Neural activation differences in amputees during imitation of intact versus amputee movements

Abstract

The mirror neuron system (MNS) has been attributed with increased activation in motor-related cortical areas upon viewing of another's actions. Recent work suggests that limb movements that are similar and dissimilar in appearance to that of the viewer equivalently activate the MNS. It is unclear if this result can be observed in the action encoding areas in amputees who use prosthetic devices. Intact subjects and upper extremity amputee prosthesis users were recruited to view video demonstrations of tools being used by an intact actor and a prosthetic device user. All subjects pantomimed the movements seen in the video while recording electroencephalography (EEG). Intact subjects showed equivalent left parietofrontal activity during imitation planning after watching the intact or prosthetic arm. Likewise, when prosthesis users imitated prosthesis demonstrations, typical left parietofrontal activation was observed. When prosthesis users imitated intact actors, an additional pattern was revealed which showed greater activity in right parietal and occipital regions that are associated with the mentalizing system. This change may be required for prosthesis users to plan imitation movements in which the limb states between the observed and the observer do not match. The finding that prosthesis users imitating other prosthesis users showed typical left parietofrontal activation suggests that these subjects engage normal planning related activity when they are able to imitate a limb matching their own. This result has significant implications on rehabilitation, as standard therapy involves training with an intact occupational therapist, which could necessitate atypical planning mechanisms in amputees when learning to use their prosthesis.

Keywords: EEG; amputee; mentalizing; mirror neuron; motor control; prosthesis; tool use; upper extremity.

Figure 1

Prosthetic limb terminal devices used in tool use video demonstrations: Hosmer Model 5X Hook (A) and simple articulated anthropomorphic hand (B).

Figure 2

Grand-averaged region-level voltage plots for the Int–Int (red) and Int–Pro (blue) groups. The planning cue is marked with a pink dotted line at –1.0 s and the execution cue is marked with a green dotted line at 0.0 s. Time bin voltage values that are statistically different between the two groups are marked with asterisks.

Figure 3

Grand-averaged electrode headplots for all experimental conditions. For each condition, four representative 100 ms timebins are shown. The planning and execution phases are marked above.

Figure 4

Grand-averaged region-level voltage plots for the Pro–Pro (red) and Pro–Int (blue) groups. The planning cue is marked with a pink dotted line at –1.0 s and the execution cue is marked with a green dotted line at 0.0 s. Time bin voltage values that are statistically different between the two groups are marked with asterisks.

Figure 5

Grand-averaged region-level voltage plots for the Int–Int (red) and Pro–Int (blue) groups. The planning cue is marked with a pink dotted line at –1.0 s and the execution cue is marked with a green dotted line at 0.0 s. Time bin voltage values that are statistically different between the two groups are marked with asterisks.

Figure 6

Grand-averaged region-level voltage plots for the Int–Pro (red) and Pro–Pro (blue) groups. The planning cue is marked with a pink dotted line at –1.0 s and the execution cue is marked with a green dotted line at 0.0 s. Time bin voltage values that are statistically different between the two groups are marked with asterisks.