Forward models in motor imagery

Motor imagery designates actions which are not actually executed but are imagined as if they are. Imagined and executed actions exhibit similar properties: they take approximately the same amount of time, follow the same motor principles and biomechanical constraints, and involve similar neuronal activity. The two currently most prominent theoretical frameworks for motor imagery are simulation theory and emulation theory. In simulation theory it is assumed that motor imagery involves motor planning processes. From a computational viewpoint this involves an inverse model, which specifies motor commands according to intended actions goals. The simulation theory of motor imagery does not fully account for the dynamic aspects, i.e. the ongoing sequence of events during motor imagery. Therefore, the emulation theory emphasizes that motor imagery also requires forward models, which predict the consequences of motor commands on the own body and on the environment. In the proposed project, three series of experiments are aimed at shedding some light on the operation and limits of forward models during motor imagery. In particular, the major aim is to investigate to what extent accurate predictions about the ongoing state of the motor system and effects in the environment are made during motor imagery. In series 1 the occurrence of action errors during motor imagery will be investigated. Using a copy-typing a task it will be analyzed whether and to what extent different types of errors are imagined, what roles different forms of action feedback play for the awareness of errors, and whether typing-style modulates the relevant representations. Using a dart throwing task it will be investigated whether imagined accuracy follows the same principles as actual accuracy, whether imagined and actual performance are more similar in dart-experts than in novices, and which role action feedback during executed actions plays for predicted accuracy. In series 2 action inhibition in motor imagery will be investigated, using the stop signal task. The questions whether it is possible to imagine inhibiting imagined actions, whether forward models predict the ongoing state of the motor system during MI in real time, whether inhibitory after-effects occur after imagined inhibition, and whether inhibitory after-effects after imagined and actual inhibition follow the same principles will be addressed. In series 3 the questions whether biomechanical constraints of bimanual coordination (i.e. better performance with symmetric than with parallel movements) are reflected in motor imagery, whether this is the case for planning and execution related constraints, and whether constraints persist when bimanual actions are imagined with one hand and executed with the other will be addressed. Results from this project will significantly enhance the understanding of the mechanisms of motor imagery and contribute to a solid theoretical and empirical basis for the application of motor imagery to mental training.

Motor imagery designates actions which are not actually executed but are imagined as if they are. Imagined and executed actions exhibit similar properties: they take approximately the same amount of time, follow the same movement principles, and involve similar neuronal activity. In the project, it was investigated whether movement effects on the body and the environment are represented in motor imagery. For this aim three series of experiments were conducted. In Series 1, the occurrence of action errors during motor imagery was investigated using a copy-typing and a dart-throwing task. It was shown that action errors are represented during motor imagery. However, they are represented to a lesser degree during motor imagery than they occur in motor execution. This shows that action errors are internally predicted in motor imagery, but that this prediction is not precise. This is partly due to the lack of feedback from the environment. In Series 2, it was investigated whether it is possible to imagine inhibiting imagined actions. The experiments show that this is possible and that this occurs approximately in real time. The imagination of action inhibition includes similar mechanisms as the inhibition of actual execution. In Series 3, it was investigated whether bimanual coordination constraints (i.e. better performance with symmetric than with parallel movements) are reflected in motor imagery. The data show that bimanual coordination constraints are reflected in motor imagery, both during the planning and the execution phase, and they show that cognitive and motor constraints are represented. Together, the results enhance the understanding of similarities and differences of motor imagery and motor execution, and they enhance the understanding of the mechanisms in motor imagery. enhance the understanding of the mechanisms in motor imagery. In particular, they show that predictions of movement effects on the body and the environment occur in motor imagery, even though those predictions are not always precise. The results a relevant for the application of motor imagery in mental training, which is used in sports and rehabilitation. Further research question arose from the project, e.g. how imagined actions are prevented (i.e. inhibited) from execution and which type of representations are acquired in mental training.