Information in the Local Field Potential: Implications for Brain-Machine Interfaces

The last decade has seen a surge in the development of brain-machine interfaces (BMIs) as assistive neural devices for paralysis patients. BMIs are devices that decode neural activity to provide control signals for external devices, computers or prostheses. Current BMI research typically involves a subject playing a computer game or controlling a robotic prosthesis through neural activity (brain control). The local field potential (LFP) is a low frequency neural signal recorded from intra-cortical electrodes, and has been recognized as one containing movement information. This thesis investigates time and frequency properties of the LFP from the perspective of developing upper limb neuroprosthetic BMIs, and touches on three major topics.