Neuroanatomy of creativity

Abstract

Creativity has long been a construct of interest to philosophers, psychologists and, more recently, neuroscientists. Recent efforts have focused on cognitive processes likely to be important to the manifestation of novelty and usefulness within a given social context. One such cognitive process - divergent thinking - is the process by which one extrapolates many possible answers to an initial stimulus or target data set. We sought to link well established measures of divergent thinking and creative achievement (Creative Achievement Questionnaire - CAQ) to cortical thickness in a cohort of young (23.7 +/- 4.2 years), healthy subjects. Three independent judges ranked the creative products of each subject using the consensual assessment technique (Amabile, 1982) from which a "composite creativity index" (CCI) was derived. Structural magnetic resonance imaging was obtained at 1.5 Tesla Siemens scanner. Cortical reconstruction and volumetric segmentation were performed with the FreeSurfer image analysis suite. A region within the lingual gyrus was negatively correlated with CCI; the right posterior cingulate correlated positively with the CCI. For the CAQ, lower left lateral orbitofrontal volume correlated with higher creative achievement; higher cortical thickness was related to higher scores on the CAQ in the right angular gyrus. This is the first study to link cortical thickness measures to psychometric measures of creativity. The distribution of brain regions, associated with both divergent thinking and creative achievement, suggests that cognitive control of information flow among brain areas may be critical to understanding creative cognition.

Figure 1

Statistical maps (P < 0.01) of significant clusters from Monte‐Carlo simulations of the CCI–cortical thickness correlation overlaid on the FreeSurfer average subject. Color bar indicates −log(10)P, where P corresponds to the CWP values in Table I. Blue color indicates a negative correlation (decreased cortical thickness correlates with the CCI) and red indicates a positive correlation (increased cortical thickness correlates with the CCI). (a) Medial left hemisphere; (b) lateral right hemisphere; (c) medial right hemisphere.

Figure 2

Statistical maps (P < 0.01) of significant clusters from Monte‐Carlo simulations of the CAQ–cortical thickness correlation overlaid on the FreeSurfer average subject. Color bar indicates −log(10)P, where P corresponds to the CWP values in Table II. (a) Lateral left hemisphere; blue indicates a negative correlation (decreased cortical thickness correlates with the CAQ) and (b) right lateral hemisphere; red indicates a positive correlation (increased cortical thickness correlates with the CAQ).