V4 N1 Paper 1
Annals of the MS in Computer Science and Information Systems at UNC Wilmington
Spring 2010

Using 3D Video Game Scenarios and Artificial Neural Networks to Classify Brain States for a Brain Computer Interface  

Maurice Benson

Committee

Karl Ricanek (chair)
Ling He
Devon Simmonds

Abstract

A brain-computer interface (BCI) is a technology that allows the user to interact with a computer without relying on any overt physical activity, i.e. hand movement as with a mouse or finger movement with a keyboard. A BCI works by monitoring the user's brain signals, extracting the important features, using the features to classify their intent, and then providing feedback to a computer application. This study focuses on an Electroencephalograph (EEG) based BCI designed to classify two dimensional manipulation of a remote control for the Nintendo Wii™ game console (Wiimote). The long-term goal of this project is to allow users the ability to interact with the system without having to physically operate the Wii remote. Such a system would allow a physically impaired user the ability to interact with the system. For this study subjects were asked to play a simple 3D video game using the Wiimote. First, subjects had to respond to stimuli by holding the wiimote with both hands and tilting it to the left, right, up or down. Next, subjects were instructed to imagine manipulating the Wiimote to the same stimuli while attempting to avoid any physical activity. During the game, EEG was recorded from 64 electrodes covering the subject's scalp. Linear regression techniques were used to extract the important components in the data. These components were then used as input features for an Artificial Neural Network to classify the user?s intentions. The results of this experiment prove that EEG contains movement related classifiable information for both physical and imagined movements.

download (pdf)

Recommended Citation: Benson, M., Ricanek, K, He, L., Simmonds, D. (2010) Using 3D Video Game Scenarios and Artificial Neural Networks to Classify Brain States for a Brain Computer Interface. Annals of the Master of Science in Computer Science and Information Systems at UNC Wilmington, 4(1) paper 1. http://csbapp.uncw.edu/data/mscsis/full.aspx.

V4 N1 Paper 1
Annals of the MS in Computer Science and Information Systems at UNC Wilmington
Spring 2010