P300 BCI: A Simulation of Random Eye Movement
People who suffer from amyotrophic lateral sclerosis (ALS) eventually lose all voluntary muscle control. In the late stages of the disease, traditional augmentative and alternative communication (AAC) devices fail to provide adequate levels of communication. Braincomputer interface (BCI) technology has provided effective communication after all other AAC devices have failed. Nonetheless, EEG-based BCI devices may also fail for people with late-stage ALS due to loss of voluntary eye movement. Specifically, some people may suffer from random eye movement (nystagmus) and/or drooping of the eyelids (ptosis). The current study is the first attempt to simulate involuntary random eye movement in able-bodied individuals. This study employs the P300-based BCI. The system presents a matrix (6x6 in this study) of letters and numbers to participants on the computer screen and their task is to focus attention to a specific item within the matrix. The attended item of the matrix will produce a P300 event-related potential (ERP). The BCI determines which item produced the largest P300 ERP and presents (types) this item on the computer screen. To simulate involuntary random eye movement the 6x6 matrix would move in random directions in increments of 1–5 pixels “Jitter 1” or increments of 10–15 pixels “Jitter 2”. Movement (i.e., jitter) occurs during the inter-stimulus interval (ISI), which is the time between the offset of one character flash and the onset of another character flash. The matrix can move in the X dimension, Y dimension, or in both dimensions simultaneously. Participants complete two conditions: 1) control (i.e., no jitter) and 2) one of the two Jitter conditions (counter-balanced). Prior to each condition, each participant completed a calibration phase with eighteen character selections (three six letter words). Following calibration, each participant was presented with 18 more characters and the computer provided online feedback to indicate if the BCI selected the character the participant intended for it to choose. To date, six participants have completed the experiment. Three participants completed the control condition and Jitter 1 and three completed the control and Jitter 2. Preliminary data indicate highest accuracy (number of characters selected correctly) in Jitter 1 (100%), followed by Control (94%), and lowest accuracy in Jitter 2 (81%). In addition, participants were surveyed in regard to distraction and attentional focus. Surveys indicate that participants in Jitter 1 found it less distracting and easier to focus in Jitter 1 than control. Whereas, participants in Jitter 2 found it more distracting and harder to focus in Jitter 2 as compared to control. The data indicate that modest amounts of matrix movement may not be deleterious to performance, and may in fact improve performance through increasing attentional resources to the task. In Jitter 2 a decline in performance was observed; however, accuracy was still adequate for effective communication.
Johnson City, TN
Shubert, Kelsey N.; Kellicut-Jones, Marissa R.; Swymer, Caitlyn N.; Wiley, T. J.; Ryan, David B.; and Sellers, Eric W., "P300 BCI: A Simulation of Random Eye Movement" (2017). ETSU Faculty Works. 898.