Improved accuracy using recursive bayesian estimation based language model fusion in ERP-based BCI typing systems.

RSVP Keyboard™ is an electroencephalography (EEG) based brain computer interface (BCI) typing system, designed as an assistive technology for the communication needs of people with locked-in syndrome (LIS). It relies on rapid serial visual presentation (RSVP) and does not require precise eye gaze control. Existing BCI typing systems which uses event related potentials (ERP) in EEG suffer from low accuracy due to low signal-to-noise ratio. Henceforth, RSVP Keyboard™ utilizes a context based decision making via incorporating a language model, to improve the accuracy of letter decisions. To further improve the contributions of the language model, we propose recursive bayesian estimation, which relies on non-committing string decisions, and conduct an offline analysis, which compares it with the existing naïve bayesian fusion approach. The results indicate the superiority of the recursive bayesian fusion and in the next generation of RSVP Keyboard™ we plan to incorporate this new approach.

On Visually Evoked Potentials in EEG Induced by Multiple Pseudorandom Binary Sequences for Brain Computer Interface Design.

Visually evoked potentials have attracted great attention in the last two decades for the purpose of brain computer interface design. Visually evoked P300 response is a major signal of interest that has been widely studied. Steady state visual evoked potentials that occur in response to periodically flickering visual stimuli have been primarily investigated as an alternative. There also exists some work on the use of an m-sequence and its shifted versions to induce responses that are primarily in the visual cortex but are not periodic. In this paper, we study the use of multiple m-sequences for intent discrimination in the brain interface, as opposed to a single m-sequence whose shifted versions are to be discriminated from each other. Specifically we used four different m-sequences of length 31. Our main goal is to study if the bit presentation rate of the m-sequences have an impact on classification accuracy and speed. In this initial study, where we compared two basic classifier schemes using EEG data acquired with 15Hz and 30Hz bit presentation rates, our results are mixed; while on one subject, we got promising results indicating bit presentation rate could be increased without decrease in classification accuracy; thus leading to a faster decision-rate in the brain interface, on our second subject, this conclusion is not supported. Further detailed experimental studies as well as signal processing methodology design, especially for information fusion across EEG channels, will be conducted to investigate this question further.

Performance of prewhitening beamforming in MEG dual experimental conditions.

This paper presents an analysis on the performance of the prewhitening beamformer when applied to magnetoencephalography (MEG) experiments involving dual (task and control) conditions. We first analyze the method's robustness to two types of violations of the prerequisites for the prewhitening method that may arise in real-life two-condition experiments. In one type of violation, some sources exist only in the control condition but not in the task condition. In the other type of violation, some signal sources exist both in the control and the task conditions, and that they change intensity between the two conditions. Our analysis shows that the prewhitening method is very robust to these nonideal conditions. In this paper, we also present a theoretical analysis showing that the prewhitening method is considerably insensitive to overestimation of the signal-subspace dimensionality. Therefore, the prewhitening beamformer does not require accurate estimation of the signal subspace dimension. Results of our theoretical analyses are validated in numerical experiments and in experiments using a real MEG data set obtained during self-paced hand movements.

Performance comparison of independent component analysis algorithms for fetal cardiac signal reconstruction: a study on synthetic fMCG data.

Independent component analysis (ICA) algorithms have been successfully used for signal extraction tasks in the field of biomedical signal processing. We studied the performances of six algorithms (FastICA, CubICA, JADE, Infomax, TDSEP and MRMI-SIG) for fetal magnetocardiography (fMCG). Synthetic datasets were used to check the quality of the separated components against the original traces. Real fMCG recordings were simulated with linear combinations of typical fMCG source signals: maternal and fetal cardiac activity, ambient noise, maternal respiration, sensor spikes and thermal noise. Clusters of different dimensions (19, 36 and 55 sensors) were prepared to represent different MCG systems. Two types of signal-to-interference ratios (SIR) were measured. The first involves averaging over all estimated components and the second is based solely on the fetal trace. The computation time to reach a minimum of 20 dB SIR was measured for all six algorithms. No significant dependency on gestational age or cluster dimension was observed. Infomax performed poorly when a sub-Gaussian source was included; TDSEP and MRMI-SIG were sensitive to additive noise, whereas FastICA, CubICA and JADE showed the best performances. Of all six methods considered, FastICA had the best overall performance in terms of both separation quality and computation times.

NUTMEG: a neuromagnetic source reconstruction toolbox.

We have developed an analysis toolbox called NUTMEG (Neurodynamic Utility Toolbox for Magnetoencephalography) for reconstructing the spatiotemporal dynamics of neural activations and overlaying them onto structural MR images. The toolbox runs under MATLAB in conjunction with SPM2 and can be used with the Linux/UNIX, Mac OS X, and even Windows platforms. Currently, evoked magnetic field data from 4-D Neuroimaging, CTF, and KIT systems can be imported to the toolbox for analysis. NUTMEG uses an eigenspace vector beamforming algorithm to generate a tomographic reconstruction of spatiotemporal magnetic source activity over selected time intervals and spatial regions. The MEG coordinate frame is coregistered with an anatomical MR image using fiducial locations and, optionally, head shape information. This allows the reconstruction to be superimposed onto an MRI to provide a convenient visual correspondence to neuroanatomy. Navigating through the MR volume automatically updates the displayed time series of activation for the selected voxel. Animations can also be generated to view the evolution of neural activity over time. Since NUTMEG displays activations using SPM2's engine, certain SPM functions such as brain rendering and spatial normalization may be applied as well. Finally, as a MATLAB package, the end user can easily add customized functions. Source code is available at http://bil.ucsf.edu/ and distributed under a BSD-style license.