Dynamic modulation of decision biases by brainstem arousal systems

Jan Willem de Gee, Olympia Colizoli, Niels A Kloosterman, Tomas Knapen, Sander Nieuwenhuis, Tobias H. DonnereLife 2017 · DOI

TL;DR

This study reveals that fluctuations in brain arousal, measured by pupil size, dynamically modulate decision biases in humans. By linking pupil responses to brainstem activity and fMRI signals, the authors demonstrate that increased arousal suppresses maladaptive decision biases, offering a new explanation for behavioral variability.

What Problem Does This Paper Address?

We often make different decisions even when presented with the same information, a phenomenon attributed to "noise" in our decision-making processes. While this variability is typically seen as a nuisance, this paper explores whether it's driven by systematic, yet hidden, factors. Specifically, the authors investigate the role of phasic arousal – rapid, transient increases in alertness – in influencing decision biases. They hypothesize that the brainstem's arousal systems, which control our overall alertness, are a key source of this trial-to-trial variability in decision-making. Understanding this link could help explain why our choices fluctuate and how the brain might optimize decision-making under uncertainty.

Methods

The researchers employed a multi-modal approach:

  • Pupillometry: They tracked pupil diameter, a well-established peripheral marker of phasic arousal, during various sensory-motor decision tasks.
  • fMRI: Functional Magnetic Resonance Imaging was used to measure brain activity, focusing on both early visual cortex and higher-order frontal and parietal regions.
  • Computational Modeling: A drift-diffusion model was fitted to reaction time distributions to understand the underlying decision-making dynamics.
  • Brainstem Imaging: Specialized fMRI techniques were used to specifically target and analyze activity in brainstem nuclei known to be involved in arousal.
  • Statistical Analysis: A range of statistical methods, including linear regression, permutation tests, and hierarchical Bayesian modeling, were used to analyze the data and establish relationships between variables.

Datasets

The study utilized data from four independent sets:

  1. Main fMRI experiment: Fifteen healthy subjects performed a visual contrast detection task with concurrent pupillometry and fMRI.
  2. Re-analyzed behavioral dataset: Data from 21 subjects who previously performed a similar visual contrast detection task were re-analyzed.
  3. Auditory tone detection task: 24 subjects performed an auditory task.
  4. Visual random dot motion task: 15 subjects performed a visual discrimination task.

No new benchmarks were introduced; existing behavioral tasks were adapted and analyzed with novel methods.

Key Insights and Ideas

  • Pupil size as a real-time arousal index: The study leverages the well-established link between pupil dilation and brain arousal to track dynamic changes in alertness during decision-making.
  • Brainstem's role in modulating bias: The findings strongly implicate specific brainstem nuclei, particularly the locus coeruleus (LC), as key players in regulating decision biases.
  • Arousal as a bias reducer: The core insight is that increased phasic arousal, signaled by larger pupil responses, actively reduces maladaptive decision biases, pushing choices towards a more neutral or optimal state.
  • Choice-specific cortical signals: Phasic arousal doesn't just affect general alertness; it specifically modulates neural signals in frontal and parietal cortices that encode the decision itself.

Key Results

  • Reduced decision bias with higher arousal: Larger task-evoked pupil responses (TPRs) consistently predicted a reduction in decision bias across different sensory modalities and task protocols. This effect was particularly pronounced in the fMRI dataset, where high TPR trials showed criteria near zero, indicating an unbiased decision.
  • Brainstem nuclei predict arousal and bias: Responses in a network of neuromodulatory brainstem nuclei (LC, VTA, SN, BF) were strongly correlated with TPR and, crucially, predicted the reduction in decision bias.
  • Arousal impacts evidence accumulation: Drift-diffusion modeling revealed that increased arousal primarily affected the "drift criterion" parameter, indicating a bias towards one choice option during evidence accumulation, rather than altering the starting point or drift rate.
  • No boost in sensory responses: Phasic arousal did not enhance sensory responses in early visual cortex, suggesting its effect is downstream of initial sensory processing.
  • Modulation of choice-encoding signals: Choice-specific signals in frontal and parietal cortex were reliably modulated by TPR, showing a greater tendency towards the 'yes'-choice under high arousal, even after accounting for stimulus evidence.

Limitations and Future Work

  • Brainstem fMRI challenges: Imaging the brainstem with fMRI is inherently difficult due to physiological noise and the small size of nuclei. While the authors employed advanced techniques to mitigate these issues, some uncertainty regarding spatial specificity remains. They emphasize that their conclusions are based on the distribution of correlations across brainstem structures, not solely on individual nuclei.
  • Single-unit measurements: To definitively link brainstem activity to specific neuron types (e.g., noradrenergic neurons), future studies would require single-unit recordings.
  • Dynamic tracking of arousal: The authors suggest that future work should aim to track TPR-linked changes in brainstem and cortical state in real-time to further elucidate the dynamic interplay.
  • Top-down vs. bottom-up influences: While the study suggests arousal might suppress top-down signals, further research is needed to fully understand how neuromodulatory signals translate into specific effects on choice behavior.
  • Mechanisms of bias suppression: Future research could explore the precise synaptic mechanisms by which neuromodulators influence evidence accumulation and bias reduction.

Summary updated

Abstract

Decision-makers often arrive at different choices when faced with repeated presentations of the same evidence. Variability of behavior is commonly attributed to noise in the brain's decision-making machinery. We hypothesized that phasic responses of brainstem arousal systems are a significant source of this variability. We tracked pupil responses (a proxy of phasic arousal) during sensory-motor decisions in humans, across different sensory modalities and task protocols. Large pupil responses generally predicted a reduction in decision bias. Using fMRI, we showed that the pupil-linked bias reduction was (i) accompanied by a modulation of choice-encoding pattern signals in parietal and prefrontal cortex and (ii) predicted by phasic, pupil-linked responses of a number of neuromodulatory brainstem centers involved in the control of cortical arousal state, including the noradrenergic locus coeruleus. We conclude that phasic arousal suppresses decision bias on a trial-by-trial basis, thus accounting for a significant component of the variability of choice behavior.
BibTeX
@inproceedings{gee2017dynamic,
  author = {Jan Willem de Gee and Olympia Colizoli and Niels A Kloosterman and Tomas Knapen and Sander Nieuwenhuis and Tobias H. Donner},
  title = {{Dynamic modulation of decision biases by brainstem arousal systems}},
  booktitle = {eLife},
  year = {2017},
  doi = {10.7554/elife.23232},
}
Figure