IKBFU Scientists Proposed a New Method for Creating a Brain-Computer Interface Using EEG Readings

The new method will allow scientists to detect event-related potentials in signals of a brain’s electric activity. The event-related potentials are an electro-physiological response that a brain gives as a reaction to changes around the organism, e.g. to light flashes, loud sounds, etc. The potentials are among the most promising fields for “brain-computer” research.

According to Elena Pitsik, a researcher at the Baltic Center for Neurotechnology and Artificial Intelligence at IKBFU, recurrence analysis will be used as a method to create such an interface.

The idea is quite simple. Many systems of the real world have a natural property of repeating their states over time. By measuring the number of such repetitions, as well as the time of their occurrence, we can make conclusions about how "complex" processes occur in the system: whether it is regular, i.e. frequently repeating in time, or whether it has more chaotic and unpredictable character. At the same time, with the help of recurrence analysis we are able to detect and describe individual processes in a system. This is why we study the event-related potentials on the EEG signal in our research.

The scientists note that the proposed method requires low computational costs and is easy to interpret. They plan to continue the research in order to configure the system to work in real time.

The results of the research by the Baltic Center for Neurotechnology and Artificial Intelligence at Immanuel Kant Baltic Federal University have been published in the Springer Nature academic journal. The work was carried out within the framework of the strategic project of IKBFU development.