Brain Computer Interface Game Controlling Using Fast Fourier Transform and Learning Vector Quantization

Abstract

Brain Computer Interface (BCI) is a direct communication pathways which enables our brain to real time control a robotic movement or a game, which uses EEG signal to capture information of a human mind. In this research, we developed an arcade game that are controlled by BCI. Player sat and used their imagination to move the object in the game, in this case, they moved up or down. To achieve that, a wireless EEG device was used to record electrical activity in the player’s brain; each action was segmented by two seconds per frame, and then extracted using Fast Fourier Transform. Continuously, the output was passed to Learning Vector Quantization networks to classify two different motor imagery-related brain patterns (imagination of limb movements: upward and downward). Before the game was used, we conducted training for ten people (subjects) with four repetitions of each thinking type. Then, it was tested with 10 other subjects in which resulted in 70% accuracy. The game was tested and then compared between the movement of the mouse and BCI of the same subject and great results were found in both conditions.