The LF-ASD brain computer interface: on-line identification of imagined finger flexions in the spontaneous EEG of able-bodied subjects

Abstract

This research has focused on developing a Brain Computer Interface (BCI) for asynchronous control applications, which are characterized by alternating periods of active control and attentive idleness. We have developed an asynchronous switch that users can control through their EEG. The on-line implementation of the Low Frequency Asynchronous Switch Design (LF-ASD) has shown promising results with actual index finger flexions with able-bodied subjects. This work reports the results of the algorithm on imagined voluntary movements with able-bodied persons, demonstrating hit (true positive) rates above 70% and false positive rates below 3%. This work is a precursor to verifying the online switch operation for people with severe motor disabilities. the brain. Since voluntary movement control is an existing, internal control system in humans, which seemed naturally suited to drive a BCI, the LF-ASD feature set was based on voluntary movement-related potentials (VMRPs). Lisogurski [3] produced an on-line implementation of the LF-ASD capable of identifying VMRPs from a continuous sampling of EEG.Promising results have been obtained to date with ablebodied subjects performing actual index finger flexions. It is yet required to prove the use of the LF-ASD with imaginary movements so that it can be applied by persons with high-level disabilities. The data presented in this paper has been obtained by applying a new methodology to able-bodied subjects who were asked to imagine a right-hand index finger Hexion. This is an important step in this research since i t establishes that the LF-ASD feature set is applicable to imagined voluntary movements as well as actual movements.

A Comprehensive Survey of Brain Interface Technology Designs

“…The column on the right summarizes the number of research groups (g), journal papers (jp) and conference papers (cp) within each attribute sub-class. 17 (30,16) In/over somatosensory cortex [22,24] 17 (29,9) In/over occipital cortex [33,54,31,32] which design approaches have received little or no attention and, finally, how well designs are reported. Each of these topics is presented in a separate subsection below.…”

A comparison approach toward finding the best feature and classifier in cue-based BCI

“…To increase the performance of BCI systems, much research has been done by BCI research groups. In this way, the Graz-BCI research group has employed discriminative features based on second order statistics such as bandpower [14], adaptive autoregressive coefficients (AAR) [18], and wavelet coefficients [19], and also combination of features by distinction sensitive learning vector quantization (DSLVQ) [8] with well-known classifiers containing Fisher's linear discriminant analysis (FLDA) [13], finite impulse response multi layer perceptrons (FIRMLP) [2], linear vector quantization (LVQ) [17], hidden Markov models (HMM) [8], and minimum distance classifiers [18] to improve the classification rate between the various movement imagery tasks. Deriche and Al-Ani [3] selected the best feature combination among the variance, AR coefficients, wavelet coefficients, fractal dimension by modified mutual information feature selection (MMIFS) method.…”

The LF-ASD brain computer interface

“…They demonstrated that the LF-ASD was capable of identifying voluntary movement-related potentials (VMRPs) from a continuous sampling of EEG. Bozorgzadeh et al [4] evaluated the first on-line implementation of the LF-ASD with able-bodied subjects using imagined finger flexions. They demonstrated that the LF-ASD capable of identifying IVMRPs from a continuous EEG.…”

“…high-level disabilities to use the LF-ASD was unknown. The data presented in this paper has been obtained by applying a similar evaluation as that used in [4] to subjects with high-level spinal cord injuries who were asked to imagine a right-hand index finger flexion. This is an important step in this research since it establishes the usability of the LF-ASD within the group of individuals with high-level spinal-cord injuries.…”