Neuronal substrates of haptic shape encoding and matching: A functional magnetic resonance imaging study

We used functional magnetic resonance imaging to differentiate cerebral areas involved in two different dimensions of haptic shape perception: encoding and matching. For this purpose, healthy right-handed subjects were asked to compare pairs of complex 2D geometrical tactile shapes presented in a sequential two-alternative forced-choice task. Shape encoding involved a large sensorimotor network including the primary (SI) and secondary (SII) somatosensory cortex, the anterior part of the intraparietal sulcus (IPA) and of the supramarginal gyrus (SMG), regions previously associated with somatosensory shape perception. Activations were also observed in posterior parietal regions (aSPL), motor and premotor regions (primary motor cortex (MI), ventral premotor cortex, dorsal premotor cortex, supplementary motor area), as well as prefrontal areas (aPFC, VLPFC), parietal–occipital cortex (POC) and cerebellum. We propose that this distributed network reflects construction and maintenance of sensorimotor traces of exploration hand movements during complex shape encoding, and subsequent transformation of these traces into a more abstract shape representation using kinesthetic imagery. Moreover, haptic shape encoding was found to activate the left lateral occipital complex (LOC), thus corroborating the implication of this extrastriate visual area in multisensory shape representation, besides its contribution to visual imagery. Furthermore, left hemisphere predominance was shown during encoding, whereas right hemisphere predominance was associated with the matching process. Activations of SI, MI, PMd and aSPL, which were predominant in the left hemisphere during the encoding, were shifted to the right hemisphere during the matching. In addition, new activations emerged (right dorsolateral prefrontal cortex, bilateral inferior parietal lobe, right SII) suggesting their specific involvement during 2D geometrical shape matching.