Differential experiences of embodiment between body-powered and myoelectric prosthesis users

Abstract

Prosthesis embodiment, the perception of a prosthesis as part of one's body, may be an important component of functional recovery for individuals with upper limb absence. This work determined whether embodiment differs between body-powered and myoelectric prosthesis users. In a sample of nine individuals with transradial limb absence, embodiment was quantified using a survey regarding prosthesis ownership and agency. The extent to which the prosthesis affected the body schema, the representation of the body's dimensions, was assessed using limb length estimation. Because body-powered prostheses offer proprioceptive feedback that myoelectric prostheses do not, it was hypothesized that both measures would reveal stronger embodiment of body-powered prostheses. However, our results did not show differences across the two prosthesis designs. Instead, body schema was influenced by several patient-specific characteristics, including the cause of limb absence (acquired or congenital) and hours of daily prosthesis wear. These results indicate that regular prosthesis wear and embodiment are connected, regardless of the actual prosthesis design. Identifying whether embodiment is a direct consequence of regular prosthesis use would offer insight on how individuals with limb absence could modify their behavior to more fully embody their prosthesis.

Figure 1

Conditions for the limb length estimation task. Participants performed the limb length estimation task by placing their arm inside an opaque tube with a sliding indicator affixed to the exterior. They performed this task using their dominant or intact limbs (a), their prosthetic limb (b), and their residual limb (c). For each condition, they estimated (red arrows) where they perceived the end of their hand, prosthesis (P-PT), residual limb (P-RL, NP-RL), or where the prosthesis would be if they were wearing it (NP-PT).

Figure 2

Self-reported embodiment. Mean ownership and agency scores for BP and MYO users. Error bars represent standard deviation across subjects and individual points represent individual participant scores.

Figure 3

Limb length estimation error. (a) Average limb length estimation for the dominant (‘o’) limbs of controls, as well as the intact limbs for all prosthesis users (‘x’). (b) Average limb length estimation error for BP (solid squares) and MYO (open squares) prosthesis users. (c) Average limb length estimation error for participants with acquired limb loss (solid triangles) and congenital (open triangles) limb absence. Error bars represent standard deviation across subjects.

Figure 4

Correlations between limb length error and participant characteristics. Correlations between (a) NP-PT error and hours of daily prosthesis wear, (b) NP-RL error and hours of daily prosthesis wear, and (c) NP-RL and agency scores.