Pixel1.gif (51 bytes)
Pixel1.gif (51 bytes)
Pixel1.gif (51 bytes) Main Page Pixel1.gif (51 bytes)
About DSP Laboratory
Pixel.gif (52 bytes)
Contact Us
Pixel.gif (52 bytes)
Go to FIU's Homepage


 Pixel1.gif (51 bytes)


Curve.gif (104 bytes) Pixel1.gif (51 bytes)

An empirical study in human-computer interface research using EEG signals recorded with an ESI-256 machine.

Pixel1.gif (51 bytes)

"An empirical study in human-computer interface research using EEG signals recorded with an ESI-256 machine.", (2000)
Cremades, J.G., Sanchez, D., Adjouadi, M., and Barreto, A.B.

Humans use interfaces (e.g. keyboard, mouse) to interact with computers. A specific task may become more or less difficult depending on the computer interface that an individual is using. To compare task difficulty, researchers have investigated the levels of mental effort utilizing current technology such as Electroencephalography (EEG) (Gevins, Smith, Leong, McEvoy, Whitfield, Du, & Rush, 1998). Alpha activity may be viewed in three distinct alpha frequency bands (8-13 Hz, 8-10 Hz, and 11-13 Hz). The alpha frequency band (8-13 Hz) of the EEG is associated with reduced levels of consciousness and awareness, and the reduction of alpha activity (alpha blocking) indicates sensory stimulation or increased mental activity (Shaw, 1992). The upper alpha band (11-13 Hz) responds selectively to the encoding of the stimulus, whereas the lower alpha component (8-10 Hz) reflects processes related not only to cognitive processing but also to the energetic mechanisms of arousal, attention and effort (Klimesch, Pfurtscheller, & Schimke, 1992). This suggests that the three alpha wavebands are appropriate measures to investigate differences in mental effort. The purpose of this study was to investigate changes in alpha activity associated with the execution of basic tasks through two types of computer interfaces. The focus was to observe differences between the type of task (visuospatial/synthetic vs. verbal/analytic), the type of computer interface (keyboard vs. mouse), and the two cerebral hemispheres (left vs. right), as related to changes in the alpha wavebands. To prepare for EEG data acquisition, a 256-electrode cap was placed on the scalp and reference electrodes were placed on both earlobes. Each electrode in the cap was filled with conductive gel, and electrode impedances were kept below 5 kilo-ohms. The tests consisted of two different tasks. In one task, subjects played a video game that exercises visuospatial skills (SP). The game chosen was 3DTetriMania, which consists of organizing figures according to their shapes. In the other task, subjects played a video game that exercises verbal skills (VE). The game chosen for this task was TrackWords, which consists of selecting letters from a given pool to create words of three letters or more. Each subject played both video games with each of two computer interfaces: the keyboard (K) and the mouse (M). Subjects were randomly assigned to counterbalanced conditions of task and computer interface (i.e., SP/K, SP/M, VE/K, and VE/M). For each game, the subjects played a 1-minute practice round to become familiar with the task. Then they played one timed round of the game for a total of 1 minute per game. EEG signals were collected with the Neuro Scan Electrical Signal Imaging System with 256 channels (ESI-256). The Acquire sub-module of the Scan 4.0 program was used to record continuous EEG data while testing each subject, at a sampling frequency of 1000 Hz, per channel. A total of 9 male subjects were tested. Four 60-second data files were obtained per subject. At the conclusion of the testing, each subject completed a questionnaire on their previous experience in using both games: all subjects had past experience with the games. The power averages (within each one of the wavebands) for each region of the brain in each of the hemispheres (left/right) was obtained. The absolute mean alpha (8-13 Hz), lower alpha (8-10 Hz), and upper alpha (11-13 Hz) power values obtained for each subject from each region of the brain was used to analyze the data. These values were entered into a repeated measures design (task x computer interface x hemisphere). Analyses were performed for each pair of active cerebral regions in the three alpha frequency bands. Results revealed that there were significant differences in hemispheric activation at all cerebral regions in the alpha waveband (p<0.05). Mean alpha power values were greater in the left hemisphere as opposed to the right hemisphere. There were no significant differences (p>0.05) in alpha activities when comparing the interfaces - mouse vs. keyboard. There were significant differences (p<0.05) at the frontal sites in the alpha (8-13 Hz) and lower alpha (8-10Hz) wavebands when comparing the Tetris vs. TrackWord tasks. Mean alpha and lower alpha power values were greater when playing the Tetris task. Furthermore, a significant interaction effect task by computer interface was found at the temporal sites (p<0.05) in the three alpha wavebands. The greater levels of alpha activity in the left hemisphere suggests that both tasks required visuospatial processes. The greater levels of alpha activity at the frontal lobe when playing Tetris suggest that TrackWord was more difficult to perform in respect to the motor domain. This could be due to the fact that subjects had prior experience on the Tetris video game. A significant interaction effect task by computer interface at the temporal sites was revealed. In this interaction, individuals showed greater alpha power when playing Tetris with the mouse and greater alpha power when playing TrackWord with the keyboard. This suggests that Tetris became a less difficult task when played with the mouse and TrackWord became a less difficult task when played with the keyboard. It appears that the integration of communication achieved from computer to user and user to computer by using the mouse as a human computer interface facilitates the Tetris task and by using the keyboard as interface facilitates the TrackWord task.