Successfully Controlled BCI Through Minimal Dry Electrodes
Date
2023-08-08
Access
Authors
Lawton, Joshua
Journal Title
Journal ISSN
Volume Title
Publisher
East Carolina University
Abstract
ABSTRACT:
There are approximately 185,000 amputations a year in the United States according to the Amputee Coalition with the number of amputations going up. While it is common for someone with a lower limb amputation to use a prosthetic, approximately 84%, it is not as common for people with upper limb amputations, approximately 56% (Raichle et al., 2008). The time it takes an amputee to get a prosthetic affects the likelihood of use, in addition to functionality (Miller et al., 2020). The purpose of this project is to show proof of concept of an EEG-controlled prosthetic, using only 2 dry-electrodes, through the use of BCI2000 using imagined movements. Eight (N-8) participants were recruited to complete a pre-training mu task, a 1D cursor training task, a 2D cursor training task, and the main 2D cursor task.
After a frequency was established for each participant, they completed 200 trials of the 1D cursor task for three different conditions (left, right, and both hand(s)) or reached a success rate of 80% for 4 trials in a row with random targets. The participants then completed the 2D cursor task with random targets until a success rate of 70% for 4 trials in a row was achieved, followed by a 2D cursor task where the targets were pre-determined. A chi-squared test determined the goodness of fit for the success rate was significant (p < 0.001) for all participants completing the 1D cursor task. The combined success rate for the participants during task 1 for their right hand was 30.16%, 47.11% for their left hand, and 61.47% for both hands. The combined success rate for task 2 was 69.40% and 79.59% for the main task.
Overall, this study successfully showed that 2 dry electrodes can be used to detect imagined movements through BCI. While the accuracy can still be improved, by enhancing the equipment and developing the training protocol, both participants that completed the main task were able to surpass the expected overall accuracy and surpass 4 out of the 6 individual accuracies. Whether it is to control a mechanical arm, leg, or other body part, the framework of this study grants development opportunities for BCI from a few dry electrodes.