 1) [1].the authors found that this task design can ensure that the binary decision-making of two sensory organs can be obtained simultaneously, and the time input after decision-making in a single experiment can be continuously measured. furthermore, the authors hope to confirm whether the activity of a single neuron is determined by the calculation of the confidence level of statistical decision. the authors recorded the activity of 1211 well separated single neurons from the middle orbitofrontal cortex of six rats using custom-made 32 or 64 channel tetrode microdrivers. the authors found that simple and correct selection of neurons had the highest activity and average confidence, and vice versa. in addition, the realization of neuron coding decision-making in a single orbitofrontal cortex is widespread in both senses. moreover, the activity of neurons in the orbitofrontal cortex can be used to predict the time input and choice updating of confidence guidance among different senses. in general, the kepecs team found a correlation between the activity of neurons in the orbitofrontal cortex of rats and the cross sensory decision-making confidence, thus supporting the metacognitive ability of rodents. the time input of rats in olfactory and auditory decision-making reflects the confidence judgment of rats. however, the ability to calculate the confidence level encoded by orbitofrontal cortex neurons is widespread in different sensory decisions. every decision people make is accompanied by the prediction and judgment of the results caused by the behavior, and this feeling will also affect the subsequent behavior. people tend to invest more time, energy or money when they are more sure of the rewards behind their actions. and this work uncovers the neurobiological structural basis of participation in the calculation of confidence and the realization of metacognitive ability in the process of decision-making. original link: plate maker: references 1. Dehaene, S., Lau, H. & amp; kouider, S. what is science, and could machines have it? Science (New York, N.Y.) 358, 486-492, doi:10.1126/science.aan8871 (2017).2 、 Lak, A. et al. Orbitofrontal cortex is required for optimal waiting based on decision confidence. Neuron 84, 190-201, doi:10.1016/j.neuron.2014.08.039 (2014). Plate maker: sy