Improving bit rate in an auditory BCI: Exploiting error-related potentials
Authors:
- 2 April 2016
- journal article
- research article
- Published by Taylor & Francis in Brain-Computer Interfaces
- Vol. 3 (2) , 75-87
Abstract
The error-related potential (ErrP) can inform the correction of brain-computer interface (BCI) mistakes, but it has thus far been incorporated only into visual BCIs, with mixed success. Given that ErrPs are thought to have higher impact when BCI accuracy is relatively low, we sought to identify the aurally evoked ErrP and investigate its auto-corrective value in auditory BCIs, which typically yield lower accuracies than visual BCIs. We implemented an auditory P300 BCI with four selectable items. Each of nine typically developed participants attempted to spell letter sequences on two separate days. Erroneous feedback was detected by (i) making use of the ErrP, (ii) assessing BCI selection confidence, and (iii) combining these two pieces of information into a hybrid detector. ErrPs were detected with an average cross-validation area under the curve of 0.946. Simulated automatic correction by reverting to the second-ranked letter improved participant-wise information transfer rate by 2.3 bits/minute when errors were detected by the hybrid method. The results suggest ErrP-based error correction can be used to make a substantial improvement in the performance of auditory BCIs.Keywords
Funding Information
- National Natural Science Foundation of China (91320202, 71201148, 61420106001)
- Natural Sciences and Engineering Research Council of Canada (RGPIN-2014-06077)
- Foundation of National Key Laboratory of Human Factors Engineering (HF2011-Z-Z-A-01)
- Holland Bloorview Foundation Graduate Student Award Program (N/A)
This publication has 50 references indexed in Scilit:
- Task-dependent signal variations in EEG error-related potentials for brain–computer interfacesJournal of Neural Engineering, 2013
- Online Adaptation of a c-VEP Brain-Computer Interface(BCI) Based on Error-Related Potentials and Unsupervised LearningPLOS ONE, 2012
- Online detection of error-related potentials boosts the performance of mental typewritersBMC Neuroscience, 2012
- Toward smarter BCIs: extending BCIs through hybridization and intelligent controlJournal of Neural Engineering, 2011
- A novel P300-based brain–computer interface stimulus presentation paradigm: Moving beyond rows and columnsClinical Neurophysiology, 2010
- Online Detection of P300 and Error Potentials in a BCI SpellerComputational Intelligence and Neuroscience, 2010
- A P300-based brain–computer interface: Initial tests by ALS patientsClinical Neurophysiology, 2006
- The Timing of Action-Monitoring Processes in the Anterior Cingulate CortexJournal of Cognitive Neuroscience, 2002
- The mental prosthesis: assessing the speed of a P300-based brain-computer interfaceIEEE Transactions on Rehabilitation Engineering, 2000
- An Information-Maximization Approach to Blind Separation and Blind DeconvolutionNeural Computation, 1995