Development of a Wearable and Cost Effective Brain-Computer Interface Assistive Device

A Brain-Computer Interface (BCI) is an interface that allows to transform the user's intent into actions without entailing muscle movements. BCIs could be used as assistive technology by individuals whose neuromuscular channels are impaired. BCIs commonly make use of bulky, expensive, and clinical-grade EEG equipment, making an in-home use uncomfortable for the user. In this work, we present the development of a wearable and cost-effective BCI assistive device. Electroencephalography signal is acquired using the OpenBCI Cyton board, which is a low-cost hardware for biological signal acquisition. A 3D printed headset is used to house the electrodes. The proposed assistive device functions based on the P300 response of the human brain, and flashing symbols on a grid are employed as stimuli to elicit a P300 response. Online detection of the P300 response is achieved using logistic regression, and a genetic algorithm is used to perform automatic feature extraction. The device allows the disabled individual to type words on a computer screen and to control a 4 DoF robotic arm. Two modes of operation can be used to control the robotic arm. One mode consists of controlling the robotic arm with discrete movements. The second control mode consists of sending high level commands to the robotic arm, that would in turn move autonomously. Therefore, the presented assistive device augments communication capabilities and allows for the control of external devices. This device serves as a proof of concept that cost-effective EEG acquisition systems along with robust classification techniques can be used in the field of assistive technologies. Future improvements are necessary to aid the definitive transition from laboratory setting to in-home use, allowing many disabled individuals to improve their quality of life.