Prediction of Brain-Computer Interface Performance: For P300 and Motor Imagery Brain-Computer Interfaces

Beschreibung

Brain-computer interfaces (BCIs) provide a communication channel that is independent of voluntary muscle control and thus enable completely paralyzed people to communicate with their environment. Despite continuous development of BCIs in the last years, about a third of users is unable to achieve significant control over the system. The goal of this dissertation was to determine predictors of the potential performance of BCI users. The prediction was to be performed on the basis of short EEG measurements, so that future users could quickly be divided into different aptitude categories. An optimal BCI paradigm and training strategy could then easily be determined. Performance prediction methods for two different BCI types were analyzed: BCIs on the basis of the modulation of the sensorimotor-rhythm (SMR) and BCIs on the basis of the visual or auditory P300 event-related potential (ERP) component. Both BCI types were evaluated with a group of healthy participants and a group of patients (SMR BCI with stroke patients, P300 BCI with amyotrophic lateral sclerosis (ALS) patients).
Brain-computer interfaces (BCIs) provide a communication channel that is independent of voluntary muscle control and thus enable completely paralyzed people to communicate with their environment. Despite continuous development of BCIs in the last years, about a third of users is unable to achieve significant control over the system. The goal of this dissertation was to determine predictors of the potential performance of BCI users. The prediction was to be performed on the basis of short EEG measurements, so that future users could quickly be divided into different aptitude categories. An optimal BCI paradigm and training strategy could then easily be determined. Performance prediction methods for two different BCI types were analyzed: BCIs on the basis of the modulation of the sensorimotor-rhythm (SMR) and BCIs on the basis of the visual or auditory P300 event-related potential (ERP) component. Both BCI types were evaluated with a group of healthy participants and a group of patients (SMR BCI with stroke patients, P300 BCI with amyotrophic lateral sclerosis (ALS) patients).

Autor*in

Sebastian Halder

Stimmen zu »Prediction of Brain-Computer Interface Performance: For P300 and Motor Imagery Brain-Computer Interfaces«

Details

Link teilen