A mechanistic account of value computation in the human brain

Abstract

To make decisions based on the value of different options, we often have to combine different sources of probabilistic evidence. For example, when shopping for strawberries on a fruit stand, one uses their color and size to infer-with some uncertainty-which strawberries taste best. Despite much progress in understanding the neural underpinnings of value-based decision making in humans, it remains unclear how the brain represents different sources of probabilistic evidence and how they are used to compute value signals needed to drive the decision. Here, we use a visual probabilistic categorization task to show that regions in ventral temporal cortex encode probabilistic evidence for different decision alternatives, while ventromedial prefrontal cortex integrates information from these regions into a value signal using a difference-based comparator operation.I n our everyday lives we often have to combine different sources of probabilistic information to make decisions that are more likely to produce desirable outcomes. For instance, imagine choosing strawberries on a fruit stand on the basis of their size and color, trying to increase the likelihood your strawberry smoothie will taste delicious. Despite much progress in understanding the neural systems that mediate reward-and value-based decision making in humans (1-8), including recent reports of value-based modulations in sensory cortex (9-11), it remains unclear how the brain represents different sources of probabilistic information and how they are used to compute the value signal necessary to make a decision.Research on perceptual decision making has already established that category-selective regions in sensory cortex encode the amount of perceptual information (i.e., sensory evidence) used in the decision process (12-17). To date it remains unclear, however, whether sensory regions also represent the amount of probabilistic reward information (i.e., probabilistic evidence) associated with different decision alternatives during value-based decisions. The lack of empirical affirmation that such regions exist has made it difficult to provide a mechanistic account of how different sources of probabilistic evidence are combined to compute value. Despite the fact that several studies on valuebased decision making have consistently implicated the medial prefrontal cortex in encoding expected value signals (18-27), it remains unknown whether it is directly involved in computing the value signal needed to make the decision (by combining different sources of probabilistic evidence) or whether it merely reflects the consequence of the decision process.Notably, the only available empirical evidence that could provide mechanistic insights into the computation of choice values comes from work on perceptual decision making. Specifically, this line of research has shown that, for binary perceptual choices, decision variables are computed by integrating the difference of the outputs of neural populations, tuned to sensory evidence for each decision alternative (12)(13)(14...

Neurobiology of Value Integration: When Value Impacts Valuation

“…This is in line with evidence suggesting that the ventral part of the medial PFC and OFC encodes the reward value of choice options Daw et al, 2006;Kim et al, 2006;Plassmann et al, 2008;Gasic et al, 2009;Hare et al, 2009;Kahnt et al, 2010Kahnt et al, , 2011Philiastides et al, 2010;Smith et al, 2010). Talmi and colleagues (2009) have demonstrated that activity in this region increases with rewards and is attenuated by the prospect of pain.…”

Quantifying the time for accurate EEG decoding of single value-based decisions

“…A widespread network of brain regions is involved in this process, with the ventromedial prefrontal cortex being a value integration region (Basten et al, 2010;Philiastides et al, 2010), and parietal regions accumulating results of this integration (Basten et al, 2010;Gluth et al, 2012;Hunt et al, 2012). Because an accumulation mechanism implies a dynamic process which evolves over time, assessing the temporal dynamics of decision-making can be highly informative of the underlying neural process.…”

“…The utility of studying single-trial activity has been highlighted by recent studies in value-based decision-making (Billeke et al, 2013;Gluth et al, 2013;Hunt et al, 2012;Philiastides et al, 2010). In particular, these studies have isolated readiness potentials in the EEG signal which reflect the emergence of value-based decisions (Gluth et al, 2013) or, in the case of MEG, have allowed tracking the information flow in brain regions linked to different stages of a decision (Hunt et al, 2012).…”

Multialternative drift-diffusion model predicts the relationship between visual fixations and choice in value-based decisions

“…The DDM has also been shown to provide highly accurate descriptions of accuracy and response times in domains such as memory retrieval and decision-making, where the stimuli are not explicitly stochastic (17,19,20,(22)(23)(24)(25)(26)(27)(28); this suggests that these decisions might be made using a similar process of random information accumulation and integration. To see why, consider the case of binary value-based choice.…”

“…The standard drift-diffusion model (DDM), as well as closely related versions, such as the leaky competitive accumulator (LCA) model (3,4,9), have been highly successful in providing quantitative explanations of the psychometrics, chronometrics, and neurometrics of binary perceptual choice (2,(10)(11)(12)(13)(14)(15)(16), and more recently in binary value-based choice (17)(18)(19)(20). These models assume that decisions are made by accumulating stochastic information over time until the net evidence in favor of one option exceeds a prespecified threshold.…”