• DocumentCode
    2026905
  • Title

    Modified pattern-reversal visual checkerboard stimuli with dual alternating frequencies for multi-class ssvep-based brain-computer interfaces

  • Author

    Chang-Hee Han ; Han-Jeong Hwang ; Chang-Hwan Im

  • Author_Institution
    Dept. of Biomed. Eng., Hanyang Univ. Seoul, Seoul, South Korea
  • fYear
    2013
  • fDate
    18-20 Feb. 2013
  • Firstpage
    86
  • Lastpage
    88
  • Abstract
    In this study, a new dual-frequency stimulation method that can produce more visual stimuli with limited number of stimulation frequencies was proposed for use in multiclass SSVEP-based BCI systems. The new stimulation was based on a conventional black-white checkerboard pattern; however, unlike the conventional approach, ten visual stimuli could be generated by combining four different stimulation frequencies. Through the offline experiments, we confirmed that all ten visual stimuli could evoke distinct and discriminable single SSVEP peaks, of which the SNRs were high enough to be used for practical SSVEP-based BCI systems. In order to demonstrate the feasibility of the proposed method, a mental keypad system was implemented and online experiments were conducted with additional ten participants, from which we achieved an average information transfer rate of 33.26 bits per minute and an average accuracy of 87.23 %.
  • Keywords
    brain-computer interfaces; medical signal processing; visual evoked potentials; SNR; black-white checkerboard pattern; dual-frequency stimulation method; mental keypad system; multiclass SSVEP-based brain-computer interface; pattern-reversal visual checkerboard stimuli; signal-to-noise ratio; steady-state visual evoked potential; stimulation frequency; Accuracy; Brain-computer interfaces; Electrodes; Signal to noise ratio; Steady-state; Time-frequency analysis; Visualization;
  • fLanguage
    English
  • Publisher
    ieee
  • Conference_Titel
    Brain-Computer Interface (BCI), 2013 International Winter Workshop on
  • Conference_Location
    Gangwo
  • Print_ISBN
    978-1-4673-5973-3
  • Type

    conf

  • DOI
    10.1109/IWW-BCI.2013.6506640
  • Filename
    6506640