Leveraging brain–computer interface for implementation of a bio-sensor controlled game for attention deficit people

Abstract

In video games, neurofeedback via Electroencephalogram (EEG) has emerged as a method for treating attention deficit, alongside preventative measures such as behavioral therapy. By 2020–21, the Neuro-Gaming industry has reached USD 6.29 billion. As a remedy to attention deficit and to take advantage of the ever-growing EEG-based gaming industry, this research work presents the design and implementation of an EEG-controlled 2D game built in the Unity 3D game engine. Our research includes steps like dataset creation, training the learning algorithms, classification, and deciding on those results in the designed game whether to shoot a target or not. We read signals from the Neurosky sensor, user orientation, and linear acceleration. We pre-process them via transforms into a processed input for various learning algorithms. The results are then exported to the game engine and used in the game. In classification, we have achieved 89% accuracy and F1 score of 87% with LSTM.