Figures and data
Canonical hand somatosensory pathway and top-down projections from the sensorimotor cortex
Schematic illustration of the somatosensory processing stream. The bottom-up somatosensory pathway (blue) conveys tactile and proprioceptive input from the hand to the ipsilateral cuneate nucleus (Cu). From there, signals transfer to the contralateral ventroposterior lateral (VPL) thalamic nucleus, and finally the hand area of the contralateral primary somatosensory cortex (S1). Top-down projections from the sensorimotor cortex (red), as described in animal studies, are shown projecting to each level of the somatosensory pathway. Following cervical spinal cord injury, the transmission of bottom-up input from below the lesion level (orange arrow) is partially or completely interrupted, providing a model to explore top-down processing in vivo. Created with BioRender.com.
Group-level activation during hand movement follows the canonical somatosensory pathway in controls and cervical SCI
Group-level activity maps for control participants (A) and individuals with cervical spinal cord injury (B) during (attempted) hand movements. Both groups showed activation clusters overlapping with the key relay stations along the somatosensory processing stream (see Figure 1 and Figure 3A), including the brainstem cuneate nuclei, the ventroposterior lateral (VPL) thalamus, and the S1 hand area. Suprathreshold clusters are outlined in black, with subthreshold activity displayed at a reduced opacity. Colour bars represent 1 − pFWE (family-wise error- corrected p-values). Warm colours (red–yellow) represent the contrast non-dominant > dominant hand (left panel) and more impaired > less impaired hand (right panel), whereas cool colours (light blue–blue) represent the inverse contrasts (dominant > non-dominant hand; less impaired > more impaired hand).
Overt and attempted hand movements elicit lateralised activity across the entire somatosensory processing stream
(A) Anatomical regions of interest (ROIs) along the somatosensory processing stream are outlined in white, showcasing the main relay stations of somatosensory afferent inputs, bilaterally. From top to bottom: the primary somatosensory cortex (S1) hand area, the ventroposterior lateral (VPL) thalamic nuclei, and the cuneate nuclei in the brainstem. (B) ROI analyses show mean activation (z-statistic) during hand movement, for the ROI on the same side as the moving hand (Ipsi) and for the opposite side (Contra). Each plot corresponds to the ROIs shown in the same row of panel. Individual control participants are plotted as blue dots, and individual SCI participants as red dots. The most impaired SCI participant (PT01), who had an anatomical and clinically complete cervical injury, i.e. no spared spinal tissue bridges, and a complete hand paralysis with no detectable EMG during voluntary hand movement (see Supplementary materials), is shown in orange. Across all three ROIs, activity was consistently stronger in the hemisphere corresponding to the moving hand, revealing robust lateralisation of the somatosensory pathway. This lateralised pattern was present not only in controls performing overt movements but also in SCI participants performing purely attempted hand movements (i.e., PT01). p-values are false discovery rate corrected for multiple comparisons within each ROI. *** = p < 0.001, ** = p < 0.01, * = < 0.05.
Structural degeneration within the cuneate nuclei of the brainstem after cervical spinal cord injury
(A) We found macrostructural volume loss in SCI participants compared with healthy controls (statistical parametric map thresholded at p < 0.001 voxelwise, uncorrected with cluster-level pFWE < 0.05). Volume loss was localised to the dorsal medulla and overlapped with the cuneate nuclei ROI (outlined in white for visualisation only). (B) Region-of-interest (ROI) analyses showing average quantitative MRI (qMRI) values sensitive to myelin (Magnetisation Transfer Saturation; MTsat) and broader tissue microstructure (R2*) within the S1 hand area, VPL, and cuneate nuclei, bilaterally. SCI participants showed significantly reduced MTsat and R2* values in the cuneate nucleus compared with controls, whereas no significant group differences were observed in the VPL or S1. Individual participants are plotted as blue (controls) or red (SCI) dots. As in Figure 2, a complete SCI participant (PT01) is shown in orange for visualisation (ASIA A, injury level C4, GRASSP score = 0; tissue bridges = 0). Error bars denote the standard error of the mean. p.u. = percentage units; 1/s = per second. *** = p < 0.001, ** = p < 0.01, * = < 0.05.
