Wearable parallel processing based high-resolution high-speed electroencephalogram monitoring integrated system

Daily use of electroencephalogram (EEG) could increase the effectiveness for medical applications. However, current brain monitoring technologies lack one or more characteristics necessary for daily use, such as portability, responsiveness or versatility. In this study we developed a portable prototype for high resolution real-time EEG monitoring fully integrated system. The prototype is composed of 112 custom made hybrid capacitive-resistive electrodes and up to 7 reference electrodes. The electrodes are connected together by using a flexible, elastic grid which is attached to an adjustable, link based mechanical headgear. Real time processing was performed using a mobile Compute Unified Device Architecture (CUDA) platform for massive parallel processing of the data. Experiments to test the capabilities of the prototype were performed. The first experiment consisted of monitoring the frontal lobe when applying visual stimulus using a lamp. When the lamp was turned on weak alpha waves were detected, while when the lamp was turned off alpha waves were detected. The second experiment consisted of monitoring the entire motor cortex while moving the right hand. At rest, strong mu-rhythm signals were detected over the entire scalp above the motor cortex whereas when moving the right hand mu-rhythm signals above the left motor cortex area were weakened. Both experiment results matched known brain activity phenomena demonstrating basic monitoring capabilities of the prototype. Furthermore we showed that our system has high portability, usability and capable of high responsiveness on high resolution conditions. The concepts introduced in this paper can not only improve EEG monitoring technologies but also contribute to other areas of wearable computing.