Probabilistic 3-dimensional column flashing based brain computer interface speller system that works simultaneously and subject independent

Brain-Computer Interfaces (BCI) are communication systems that enable people to use computers, wheelchairs, or neuroprostheses without muscle systems. The primary purpose of research in EEG-based BCI is to develop a new method with higher classification accuracy and data rate than existing methods and reveal new applications. One of the most popular EEG-based BCI systems is the P300-based spelling model. This model works based on capturing the user's P300 signals when the row and column of the target character are flashed, along with randomly flashing the rows and columns of the letters in a matrix. In this thesis, a fast and high-performance brain-computer interface has been developed that enables individuals with paralysis such as ALS to communicate with their environment in real time by writing. Another unique aspect of this study is that the developed model is independent of the person; it has a plug-and-use feature without extra training or calibration steps for a new user. By calculating the probability of the desired word from the dictionary available in the database, the BCI system has been developed, which enables the next letter to be flashed more than conventional methods. In the BCI system, a 3-dimensional column stimulus presentation paradigm was used, which is probabilistic and ensures that the P300 amplitude is large. The proposed person-independent probabilistic model working online was tested with 15 people, and character detection was performed with an accuracy of 92.17%.