Aquisição e condicionamento de sinais cerebrais por eletroencefalografia

Abstract

Recent developments of studies in the area of signal processing have made the development of new techniques of biopotential analysis possible, such as electroencephalography signals (EEG). This evolution allows the development of new brain-machine interfaces (BMI) capable of offering alternative solutions that help people with motor diseases. In this context, the current work proposes a platform for the acquisition of EEG signals and an algorithm for their processing and analysis, capable of identifying if the studied signal corresponds to a subject who is blinking or with closed eyes. The proposed platform is based on the demonstration kit ADS1299EEG-FE together with the Arduino Uno, making it possible from the biosignal manipulation until their sending to the computer. The algorithm performs the digital filtering of the received data and their analysis in the time and frequency domains, making pattern identification related to the ocular movement feasible. The platform is tested through the signals of 30 individuals available in the PhysioNet databank, reaching a success rate of 87% for blinking identification and of 67% for the identification of subjects with closed eyes. From these results, this study suggests the application of the solution presented in future studies of the acquisition of EEG signals.