Decision Making: Unraveling the enigmatic role of the subthalamic nucleus

Subpopulations of neurons in the subthalamic nucleus have distinct activity patterns that relate to the three hypotheses of the Drift Diffusion Model.
  1. Qianli Yang  Is a corresponding author
  1. Institute of Neuroscience, Key Laboratory of Brain Cognition and Brain-inspired Intelligence Technology, Center for Excellence in Brain Science and Intelligence Technology, Chinese Academy of Sciences, China

Pondering whether “to be or not to be”, or impulsively “tilting at windmills”, decision-making comes in many forms, as so eloquently demonstrated in many great works of literature. But what causes us to sometimes come to quick conclusions while we struggle to make up our minds at other times?

Making decisions involves a complex interplay between numerous neurons located in various regions of the brain. The neural circuitry involved in this process remains, however, enigmatic. To study the cognitive processes implicated in making simple two-choice decisions, researchers often employ the Drift Diffusion Model. This mathematical framework assumes that information accumulates over time until a decision threshold is reached, and has three main parameters: the drift rate (the rate information accumulates), the decision threshold (the amount of information needed to make a decision), and the non-decision time (the time taken for processes not directly related to decision-making, such as stimulus encoding and response execution; Ratcliff, 1978; Ratcliff and McKoon, 2008).

Two of the most critical brain areas involved in decision-making are the prefrontal cortex and the hippocampus. More recently, it has been proposed that a region known as the subthalamic nucleus, which is part of the basal ganglia – and is therefore involved in motor control and integration – also has an integral role. Now, in eLife, Kathryn Rogers, Joshua Gold and Long Ding from the University of Pennsylvania report new insights into how the subthalamic nucleus contributes to decision-making (Rogers et al., 2024).

Previous computational models propose three roles for the subthalamic nucleus in decision-making and how it interacts with other parts of the brain (Figure 1): the subthalamic nucleus works with the medial prefrontal cortex to set thresholds for decision-making. These thresholds determine when enough information has been gathered to make a decision, helping control impulsivity (Hypothesis 1). Through its interaction with the external segment of the globus pallidus (a component of the basal ganglia), the subthalamic nucleus helps to calibrate how different options are evaluated, making sure the choices are assessed properly (Hypothesis 2). It helps to implement step-like, all-or-none nonlinear computations to improve the basal ganglia’s efficacy in adjusting decision bounds (Hypothesis 3; Frank, 2006; Cavanagh et al., 2011; Ratcliff and Frank, 2012; Bogacz and Gurney, 2007; Coulthard et al., 2012; Green et al., 2013; Lo and Wang, 2006; Wei et al., 2015).

The Drift Diffusion Model illustrated by Don Quixote’s decision-making.

Over time (t), Don Quixote (black mask) accumulates information that allows him to decide whether he is looking at a giant (option 1) or a windmill (option 2), and act accordingly. The subthalamic nucleus is thought to influence this decision-making process in three different ways, per the Drift Diffusion Model. First (H1), by adjusting the amount of information needed to make a decision, known as the decision threshold (represented as the gap between the two red unbroken lines increasing to the gap between the two dashed lines). Second (H2), by calibrating different pieces of evidence so they can be used to evaluate the two possibilities (whether an object is a windmill or a giant). This enables the evidence accumulation process to shift from weighting unreliable perceptual evidence too heavily (represented by the solid black line) to favoring more reliable evidence later in the decision process (represented by the dashed black line). Third (H3), by helping neural pathways sent from the basal ganglia to precisely adjust the boundaries of the decision threshold.

Image credit: Qianli Yang (CC BY 4.0).

Don Quixote’s mistaking of windmills for giants can be explained by all three hypotheses. His decision threshold may have been too low, causing him to react based on initial false evidence that the windmills’ rotating blades are the giants' flailing arms (Hypothesis 1). Similarly, he may have failed to adjust his decision threshold according to his peaceful surroundings (and thus failing to realize that he was surrounded by windmills, rather than giants (Hypothesis 3)), or to properly weigh his unreliable perception of the environment around him, which favored seeing an angry giant over a harmless windmill (Hypothesis 2).

To determine which hypothesis best explains the role of the subthalamic nucleus in decision-making, Rogers et al. recorded the activity of single neurons in the subthalamic nucleus of monkeys during a visual-saccadic decision task. In the experiment, the monkeys undertook a direction-discrimination task, during which they reported the perceived motion direction of a random dot stimulus by making a rapid eye movement (saccade) towards the corresponding choice target.

This revealed three distinct neuronal subpopulations in the subthalamic nucleus, each corresponding to one of the hypothetical models: The first subpopulation showed choice- and coherence-dependent activity that ramped up during motion viewing, consistent with the theory that these neurons pool and normalize evidence-related signals (Hypothesis 2). The second subpopulation exhibited an early, sharp rise in activity that was independent of choice and coherence during motion viewing, and gradually decreased toward saccade onset. This matches the prediction that the neurons provide an early signal to suppress immature choices (Hypothesis 1). The third subpopulation showed choice- and coherence-dependent ramping activity during motion viewing and a short burst of activity for one choice just before saccade onset. This aligns with the prediction that the subthalamic nucleus balances evidence-related signals until decision time (Hypothesis 3). This heterogeneity suggests that the subthalamic nucleus influences perceptual decision-making (where sensory information is used to guide behavior) in multiple ways.

