A novel probabilistic and 3D column P300 stimulus presentation paradigm for EEG-based spelling systems


KORKMAZ O. E., AYDEMİR Ö., ORAL E. A., ÖZBEK İ. Y.

NEURAL COMPUTING & APPLICATIONS, cilt.35, sa.16, ss.11901-11915, 2023 (SCI-Expanded) identifier identifier

  • Yayın Türü: Makale / Tam Makale
  • Cilt numarası: 35 Sayı: 16
  • Basım Tarihi: 2023
  • Doi Numarası: 10.1007/s00521-023-08329-y
  • Dergi Adı: NEURAL COMPUTING & APPLICATIONS
  • Derginin Tarandığı İndeksler: Science Citation Index Expanded (SCI-EXPANDED), Scopus, Academic Search Premier, PASCAL, Applied Science & Technology Source, Biotechnology Research Abstracts, Compendex, Computer & Applied Sciences, Index Islamicus, INSPEC, zbMATH
  • Sayfa Sayıları: ss.11901-11915
  • Karadeniz Teknik Üniversitesi Adresli: Evet

Özet

Brain-computer interfaces can allow users to select letters or other targets on a computer screen without any muscular activity. One of the most popular EEG-based spelling systems is employed with P300 potentials, eliciting the brain's positive electrical response to a flash-denoting gaze at a target character. A P300 spelling system generally uses the row-column paradigm, which displays letters in a matrix and alternately flashes the rows and columns in a randomized order. At the same time, the participant focuses on a target letter. Afterward, a classification algorithm determines the row and column that elicited the largest P300 amplitude and finds the intersection of the identified row and column that indicates the target letter. This paper proposes a probabilistic and 3D column P300 stimulus presentation paradigm for a high-performance spelling system. In contrast to the classical 2D row-column paradigm, we utilized a 3D column stimulus presentation in the proposed paradigm, and we adaptively determined the flashing numbers of columns according to the probability of the following letter in the dictionary between 4 and 13. While the subject-independent P300 speller model was trained with the EEG signals of the first 10 participants, it was independently tested with the second group, including 15 volunteers. The results showed that the proposed probabilistic and 3D column P300 stimulus presentation paradigm-based method achieved 92.69% average classification accuracy, which is 9.94% higher than that of the non-probabilistic 3D column paradigm. The obtained experimental results show the effectiveness and potential of the proposed method to achieve very high performance for brain-computer interface spelling systems.