Neural correlates of distorted body representations underlying tactile distance perception

Tactile distance perception is believed to require that immediate afferent signals be referenced to a stored representation of body size and shape (the body model). For this ability, recent studies have reported that the stored body representations involved are highly distorted, at least in the case of the hand, with the hand dorsum represented as wider and squatter than it actually is. Here, we investigated whether we can elicit the same type of distortions by using an MR-compatible system for somatosensory stimulation (airpuff) and in turn define the neural basis of this phenomenon. In each trial, participants received sequential tactile stimulations via airpuff on two out of nine points on a squared grid (5x5cm) on the dorsum of heir hand. In a behavioural experiment (Exp. 1) they have to estimate the distance between the two points by adjusting the length of a visually-presented line on the screen. The technique of multidensional scaling (MDS) was used to reconstruct the grid of the actual and perceptual maps of points’ configuration. Procrustes distance showed a stretch of the estimated map in the mediolateral hand axis. These behavioural results are consistent with previous studies, in which there were clear biases to overestimate distances oriented along the medio-lateral axis of the hand compared to the proximo-distal axis. In order to determine the neural correlates of these body distortions, we performed an fMRI study (Exp. 2) - functional scans in a random block design and a tactile localiser - in which the same nine points were this time individually stimulated. For each participant, we measured the response pattern generated by each stimulated point by contrasting each stimulation condition (e.g., repeated stimulation of point 1, 2, etc) against baseline. In order to relate the representations between the different points and to computational models, we compare response-pattern dissimilarity matrices in several regions of interests (ROIs). ROIs were defined using both anatomical (i.e., probabilistic map) and functional (i.e., functional localiser) criteria at individual level. Similar to the behavioural experiment, we used the technique of MDS to reconstruct the spatial relationships between the points using the values from the dissimilarities matrices. We expect to find a similar pattern of results (i.e., distortions) in the brain areas, of the tactile representation processing, that generate the distorted body representations we found in the behavioural experiment