Figure 4. Topography is preserved despite the absence of peripheral input.
To rule out the potential contribution of peripheral inputs from the injured nerve to preservation of the phantom hand map, we tested an upper-limb amputee suffering from brachial plexus avulsion injury with vivid kinaesthetic phantom digit sensations. This injury involves the tearing of the nerve from its attachment at the spinal cord, abolishing afferent inputs and efferent outputs to the residual limb. 3 tesla MRI was used here due to safety restrictions, meaning that standard field digit preference maps were acquired. The bimanual travelling wave task (A, Figure 2) elicited maps in the controls with reversed colour gradients (B - left 2 panels). In comparison to the ultra high-field left hand maps (right panel), the standard field left hand maps were cruder (e.g. less digit specificity, B - middle panel). (C) The brachial plexus amputee showed two maps with reversed colour gradients, comparable to those found in controls. The map in the missing hand hemisphere showed preference for digits in primary somatosensory cortex. (D) Inter-digit representational similarity was assessed for multivoxel patterns underlying the missing hand map. The correlation value of the brachial plexus amputee’s inter-digit overlap pattern with the controls exemplified in Figure 3 fell within the normal controls-to-controls correlation range. Together, these results suggest that preserved representation of the missing hand existed independently of peripheral inputs. Other annotations are as in Figure 1.