Novel vibrotactile discrimination task for investigating the neural correlates of short-term learning with fMRI.

Innovative perceptual-motor learning paradigms applicable for functional magnetic resonance imaging (fMRI) offer much potential for elucidating the specific cortical mechanisms that underpin short-term learning. In this study, a novel, fMRI-compatible, vibrotactile discrimination task, adapted from a tactile version of the Morse-code, was introduced. Uniquely, this task featured distinct components of training and testing, such that cortical changes associated with these modalities of learning may be dissociated. During testing, the ability to correctly associate matching and non-matching vibrotactile/visual stimulus pairs was assessed. Initial testing in naïve healthy subjects (n = 23) revealed a training-dependent decrease in discrimination error rates and discrimination reaction time over the course of a single fMRI session. The rate and extent of learning were significantly decreased when the complexity of vibrotactile stimuli to be discriminated was increased. In a subgroup of subjects (n = 15) who participated in repeated testing, it was revealed that both initial testing and retesting sessions were characterized by similar within-subject training-related behavioral properties, when different vibrotactile patterns of similar challenge level were presented for the retesting session. Preliminary functional imaging data from a single subject case-study revealed task-related cortical activations over a widely distributed frontoparietal network, which demonstrated spatial consistency within- and also between-sessions (test-retest). Observed behavioral and cortical properties suggest that the current methodology may be suitable for assessing neural changes linked to short-term vibrotactile learning. In addition, demonstrated test-retest capability of the proposed task may uniquely permit applications where test conditions are to be manipulated within-subjects.