Neurons in human pre-supplementary motor area encode key computations for value-based choice

Abstract

Adaptive behavior in real-world environments demands that choices integrate over several variables, including the novelty of the options under consideration, their expected value, and uncertainty in value estimation. We recorded neurons from the human pre-supplementary motor area (preSMA), ventromedial prefrontal cortex (vmPFC) and dorsal anterior cingulate to probe how integration over decision variables occurs during decision-making. In contrast to the other areas, preSMA neurons not only represented separate pre-decision variables for each choice option, but also encoded an integrated utility signal and, subsequently, the decision itself. Conversely, post-decision related encoding of variables for the chosen option was more widely distributed and especially prominent in vmPFC. Our findings position the human preSMA as central to the implementation of value-based decisions.

Encoding of predictive associations in human prefrontal and medial temporal neurons during Pavlovian conditioning

“…One possible interpretation is that weakly coding neurons provided a distributed code for discriminating between positive and neutral outcomes when considered jointly (Stefanini et al 2020). These results are consistent with previous findings on tracking feedback in medial frontal cortex neurons (Fu et al 2019, Aquino et al 2021).…”

“…25,[37][38][39][40][41][42][43][44][45][46] We also obtained recordings from the hippocampus, a brain region that has been previously implicated in model-based inference, 28,47,48 as well as from the anterior cingulate cortex and pre-supplementary motor cortex (preSMA) in dorsomedial frontal cortex, a region that has been previously found to prominently encode error signals. [49][50][51][52][53] We hypothesized that neurons in ventromedial prefrontal cortex and amygdala might play a role in encoding predictive information about stimulus identity, consistent with a role for these structures in stimulus-stimulus learning, and in model-based inference more generally. 18,24,25,28 In addition to testing for neurons correlating with stimulus identity we also tested for neurons encoding stimulus-value associations.…”

“…We obtained simultaneous recordings from neurons in the ventromedial prefrontal cortex (including neurons on the medial orbital surface) as well as the amygdala, two brain areas that have long been suggested to play a role in value-based learning and Pavlovian learning in particular (Schoenbaum et al 1998, Gottfried et al 2003, Holland & Gallagher 2004, Hampton et al 2006, Ostlund & Balleine 2007, Kennerley & Wallis 2009, Salzman & Fusi 2010, Wang et al 2017, Rudebeck et al 2017, Paton et al 2006. We also obtained recordings from the hippocampus, a brain region that has been previously implicated in model-based inference (Wimmer & Shohamy 2012, Boorman et al 2016, Miller et al 2017, as well as from the anterior cingulate cortex and pre-supplementary motor cortex (preSMA) in dorsomedial frontal cortex, a region that has been previously found to prominently encode error signals (Bonini et al 2014, Fu et al 2019, Sajad et al 2019, Aquino et al 2021, Fu et al 2022.…”

A functional overlap between evidence accumulation, confidence, and changes of mind in the pre-supplementary motor area and insula

“…Activity in the pSMA reflected confidence following stimulus onset until the decision time, in line with a possible role of the pSMA in tracking the uncertainty of a task as it unfolds until a decision is reached (Aquino et al, 2023;Bonini et al, 2014;Fleming et al, 2018;Nachev et al, 2008;Ullsperger et al, 2014). This mechanism may be useful in situations where one needs to adapt behaviour before a decision, for instance when deciding to opt-out before the decision is taken if it is deemed too difficult (Kepecs et al, 2008;.…”

The neurocomputational link between defensive cardiac states and approach-avoidance arbitration under threat

“…Adapting the relative weight of potential positive vs. negative outcomes might enable more optimal decisions to approach rather than avoid when the potential reward (relative to threat) is deemed sufficiently large 28 . Such value integration has previously been suggested to be driven by the dorsal anterior cingulate cortex (dACC) 32,36,42,43 and pre-motor areas such as the supplemental motor area (SMA) 29,44 . We predict that value integration in these regions might be driven by freezing states.…”