Tactile Stimulation Designs Adapted to Clinical Settings Result in Reliable fMRI-based Somatosensory Digit Maps

Abstract

A wide range of neurological diseases with impaired motor functioning of the upper extremities are accompanied by impairments of somatosensory functioning, which are often undescribed but can provide crucial information for diagnostics, treatment selection, and follow-up. Therefore, a reliable description of the functional representation of the digits in the somatosensory cortex would be a highly valuable, but currently lacking, tool in the clinical context. Task-based functional Magnetic Resonance Imaging of passive tactile stimulation provides an indirect, but valid description of the layout of the digit map in the primary somatosensory cortex. However, to fulfill the specific requirements for clinical application, the presently established approaches need to be adapted and subsequently assessed for feasibility and retest reliability, in order to provide informative parameters for the description of the evoked digit activations. Accordingly, the present high-field 3T fMRI study compares the performance of two established digit mapping designs - travelling wave (TW) and blocked design (BD) - for passive tactile stimulation of the five digits, adapted to reduce the time requirements to just below 15 minutes. To be able to assess the retest reliability unaffected by any clinical conditions, the study was performed on neurotypical participants. The results show that both stimulation designs evoke significant and distinct activation clusters in the primary somatosensory cortex of all participants for all five digits. The average spatial locations of the center of gravities across participants show the common succession of distinct digit representation along the central sulcus. The cortical extent elicited activation, which is generally larger for the thumb and the index finger, also shows comparable average values across the two approaches. Less overlap of activation between neighboring digits was obtained in BD, consistent with the distinct single digit neuronal representations. A high retest reliability was obtained for the location of the digit activation, displaying stable center of gravity locations across sessions for both stimulation designs. This is contrasted by only medium to low retest reliability for the extent and overlap of the digit activations, indicating discrepancies across sessions. These results demonstrate the capacity of shortened fMRI digit mapping approaches (both TW and BD) to obtain the full layout of single digit cortical activations on the level of the individual, which together with the high reliability of the location of the digit representation over time indicates both approaches are clinically applicable.