Brain-Computer Interface for the Muscularly Disabled 

Muscular disability may deny computer access. Brain-computer interfaces (BCI’s) using non-invasive electroencephalographs (EEGs) may allow computer access by substituting for a keyboard and/or mouse. This invention is a program to interpret EEG signals as commands for a computer. The system uses state-of-the-art signal analysis and pattern recognition techniques to achieve internationally competitive accuracy on a BCI task, which was to predict which hand was about to move based on the subject’s EEG. In the process, a filter specialized for EEG signals was created from a simple model of the neuron action potentials the EEG measures. The first version of this filter (wavelet) tied a standard wavelet in performance, and outperformed all other standard wavelets tested. This is a critical step in the development of a working prototype.