Rogers et al. then applied electrical microstimulation (weak electric currents that affect neurons near the electrode) to the subthalamic nucleus during the task to perturb its activity. This caused the biases influencing the monkeys’ choice biases to change, reduced the influence of motion strength (i.e., the speed of the moving dot) on choices, and decreased response times. Fitting the data with a Drift Diffusion Model revealed that microstimulation of the subthalamic nucleus affected the decision threshold, evidence accumulation, as well as processes not involved in decision-making.

This groundbreaking study provides new insights into the causal roles of the subthalamic nucleus in perceptual decision-making, highlighting its involvement in modulating various aspects of the decision process through distinct neural subpopulations. These findings advance our understanding of how the basal ganglia contribute to decision-making and cognitive function.

However, it remains unclear if the subthalamic nucleus interacts with other brain regions, and if potential anatomical and/or functional alterations to this region could be implicated in neurological and psychiatric disorders. Future research could explore these issues and investigate therapy avenues that target the subthalamic nucleus for conditions involving impaired decision-making, potentially offering hope for individuals like Don Quixote.

References

Article and author information

Author details

  1. Qianli Yang

    Qianli Yang is in the Institute of Neuroscience, Key Laboratory of Brain Cognition and Brain-inspired Intelligence Technology, Center for Excellence in Brain Science and Intelligence Technology, Chinese Academy of Sciences, Shanghai, China

    For correspondence
    qlyang@ion.ac.cn
    Competing interests
    No competing interests declared
    ORCID icon "This ORCID iD identifies the author of this article:" 0000-0003-4226-2319

Publication history

  1. Version of Record published: July 10, 2024 (version 1)

Copyright

© 2024, Yang

This article is distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use and redistribution provided that the original author and source are credited.

Metrics

  • 364
    views
  • 44
    downloads
  • 0
    citations

Views, downloads and citations are aggregated across all versions of this paper published by eLife.

Download links

A two-part list of links to download the article, or parts of the article, in various formats.

Downloads (link to download the article as PDF)

Open citations (links to open the citations from this article in various online reference manager services)

Cite this article (links to download the citations from this article in formats compatible with various reference manager tools)

  1. Qianli Yang
(2024)
Decision Making: Unraveling the enigmatic role of the subthalamic nucleus
eLife 13:e100598.
https://doi.org/10.7554/eLife.100598

Further reading

    1. Neuroscience
    Shanka Subhra Mondal, Steven Frankland ... Jonathan D Cohen
    Research Article

    Deep neural networks have made tremendous gains in emulating human-like intelligence, and have been used increasingly as ways of understanding how the brain may solve the complex computational problems on which this relies. However, these still fall short of, and therefore fail to provide insight into how the brain supports strong forms of generalization of which humans are capable. One such case is out-of-distribution (OOD) generalization – successful performance on test examples that lie outside the distribution of the training set. Here, we identify properties of processing in the brain that may contribute to this ability. We describe a two-part algorithm that draws on specific features of neural computation to achieve OOD generalization, and provide a proof of concept by evaluating performance on two challenging cognitive tasks. First we draw on the fact that the mammalian brain represents metric spaces using grid cell code (e.g., in the entorhinal cortex): abstract representations of relational structure, organized in recurring motifs that cover the representational space. Second, we propose an attentional mechanism that operates over the grid cell code using determinantal point process (DPP), that we call DPP attention (DPP-A) – a transformation that ensures maximum sparseness in the coverage of that space. We show that a loss function that combines standard task-optimized error with DPP-A can exploit the recurring motifs in the grid cell code, and can be integrated with common architectures to achieve strong OOD generalization performance on analogy and arithmetic tasks. This provides both an interpretation of how the grid cell code in the mammalian brain may contribute to generalization performance, and at the same time a potential means for improving such capabilities in artificial neural networks.

    1. Neuroscience
    Sanggeon Park, Yeowool Huh ... Jeiwon Cho
    Research Article

    The brain’s ability to appraise threats and execute appropriate defensive responses is essential for survival in a dynamic environment. Humans studies have implicated the anterior insular cortex (aIC) in subjective fear regulation and its abnormal activity in fear/anxiety disorders. However, the complex aIC connectivity patterns involved in regulating fear remain under investigated. To address this, we recorded single units in the aIC of freely moving male mice that had previously undergone auditory fear conditioning, assessed the effect of optogenetically activating specific aIC output structures in fear, and examined the organization of aIC neurons projecting to the specific structures with retrograde tracing. Single-unit recordings revealed that a balanced number of aIC pyramidal neurons’ activity either positively or negatively correlated with a conditioned tone-induced freezing (fear) response. Optogenetic manipulations of aIC pyramidal neuronal activity during conditioned tone presentation altered the expression of conditioned freezing. Neural tracing showed that non-overlapping populations of aIC neurons project to the amygdala or the medial thalamus, and the pathway bidirectionally modulated conditioned fear. Specifically, optogenetic stimulation of the aIC-amygdala pathway increased conditioned freezing, while optogenetic stimulation of the aIC-medial thalamus pathway decreased it. Our findings suggest that the balance of freezing-excited and freezing-inhibited neuronal activity in the aIC and the distinct efferent circuits interact collectively to modulate fear behavior.