Tuesday, December 12, 2006

Brain imaging of intuition during perceptual discovery

Volz and von Cramon at the Max Planck Institute for Human Cognitive and BrainSciences in Leipzig, Germany, have done an interesting piece of work on imaging intuition, reported in the Journal of Cognitive Neuroscience. Here is their abstract:

"According to the Oxford English Dictionary, intuition is "the ability to understand or know something immediately, without conscious reasoning." Most people would agree that intuitive responses appear as ideas or feelings that subsequently guide our thoughts and behaviors. It is proposed that people continuously, without conscious attention, recognize patterns in the stream of sensations that impinge upon them. What exactly is being recognized is not clear yet, but we assume that people detect potential content based on only a few aspects of the input (i.e., the gist). The result is a vague perception of coherence which is not explicitly describable but instead embodied in a "gut feeling" or an initial guess, which subsequently biases thought and inquiry. To approach the nature of intuitive processes, we used functional magnetic resonance imaging when participants were working at a modified version of the Waterloo Gestalt Closure Task. Starting from our conceptualization that intuition involves an informed judgment in the context of discovery, we expected activation within the median orbito-frontal cortex (OFC), as this area receives input from all sensory modalities and has been shown to be crucially involved in emotionally driven decisions. Results from a direct contrast between intuitive and nonintuitive judgments, as well as from a parametric analysis, revealed the median OFC, the lateral portion of the amygdala, anterior insula, and ventral occipito-temporal regions to be activated."

The figure above indicates some of the task and stimuli conditions used. In the object condition, participants were presented with fragmented black-and-white line drawings of common objects which were subsequently fragmented with which differed in their potential to mask the drawing. In the nonobject condition, participants were presented with meaningless fragmented black-and-white line drawings. The upper left panel (A) shows an example of a coherent trial (violin), the upper right panel (B) an example of an incoherent trial. Stimuli were presented for 400 msec and participants had 2 sec to indicate whether the fragmented line drawing depicted a possible object (left response button) or an impossible object (right response button). Stimuli were neither presented repeatedly nor were images concurrently presented in different levels of fragmentation within one individual session. Participants were encouraged to base their decision on an "initial guess" whether or not the drawing was coherent.

This next figure shows MRI data. Group-averaged activations are shown on coronal, sagittal, and axial slices of an individual brain normalized and aligned to the Talairach stereotactic space. The upper left panel shows the imaging results of the direct contrast between trials that participants judged to be meaningful with trials that participants judged to be meaningless. The upper right panel shows the imaging results from the correlational analysis of the median OFC's (mOFC) time course. In the lower panel, imaging results from the parametric analysis are shown that used a performance-dependent regressor (i.e., the percentage of correct answers per level).

The data suggest that activation within the median orbito-frontal cortex reflects intuitive processing, while activation within ventral occipito-temporal regions reflects object recognition processes.


  1. Glad to run into your blog. Very interesting reading and I will definitely keep coming back.
    I have similar interests too!!

  2. great! It is gratifying to find that others find this effort useful...

  3. Deric,
    Could this be the physiological evidence for top down processing? It is almost like the median OFC acts as a gatekeeper. I wonder if an autistic brain functions the same, or if the VOT region lights up more?

    Fascinating stuff!