Design And Implementation Of A Brain Computer Interface System Based On Steady State Visual Evoked Potentials

Brain Computer Interface(BCI)is an external information exchange and control technology between human brain and other electronic devices,which does not rely on general information output pathways of human brain(peripheral nerve and muscle tissue).The main purpose of BCI is to help disabled people without muscle activities with a new way of communicating and controlling the environment.Recently great progresses can be seen in the development of BCI technology in the past 17 years.However,troubled by the facts that EEG signal has low signal-to-noise ratio,high complexity and poor resistance to noise,crucial challenges still exist in developing mature BCI systems.This paper focused on the research of key technologies of BCI system and managed to develop a stable SSVEP-based BCI system with four different options for selection.This system is consisted of an 8-trails EEG recorder and SSVEP controller and both were portable,simple,low cost and mature.Experiments have confirmed that the average accuracy of this system has reached 96.5% and performed as well as,or even better than published papers.This BCI system is practical in helping disabled people in their daily activities and provides them with a new method in communicating with external environment,and therefore,improve their quality of life.Compared with existing SSVEP-based BCI systems,the innovations from this paper are the followings:1)This paper designed all the hardware circuits of the entire BCI system,rather than completing the system by splicing purchased equipment.As a result,the functional division of the system,mode of operation,equipment size,equipment costs could be fully customized to achieve the system’s portability,ease of use,low cost and practicality.2)In this paper,the SSVEP recognition algorithm was designed and implemented in the selfdesigned EEG recorder.The recorder completes signal acquisition and SSVEP recognition and directly outputs the identification results.The system does not require a separate computing device(computer or DSP device)for signal identification.This is the most significant difference between this paper and existing BCI systems.This design eliminated the hardware redundancy of existing BCI systems,and further enhanced the portability,ease of use and usability of the system,reduced system cost.3)In the SSVEP recognition algorithm,this paper proposed and designed a DC baseline calibration preprocessing algorithm based on the segmentation mean value to solve the problem of data overflow,resources demand and operational efficiency in hardware programming.At the same time,this paper combined the fixed-point fraction with Fast Fourier Transformation(FFT),the FFT algorithm based on fixed-point fraction was applied to SSVEP recognition for the first time,which reduced hardware resources demand of the SSVEP recognition algorithm.Thus,SSVEP recognition can be quickly processed in microcontroller which has limited resources and no float-point computing ability.4)The system designed in this paper provides two developmental output interfaces: Simulated keyboard output simulates the entire BCI system into a keyboard,which can directly manipulate existing computer software;Pulse output is used to control other electronic devices.Compared with other existing BCI systems,the system designed in this paper is no longer limited to only operate self-developed demo software.