DescriptionElectroencephalographic (EEG) and magnetoencephalographic (MEG) signals carry information about upper-limb kinematics. Their expression in M/EEG signals is potentially influenced by many factors. We believe that vision and movement type (i.e., discrete vs continuous) are key factors. To identify their effects, we investigated visuomotor tasks in several studies.
In a first EEG study (15 healthy subjects), we investigated the influence of vision during goal-directed upper-limb movements . Our paradigm implemented a center-out task (COT) followed by a pursuit tracking task (PTT). It involved two visual stimuli (target and cursor) in two conditions. In one condition (exe), subjects controlled the cursor with their right hand. In the second condition (obs), subjects observed a computer controlled cursor. Regarding the PTT, our findings indicate that premotor and primary sensorimotor areas carried significant information about cursor velocity in the exe condition, whereas parieto-occipital areas carried significant directional information in either condition. Regarding the COT, we could classify condition (exe vs obs) and direction (left, up, right and down) above significance level. The two classifiers relied on different cortical areas. Contra-lateral pre- and primary motor areas predicted condition, while parieto-occipital areas predicted direction in both conditions.
In a second EEG study (10 healthy subjects), we investigated how the results of the PTT translate to an online control scenario. In both conditions (hand movement based control, EEG based control) the kinematics could be decoded above the significance level, with parieto-occipital areas carrying significant information about end-effector velocity. We observed an amplitude-scaling mismatch between the decoded and actual kinematics, which resulted in declined tracking quality in the EEG based control condition.
To address the amplitude-scaling mismatch, we conducted an MEG study (20 healthy subjects), where we investigated a similar PTT as in the first EEG study. In this study, the subjects controlled the cursor with their right index finger . We found that the MEG signals carried not only significant information about velocity, but also about speed in primary-sensorimotor areas.
We believe that our findings can aid to a new generation of non-invasive brain-computer interfaces.
 Kobler RJ et al. Sci.Rep. 2018  Kobler RJ et al. Proc. 8thGBCIC. 2019
|Period||23 Oct 2019|
|Event title||Neuroscience 2019|
|Location||Chicago, United States, Illinois|
- Brain-Computer Interfaces
ASJC Scopus subject areas
- Neuroscience (miscellaneous)
- Signal Processing
- Biomedical Engineering
Fields of Expertise
- Human- & Biotechnology
Research output: Contribution to conference › Abstract › peer-review