Search

Simple imaginary movements activate similar cortical and subcortical areas to actual movements, chiefly in the sensory-motor network. However, only a few studies also examined the imagery of more skilful movements such as reaching. Ten volunteers performed reaching movements or imagined the same movements. EEG was simultaneously recorded and analysed with sLORETA, which compared the preparation for actual and imagined reaching with respect to their baseline and between tasks. Major differences between them were found at three time intervals after target presentation, always in favour of the actual reaching condition. The first one was from 160 to 220 msec in the frontal and parietal regions. The second difference was evident from 220 to 320 msec in the premotor cortex. The third difference was evident from 320 msec, mainly in the perirolandic region. Also, the anterior and posterior cingulate cortices were widely involved, in both tasks. We suggest the existence of two separate systems which may work together during actual reaching programming. The first one involves structures such as the premotor cortex, supplementary motor area and primary motor cortex, together with the parietal and occipital cortex. This system may integrate extrinsic target coordinates with proprioceptive information from the reaching arm and pre-stored programs in the associative motor cortex. It is activated strongly and involves more cortical areas in actual than imagined reaching. The second system, common to both tasks, involves anterior and posterior cingulate cortices, with the possible role of contributing awareness and focusing the various components of the process.

Current experimental methods for the study of reaching in the MRI environment do not exactly mimic actual reaching, due to constrains in movement which are imposed by the MRI machine itself. We tested a haptic robot (HR) as such a tool. Positive results would also be promising for combined use of fMRI and EEG to study reaching. Twenty right-handed subjects performed reaching tasks with their right hand with and without the HR. Reaction time, movement time (MT), accuracy, event-related potentials (ERPs) and event-related desynchronisation/synchronisation (ERD/ERS) were studied. Reaction times and accuracies did not differ significantly between the two tasks, while the MT was significantly longer in HR reaching (959 vs. 447 ms). We identified two positive and two negative ERP peaks across all leads in both tasks. The latencies of the P1 and N2 peaks were significantly longer in HR reaching, while there were no significant differences in the P3 and N4 latencies. ERD/ERS topographies were similar between tasks and similar to other reaching studies. Main difference was in ERS rebound which was observed only in actual reaching. Probable reason was significantly larger MT. We found that reaching with the HR engages similar neural structures as in actual reaching. Although there are some constrains, its use may be superior to other techniques used for reaching studies in the MRI environment, where freedom of movement is limited.