Department of Exercise Sciences

Kinesthetic, but not visual, motor imagery modulates corticomotor excitability

Cathy M. Stinear, Winston D. Byblow, Maarten Steyvers, Oron Levin & Stephan P. Swinnen.
Published in Experimental Brain Research (2006) 168, 157-164


The areas of the brain that are active when we imagine moving are similar to those that are active when we actually perform the movement. However, there has been some debate regarding the extent of primary motor cortex (M1) activity during motor imagery. This is partly due to previous research instructing participants to use a variety of motor imagery strategies. There are two main modes of motor imagery: Kinesthetic motor imagery (KMI) and Visual motor imagery (VMI). KMI involves imagining the feeling of performing a certain movement. VMI involves imagining seeing yourself performing a certain movement. Little is known about how KMI and VMI may differentially affect the extent of primary motor cortex activation.


To examine the modulation of corticomotor excitability during KMI and VMI of a phasic thumb movement task.


  • Participants: 20 healthy right-handed adults.
  • Single-pulse Transcranial Magnetic Stimulation (TMS) was used to assess corticomotor excitability during imagined tapping of the right thumb, in time with a 1 Hz auditory metronome. At the beginning of each block of trials, they were instructed to imagine this movement using either a kinesthetic or visual strategy. They also completed blocks of trials under two control conditions. In one condition, they imagined a visual scene, in the other they thought of nothing in particular. The auditory metronome was present throughout both control conditions.
  • Surface EMG was recorded from the abductor pollicis brevis (APB) and abductor digiti minimi (ADM) muscles. Motor evoked potentials (MEPs) were recorded in both muscles. TMS was delivered with the metronome beat, when APB would be activated during actual performance of the task. This is the ‘on’ phase. TMS was also delivered between metronome beats, when APB would be relaxed during actual task performance. This is the ‘off’ phase. Ten participants completed this experiment.
  • F-waves were also recorded in APB under imagery and control conditions, to check for changes in excitability at the spinal level. Peripheral electrical stimulation of the right median nerve was delivered during the ‘on’ and ‘off’ phase of imagined movement. F-wave amplitude and persistence were determined, for the right APB. A separate group of ten participants completed this experiment.
  • The mean MEP amplitudes in each APB and ADM, during the ‘on’ and ‘off’ phases of each imagery and control condition were calculated. The mean pre-trigger EMG activity was also calculated for each muscle and condition, to ensure that all muscles stayed at rest during imagery.


Fig. 1 Example traces from APB muscle of a typical subject (2 traces overlaid per condition), representing the central tendency of the data for this subject. Calibration bar: 1 mV. 20 ms


  • Visual static imagery (VSI) and visual motor imagery (VMI) did not affect APB MEP amplitude (Fig 1).
  • KMI facilitated APB MEP amplitude during the ‘on’ phase of imagined movement.
  • The facilitation of MEP amplitude was specific to the APB, which was engaged by actual task performance. ADM MEP amplitude was not facilitated by motor imagery, as it is not engaged in actual task performance (Fig 2).
  • F-wave amplitude and persistence were not affected by motor imagery.
Fig.2 Mean MEP amplitude a and mean pre-trigger EMG b as a function of Muscle, Task and Phase. Error bars - 1 s.c., ***p<0.01


  • The present study indicates that kinesthetic motor imagery facilitates the excitability of M1, while visual motor imagery strategies do not. The facilitation of excitability by KMI was both muscle-specific and temporally modulated, in the absence of changes in excitability at the spinal level.
  • These findings have implications for the use of motor imagery as a rehabilitation strategy to enhance motor cortex excitability following stroke. It may be that KMI strategies will be more effective than VMI strategies.


We would like to thank Craig Hall for helpful advice regarding the MIQ-R, and Cheryl Murphy and Melanie Fleming for their assistance in data collection and analysis. Support for the present study was provided through grants from the Research Council at K. U. Leuven, Belgium (contract no. OT/03/61), the Research Programme of the Fund for Scientific Research, Flanders (FWO-Vlaanderen # G.0460.04 and G.0245.05), and the Auckland Medical Research Foundation (81475).