I. Introduction
Brain-computer interface (BCI) technology, which links neural signal based commands to computer based outputs, has advanced significantly over the past decade. The BCI research community aims to give people with motor disabilities the ability to use their neural signals to control devices ranging from computers to wheelchairs to prosthetic limbs. A major advantage of BCIs over other augmentative communication methods is its independence from muscle movement, which is required for all other devices. Many BCI studies have demonstrated that people can move a computer cursor to a target using neural signals collected from the scalp (electroencephalogram - EEG) [1]–[3], surface of the brain (electrocorticogram - ECoG) [4], [5], or within the brain (local field potential and action potential recordings) [6], [7]. This is typically accomplished by training the user to modulate their neural signals by performing motor imagery [1], [4], [5], [8]. However, with few exceptions [9], [10], even very simple computer based applications controlled with neural signals are not available for people with motor disabilities.