Perceptual Decision Making in the Human Brain

Perceptual decision making deals primarily with the evaluation of sensory information for the decision to be made. Furthermore, because our sensory systems tend to be noisy and physical stimuli can often not be represented unambiguously, any interpretation of sensory information requires - an often implicit - decision process.

Neurophysiological studies in monkeys and humans have recently begun to investigate perceptual decision processes mainly in the visual and somatosensory domain. These studies have revealed activity patterns in sensory brain areas that showed a stronger contingency on the behavioural response the subject has made (to that stimulus) than on the physical stimulus presented.

Single unit recording studies in monkey cortex (Mazurek et al., 2003) have given us insights into the mechanics how such decisions are formed. In particular, they have shown that at least in certain scenarios using temporally extended visual stimulus presentation, sensory information (from lower level sensory brain areas) is integrated in higher level brain structures until a certain threshold of activity is reached and the response is more or less deterministically executed.

Such processes (called diffusion-to-barrier-process) have been investigated for more than 30 years in the field of mathematical psychology and are well capable of reproducing the statistics of human behavioural data (such as response-time-distributions, error-rates, speed-accuracy-trade-off) accurately. Studies in monkeys and more recently in humans, have now provided first evidence that there is a direct neuronal implementation of such diffusion mechanisms (Philiastides et al., 2006).

In our experiments we are investigating perceptual decision making processes with non-invasive whole-brain recording techniques (EEG, MEG & fMRI) in the human brain. While the monkey studies provided very detailed insights into the behaviour of single neurons during such tasks, only a small part of the brain can be looked at with single unit recording techniques and many of the interactions these neurons are involved in with other brain areas remain undiscovered.

In a previous fMRI study we found that even for complex object categories (such as houses and faces), the comparison of the outputs of different pools of selectively tuned neurons could be a general mechanism by which the human brain computes perceptual decisions and that this difference operation localizes to the left posterior dorsolateral prefrontal cortex  (Heekeren et al., 2004). Furthermore, when subjects performed the same direction‑of‑motion discrimination task used in the single-unit recording studies, the identical brain region showed a greater fMRI response to stronger motion signals whether the subjects responded with a button press or an eye movement (Heekeren et al., 2006).

It is well known that not only top-down-information is relevant in complex and even in simple-decision processes, but this is highly influenced by contextual variables such as prior probabilities of the possible events, the outcome associated with a given response, or speed accuracy trade-off. Such variables are likely not encoded in low-level sensory areas (which may often be sufficient to differentiate between different stimulus alternatives) but their empirically proven influence on the decision process requires interaction of a multitude of different areas such as motor-planning regions, the reward system, and brain regions involved in affective processing. This, studying this complex system requires whole-brain recording techniques.

EEG/MEG is a technique that covers signals generated from the whole cortex, while being less sensitive to subcortical structures. It provides accurate temporal information about the neural dynamics involved in the sensory-motor-conversion processes described above (e.g. Bauer et al., 2006). In recent years several techniques (source analysis techniques such as beamformers, minimum-norm-estimates as well as decomposition techniques like ICA) have been developed that allow at least some inference of human electrophysiological data on the level of neural sources.

Our aim is to investigate the influence of both stimulus-intrinsic parameters (certainty, noise) as well as contextual parameters, as reward and prior-probabilities and task-demands like attention on local processing in individual brain regions. The interaction of different brain regions is of particular interest on this context.

Therefore we will investigate the behaviour of classical ERP components as well non-phase-locked induced oscillatory activity in such tasks. Recently several techniques have been developed to analyze coupling between brain regions and interactions in general.

Another focus will be to link the electrophysiological data on the single trial level to the behavioural outcomes as well as experimentally controlled stimulus parameters. Furthermore we aim to link parametric models (such as the diffusion model) with electrophysiological data to identify the components necessary for such processes.


Hauke Heekeren
Markus Bauer
Hermine Wenzlaff

Key References

Heekeren, H. R., Marrett, S., Bandettini, P. A., & Ungerleider, L. G. (2004). A general mechanism for perceptual decision-making in the human brain. Nature, 431 (7010), 859-861.

Heekeren, H. R., Marrett, S., Bandettini, P. A., & Ungerleider, L. G. (2006). Involvement of human left dorsolateral prefrontal cortex in perceptual decision-making is independent of response modality. Proceedings of the National Academy of Sciences of the United States of America, 103, 10023-10028.

Mériau, K., Wartenburger, I., Kazzer, P., Prehn, K., Lammers, C. H., van der Meer, E., Villringer, A., & Heekeren, H. R. (2006). A neural network reflecting individual differences in cognitive processing of emotions during perceptual decision-making. NeuroImage, 33, 1016-1027.

Bauer, M., Oostenveld, R., Peeters, M., & Fries, P. (2006).Tactile Spatial Attention Enhances Gamma-Band Activity in Somatosensory Cortex and Reduces Low-Frequency Activity in Parieto-Occipital Areas. Journal of Neuroscience, 26, 490–501.

Philiastides, M. G., Ratcliff, R., & Sajda, P. (2006). Neural representation of task difficulty and decision making during perceptual categorization: a timing diagram. Journal of Neuroscience, 26, 8965-75.