Beyin bilgisayar arayüzlerinin video oyunlarının ana kontrolünde ve dinamik zorluk ayarlamasında kullanımı

Brain computer interface is a communication pathway between humans and computers. The acquired EEG signals are translated into a command using signal processing, machine learning and statistical methods. Brain computer interfaces are generally designed for patients with severe disabilities such as motor impairments, but there are also many applications for healthy users. In this study, two BCI based video games were developed. In one of the games, the game is fully controlled by making use of steady state visual evoked potentials. The average accuracy of the controls of this game was found to be 0.87. In the second game, even though the game is controlled by keyboard, the affective states of the player are predicted and the game difficulty is arranged accordingly. The valence and arousal predictions were tested in an offline analysis, which yielded 0.72 and 0.85 accuracy for valence and arousal levels respectively. The implemented dynamic difficulty adjustment system making use of the valence and arousal levels predicted in real time performed well enough to change the game difficulty to an enjoyable level.