Identification of Wrist Movements from EEG Signals using Methods of Artificial Intelligence

V. Logar, A. Belič, and I. Škrjanc (Slovenia)


Dynamic visuo-motor task, EEG, Information transfer, Phase demodulation, Wrist movement estimation


Lately, a new concept of information coding and transfer in the human brain has been proposed. According to some authors the information is transferred between the active re gions of the brain as a phase coded content. In some opin ions, the brain’s oscillatory activity in cooperation with the phase coding represents one of the basic information cod ing mechanisms in the brain. In the past, some research has already been done regarding visuo-motor integration to support this theory. Therefore, this paper tries to in vestigate whether the phase coding theory is also valid for other, more complex visuo-motor tasks. For this reason we measured EEG signals from four subjects performing a dynamic visuo-motor task and try to extract the infor mation about their efficiency. The study revealed that it is possible to decode some information about the subjects’ performance from the EEG signals using signal filtering, a phase demodulation method and a principal component analysis. The information we were trying to find were the wrist movements of the subjects, when they performed tar get signal tracking visuo-motor task. The study revealed that using the above mentioned methods of signal process ing and a fuzzy model, it is possible to successfully predict the wrist movements of all the subjects involved. There fore, we can conclude that phase coding definitely is one of the integral aspects of the brain function, while phase demodulation concept proved to be efficient in decoding information from more complex cognitive tasks.

Important Links:

Go Back