Imagery of movements immediately following performance allows learning of motor skills that interfere

www.nature.com/scientificreports OPEN Received: 3 May 2018 Accepted: 5 September 2018 Published: xx xx xxxx Imagery of movements immediately following performance allows learning of motor skills that interfere Hannah R. Sheahan1, James N. Ingram1,2, Goda M. Žalalytė1 & Daniel M. Wolpert1,2 Motor imagery, that is the mental rehearsal of a motor skill, can lead to improvements when performing the same skill. Here we show a powerful and complementary role, in which motor imagery of different movements after actually performing a skill allows learning that is not possible without imagery. We leverage a well-studied motor learning task in which subjects reach in the presence of a dynamic (force-field) perturbation. When two opposing perturbations are presented alternately for the same physical movement, there is substantial interference, preventing any learning. However, when the same physical movement is associated with follow-through movements that differ for each perturbation, both skills can be learned. Here we show that when subjects perform the skill and only imagine the follow-through, substantial learning occurs. In contrast, without such motor imagery there was no learning. Therefore, motor imagery can have a profound effect on skill acquisition even when the imagery is not of the skill itself. Our results suggest that motor imagery may evoke different neural states for the same physical state, thereby enhancing learning. The ability to acquire new motor skills without disrupting existing ones is critical to the development of a broad motor repertoire. We have previously suggested that the key to representing multiple motor memories is to have each associated with different neural states, rather than physical states of the body1. Specifically, we proposed that when reaching in two opposing force-field environments which alternate randomly from trial to trial, the inability of subjects to learn2–7 is due to the fact that each movement is associated with the same neural states. However, contexts which separate neural states for the same physical states should allow learning by enabling the same physical movement to be associated with different motor commands. For example, if each movement through the force-field is part of a larger motor sequence comprised of a different follow-through movement, two opposing perturbations can be learned1,6. As motor preparation is thought to involve setting the initial neural state8, just planning different follow-through movements, without execution, results in learning of distinct representations1. From this perspective, other behaviours that create different neural states for the same physical states may also enable the learning of distinct motor memories. Many studies have suggested that imagining a movement and physically executing it may engage similar neural substrates. For example, human neuroimaging studies have shown similar motor-related activity when imagining and executing movements9–12. Moreover, simply imagining moving a body part increases the EMG response of the associated muscles to TMS over primary motor cortex, suggesting that the circuits involved in action are at least partially active during imagery13,14. Similarly, direct recording of neural populations have recently revealed that when monkeys covertly control a BMI-cursor, the evolution of neural states associated with the preparation and execution of the BMI movements are similar and specific to those observed during the corresponding physical reaches15. Given that similar motor cortical dynamics are seen in human and non-human primates16, we hypothesized that the same overlap of dynamical neural states may also exist when humans execute or imagine movements. That is, if the neural states of a motor area involved in generating a physical movement can be made 1 Computational and Biological Learning Laboratory, Department of Engineering, University of Cambridge, Cambridge, UK. 2Department of Neuroscience, Zuckerman Mind Brain Behavior Institute, Columbia University, New York, NY, USA. Correspondence and requests for materials should be addressed to H.R.S. (email: sheahan.hannah@ gmail.com) SCIENtIfIC REPOrtS | (2018) 8:14330 | DOI:10.1038/s41598-018-32606-9 1 www.nature.com/scientificreports/ Exposure trials A Planning only Motor imagery Secondary target disappears mid-movement Movement to secondary target imagined Follow through CW Field No motor imagery CCW Field B Channel trials Follow through channels Secondary target Motor imagery channels All groups Motor imagery group only No motor imagery channels No motor imagery group only Executed movement Imagined movement Central target Starting location Figure 1. Experimental paradigm. Subjects performed reaching movements that were either (A) exposure trials or (B) channel trials. On all trials, a starting location, central target and one secondary target (at either −45° or +45° relative to the initial movement direction) were displayed from the start of the trial. (A) On exposure trials, a velocity-dependent curl force field (blue arrows) was applied on the initial movement. The field direction, clockwise (CW) or counter-clockwise (CCW) was determined by the secondary target location. The exposure trials varied across the groups. The Follow through group continued the initial movement to the secondary target (null field as in channel trials). For the Planning only group, the secondary target disappeared late in the initial movement and they were required to stop at the central target. Both the Motor imagery and No-motor imagery groups were cued by a blue central target, displayed from the start of the trial, indicating that they should stop the movement at the central target. In addition, the motor imagery groups were asked to imagine making a movement to the secondary target and press a button when the imagined movement was complete. (B) On follow through channel trials (left), subjects made a movement to the central target followed immediately by a movement to the secondary target. A channel was applied on the initial movement, allowing an assessment of adaptation measured as the forces applied into the channel wall. A null field was applied on the secondary movement. For half of participants in the motor imagery group, we also included channels for imagined follow though trials (middle) at the end of the exposure phase. Likewise, for half of participants in the no-motor imagery group we included channels for movements just to the central target (right). Note that for clarity in all panels the trials for the two different secondary targets are shown separated, but in the experiment the starting and central targets were in identical locations so that the initial movements were the same. In the experiment there were 4 possible starting locations but for clarity we display only one. different (even partially) by motor imagery, then each of these different neural states could be assoc (...truncated)