Motor imagery based EEG features visualization for BCI applications

Abstract

Over recent years, electroencephalography’s (EEG) use in the state-of-the-art brain-computer interface (BCI) technology has broadened to augment the quality of life, both with medical and non-medical applications. For medical applications, the availability of real-time data for processing, which could be used as command signals to control robotic devices, is limited to specific platforms. This paper focuses on the possibility to analyse and visualize EEG signal features using OpenViBE acquisition platform in offline mode apart from its default real-time processing capability, and the options available for processing of data in offline mode. We employed OpenViBE platform to acquire EEG signals, pre-process it and extract features for a BCI system. For testing purposes, we analysed and tried to visualize EEG data offline, by developing scenarios, using method for quantification of event-related (de)synchronization ERD/ERS patterns, as well as, built in signal processing algorithms available in OpenViBE-designer toolbox. Acquired data was based on deployment of standard Graz BCI experimental protocol, used for foot kinaesthetic motor imagery (KMI). Results clearly reflect that the platform OpenViBE is a streaming tool that encourages processing and analysis of EEG data online, contrary to analysis, or visualization of data in offline, or global mode. For offline analysis and visualization of data, other relevant platforms are discussed. In online execution of BCI, OpenViBE is a potential tool for the control of wearable lower-limb devices, robotic vehicles and rehabilitation equipment. Other applications include remote control of mechatronic devices, or driving of passenger cars by human thoughts.