Real-time Human Interface Driven By Eye Movement Event-Related Potential Pattern Recognition

In recent, brain function field analysis attracts concentrated attention, especially on the significant study about BMI (Brain-machine Interface) using fMRi, NIR[1-5], EEG. However, it is known that there exists a problem for the use of an input support device with this serious problem on considerable time for extracting characteristic event related pattern from brain wave and for the large-scale and large-amount device itself such as the MRI equipment. This study aims at rapid BMI pattern recognition for the eye-ball movement, which is considered to be removed a factor from EEG as an artifact [6]. We investigated the repeatability of eye-ball movement Event Related scalp electroencephalogram Potential (ERP) and the characteristics, which possess steady, high voltage and 50ms prompt reaction. As the ERP pattern discriminator, this paper proposes 2 methods, ISE based Euclid Norm and linear mapped Euclid Norm methods.