Test Bed for Performance of Single Degree Freedom Movement

Abstract

Research in brain computer interfaces (BCI) focus mainly on neural-prosthesis. Recently commercial electroencephalograms (EEG), a non-invasive BCI, have been introduced into the market which decreases the cost and increases the availability for everyday use. A standard evaluation of BCI performance is lacking, therefore in thesis work, a test bed is developed to measure performance of a BCI for single degree freedom movement. The classic video game Pong is used to measure and compare the performance of different users, input devices, and control algorithms. The first experiment recreated the Fitts' Law reciprocal tapping task to validate the of measure performance for different paddle sizes and distance between targets. The second experiment extends Fitts' Law to a Pong game. Our results for the second experiment show there is a main effect where paddle size affects throughput (F(5,812) = 19.77, p < 0.0001). Since there was a main effect, a pairwise t-test was used to compare performance between different paddle sizes. Future research will compare the performance for different input devices and signal processing algorithms.