Neural correlates of evidence and urgency during human perceptual decision-making in dynamically changing conditions

Abstract

Current models of decision-making assume that the brain gradually accumulates evidence and drifts towards a threshold which, once crossed, results in a choice selection. These models have been especially successful in primate research, however transposing them to human fMRI paradigms has proved challenging. Here, we exploit the face-selective visual system and test whether decoded emotional facial features from multivariate fMRI signals during a dynamic perceptual decision-making task are related to the parameters of computational models of decision-making. We show that trial-bytrial variations in the pattern of neural activity in the fusiform gyrus reflect facial emotional information and modulate drift rates during deliberation. We also observed an inverse-urgency signal based in the caudate nucleus that was independent of sensory information but appeared to slow decisions, particularly when information in the task was ambiguous. Taken together, our results characterize how decision parameters from a computational model (i.e., drift rate and urgency signal) are involved in perceptual decision-making and reflected in the activity of the human brain.YY and AD designed research; YY and MT collected the data; YY, MD, and YZ analyzed the data and contributed methods; and YY, AD, LF, and PC contributed to result interpretation. YY drafted the initial manuscript; all authors contributed to writing of this manuscript.

Evidence and urgency related EEG signals during dynamic decision-making in humans

“…As with our previous work (Yau et al, 2020), we observed two distinct groups of individuals based on performance under the ambiguous condition ( Fig. 3) : (1) early responders (n=32), who on >=80% of ambiguous trials responded during the first two-thirds of the trial before information ramped towards one direction and (2) the rest, who were categorized as late responders (n=25).…”

“…At any given time t, the evidence E (i.e., level of information/facial emotion level) is multiplied by an attentional fixed gain term g. An intra-trial Gaussian noise variable G(0,N) with a mean of 0 and a standard deviation of N was added. A N of 6 was chosen based on previous work (Carland et al, 2015;Yau et al, 2020) and because it gave a range of simulated RTs with similar variability as the observed data in our current study. The time constant determines how far back in time sensory information is considered by the model.…”

“…Where ( ) = • and represents the urgency signal that rises from zero with a slope . The UGM assumes that evidence is multiplied by the urgency signal which increases linearly with time (Cisek et al, 2009;Yau et al, 2020). An alternative non-linear model fitted with an additional 2 nd order polynomial parameter did not significantly improve model fit, as such, we opted to keep the linear model with the minimum number of parameters.…”

“…It may be that sudden-onset and discrete trial presentation coupled with short trial times often favoured in ERP may partly obscure the dynamics of an unfolding decision process. Attempts to use functional magnetic resonance imaging (fMRI) with evidence accumulation paradigms have provided evidence that a basal ganglia based urgency signal exists in humans (Nagano-Saito et al, 2012;Yau et al, 2020), as shown in primates (Thura and Cisek, 2017), but these lack the temporal resolution to precisely resolve urgency signalling.…”

“…In order to disambiguate sensory processing, decisional evidence accumulation, and urgency, we designed a dynamic decision-making task with a slowly morphing stimulus consisting of faces whose emotions transitioned from neutral to happy or sad (Yau et al, 2020). We exploited the high temporal resolution of EEG to tease apart neural determinants of human decision formation.…”