Motor imagery enhances performance beyond the imagined action

Abstract

Motor imagery is frequently utilized to improve the performance of specific target movements in sports and rehabilitation. In this study, we show that motor imagery can facilitate learning of not only the imagined target movements but also sequentially linked overt movements. Hybrid sequences comprising imagined and physically executed segments allowed participants to learn specific movement characteristics of the executed segments when they were consistently associated with specific imagined segments. Electrophysiological recordings revealed that the degree of event-related synchronization in the alpha and beta bands during a basic motor imagery task was correlated with imagery-evoked motor learning. Thus, both behavioral and neural evidence indicate that motor imagery's benefits extend beyond the imagined movements, improving performance in linked overt movements. This provides decisive evidence for the functional equivalence of imagined and overt movements and suggests applications for imagery in sports and rehabilitation.

Keywords: adaptation; force field; motor imagery; motor learning; motor sequence.

Fig. 1.

Experimental design and protocol. (A) Kinarm Exoskeleton Robot Lab. The screen/mirror is displayed transparent here for illustration; however, in the experiment participants were not able to see their arms. (B) Exemplary trial of the reaching task separately for each experimental group. White dot—cue, gray dot—middle target, yellow dot—final target, red dot—hand position at the beginning of a trial, black arrow—instructed reaching path during the trial, red arrow—imaginary reaching path, big teal arrow—exemplary force field direction. (C) Reaching task trial sequence. The prior movement, that is, the reach from cue to middle target was only overtly performed by the active group. The imagery group only imagined the prior movement while the hand was already positioned at the middle target. The control group did not perform or imagine a prior movement at all. Feedback about the movement time between middle and final target was given after every trial to encourage a similar reaching speed across participants. (D) Experimental phases. The timing check task consisted of active reaches regardless of group membership. The imagined fist clenching task was only performed by participants in the MI group. (E) Exemplary cue/final target combinations with big arrows representing force field direction depending on the cue’s location. White dot with teal/gray border—cue, gray dot—middle target, yellow dot—final target, black arrow—desired reaching path, and big teal/gray arrow—force field direction. (F) Imagined fist clenching task sequence. Participants were instructed to relax or to perform kinesthetic motor imagery of clenching their fist around the Kinarm’s handle when the fixation cross turned white or red, respectively. (A–E) were adapted from .

Fig. 2.

Behavioral results. Data of the Control and Active groups were also reported in ref. . (A) Reaching trajectories. Single trial trajectories from middle to final targets of all participants (N = 20 in each group) in specific blocks of the experiment. For each final target/cue position combination, one reaching trajectory per participant per block is shown. Force fields were only present in adaptation blocks. (B) Averaged maximal perpendicular error (MPE) over trials and participants for each block. Force fields were present from block 7 to block 56 (adaptation blocks = white background). Error bands indicate 95% CIs. (C) Change of MPE from beginning to end of adaptation. Each colored dot depicts the difference between average performance in the first and last two adaptation blocks of one participant. Black dots mark the respective group averages. Lines denote significant differences between groups P < 0.05. Stars mark significant within-group effects P < 0.05. (D) Change of MPE from baseline to washout. The differences between the first washout and the last baseline block are displayed per participant and as a group average. (E) Averaged FFC over trials and participants for each block. (F) Average FFC in the last 4 adaptation blocks. (G) Comparison of median dwell time in the middle target in the timing check task. Each colored dot depicts the median dwell time of one participant.

Fig. 3.

EEG data imagined fist clenching task. (A) Time-frequency representation of grand averaged (N = 16) data at electrode standard positions C3 and C4. Color represents power change in percent relative to the baseline window from −0.75 to −0.25 s. Red cross—start of motor imagery, white cross—start of relaxation. (B) Power change in percent over time relative to baseline activity in 3 frequency ranges averaged across participants and respective frequencies. Orange line—C3 electrode; black line—C4 electrode. (C) Distribution of averaged alpha activity across the scalp in specified time windows (Left—ERD; Right—ERS), averaged across participants. (D) Source reconstruction of averaged alpha activity in chosen time windows, averaged across participants. The strongest 50% of values are shown in color.

Fig. 4.

Correlation analysis results. (A) Pearson’s correlation of change of error in the reaching task with power change in the imagined fist clenching task across participants in C3 and F3. The pixel color indicates the respective correlation value at that location. A more negative change of error denotes better adaptation performance. Red cross—start of motor imagery, white cross—start of relaxation. The significant cluster is outlined in black. (B) Topoplots of different time windows and frequency ranges displaying the distribution of thresholded correlation values. Only the significant cluster is shown. Correlation values are averaged in the specified time and frequency ranges. (C) Relationship between change of error in the reaching task and the maximal positive power change value (ERS) in the imagined fist clenching task across participants. Each point depicts a participant. The maximal positive power change value was taken for each participant in the alpha frequency range between 2 and 4 s. (D and E) Pearson’s correlation of change of error in the reaching task with power change in the imagined fist clenching task—averaged over specified time and frequency range—across participants in source space. The strongest 50% of values are shown in color in (D) and the color bar is kept identical for (E).