Shaping corticospinal pathways in virtual reality: effects of task complexity and sensory feedback during mirror therapy in neurologically intact individuals

Abstract

Background: Restoration of limb function for individuals with unilateral weakness typically requires volitional muscle control, which is often not present for individuals with severe impairment. Mirror therapy-interventions using a mirror box to reflect the less-impaired limb onto the more-impaired limb-can facilitate corticospinal excitability, leading to enhanced recovery in severely impaired clinical populations. However, the mirror box applies limitations on mirror therapy, namely that all movements appear bilateral and are confined to a small area, impeding integration of complex activities and multisensory feedback (e.g., visuo-tactile stimulation). These limitations can be addressed with virtual reality, but the resulting effect on corticospinal excitability is unclear.

Objective: Examine how virtual reality-based unilateral mirroring, complex activities during mirroring, and visuo-tactile stimulation prior to mirroring affect corticospinal excitability.

Materials and methods: Participants with no known neurological conditions (n = 17) donned a virtual reality system (NeuRRoVR) that displayed a first-person perspective of a virtual avatar that matched their motions. Transcranial magnetic stimulation-induced motor evoked potentials in the nondominant hand muscles were used to evaluate corticospinal excitability in four conditions: resting, mirroring, mirroring with prior visuo-tactile stimulation (mirroring + TACT), and control. During mirroring, the movements of each participant's dominant limb were reflected onto the nondominant limb of the virtual avatar, and the avatar's dominant limb was kept immobile (i.e., unilateral mirroring). The mirroring + TACT condition was the same as the mirroring condition, except that mirroring was preceded by visuo-tactile stimulation of the nondominant limb. During the control condition, unilateral mirroring was disabled. During all conditions, participants performed simple (flex/extend fingers) and complex (stack virtual blocks) activities.

Results: We found that unilateral mirroring increased corticospinal excitability compared to no mirroring (p < 0.001), complex activities increased excitability compared to simple activities during mirroring (p < 0.001), and visuo-tactile stimulation prior to mirroring decreased excitability (p = 0.032). We also found that these features did not interact with each other.

Discussions: The findings of this study shed light onto the neurological mechanisms of mirror therapy and demonstrate the unique ways in which virtual reality can augment mirror therapy. The findings have important implications for rehabilitation for design of virtual reality systems for clinical populations.

Fig. 1

(A) Demonstration of NeuRRoVR, a virtual reality platform that can display and mirror the trunk, upper extremity, lower extremity, and finger movements. (B) Experimental protocol. Participants used NeuRRoVR in four conditions: rest, mirroring, mirroring + TACT, and control. During mirroring and mirroring + TACT, movements of the participant’s dominant limb were reflected onto the avatar’s nondominant limb and the avatar’s dominant limb was immobilized (i.e., unilateral mirroring). Prior to the mirroring + TACT condition, participants experienced a two-minute period of visuo-tactile stimulation. In the control condition, unilateral mirroring was disabled. During the mirroring, mirroring + TACT, and control conditions, participants performed a simple activity (finger flexion/extension) and a complex, task-oriented activity (stacking virtual blocks). (C) Transcranial magnetic stimulation experimental set-up. The TMS coil was positioned over the primary motor cortex contralateral to the nondominant (mirroring) limb, and motor evoked potentials were measured with a surface electromyography sensor on the flexor digitorum superficialis. A motion capture system measured the position and orientation of the coil and the participant’s head for consistent coil placement throughout the experiment

Fig. 2

Motor evoked potentials (MEPs) from a representative participant in all conditions and activities

Fig. 3

Raincloud plots [76] comparing corticospinal excitability during (Left) no mirroring during the simple activity to mirroring during the simple activity, (Middle) mirroring during the simple activity to mirroring during the complex activity, and (Right) mirroring during the simple activity to mirroring during the simple activity following visuo-tactile stimulation (TACT). Here, the thick horizontal line contained by the box-and-whisker indicates the median and the thin horizontal line extending beyond the box-and-whisker indicates the mean

Fig. 4

Raincloud plots [76] of our mixed-model analysis comparing the interaction between mirroring and task complexity (Top) and task complexity and visuo-tactile stimulation (Bottom). Here, the thick horizontal line contained by the box-and-whisker indicates the median and the thin horizontal line extending beyond the box-and-whisker indicates the mean. The “*” denotes a significant main effect of task complexity and the “†” denotes a significant main effect of mirroring. The significance level was p < 0.05