Lee, P.-L., Chang, H.-C., Hsieh, T.-Y., et al (2012). A brain-wave-actuated small robot car decomposition-based approach. IEEE Transactions on Systems, Man, and Cybernetics - Part A: Systems and Humans, 42(5):1053-1064 Details
Customer study highlights
Smart devices that interact directly with the brain were first researched in the 1970s under a US government funded research program. Today, the ‘brain computer interface’ (BCI) exists in many forms, ranging from consumer-level gaming equipment to prosthetics to student learning systems.
Being non-invasive, the electroencephalogram (EEG) remains an attractive choice for BCI devices. However, the challenge is in finding ways to maximise the information transfer rate of the signal for useful control over an external device.
Lee et al (2012) chose to use visual-evoked potentials (VEPs), signals in the EEG that are induced by visual stimuli, to control a small wireless robot car. Highly reproducible, VEPs have fast response times and are time- and phase-locked to onset of visual stimuli, making them potentially very useful for a BCI device.
EEG signals from 11 male participants (22-27 yrs), recorded using a BioAmp, were used to code left, forward, and right movements of the robot car. To maximise the information transfer rate, VEP signals were extracted using ensemble empirical mode decomposition (EEMD) and a matched filter detector.
All 11 subjects who attempted to navigate the robot car through its 1.5m 'S-shaped' course were able to do so in an average time of 84.5 seconds. This method provided improved accuracy and an average information transfer rate of 42.1 bits/min, compared to <33 bits/min from similar studies using FFT-based signal decomposition.
Despite being computationally more intensive, EEMD-based methods may represent a favourable compromise between accuracy and processing delay compared to other methods.