An amplitude-modulated visual stimulation for reducing eye fatigue in SSVEP-based brain-computer interfaces

作者：Chang Min Hye; Baek Hyun Jae; Lee Seung Min; Park Kwang Suk^*

来源：Clinical Neurophysiology, 2014, 125(7): 1380-1391.

DOI：10.1016/j.clinph.2013.11.016

• Brain-computer interface (BCI)
• Steady-state visual evoked potential (SSVEP)
• Amplitude-modulated stimulus
• Electroencephalogram (EEG)

Objective: A high-frequency steady-state visual evoked potential (SSVEP) has been suggested for the reduction of eye fatigue for SSVEP-based brain-computer interfaces (BCIs). However, the poor performance of high-frequency SSVEP requires a novel stimulus of better performance even with low eye fatigue. As an alternative to the high-frequency SSVEP, we explore the SSVEP response to an amplitude-modulated stimulus (AM-SSVEP) to verify its availability for brain-computer interfaces (BCIs).
Methods: An amplitude-modulated stimulus was generated as the product of two sine waves at a carrier frequency (f(c)) and a modulating frequency (f(m)). The carrier frequency was higher than 40 Hz to reduce eye fatigue, and the modulating frequency ranged around the alpha-band (9-12 Hz) to utilize low-frequency harmonic information. Four targets were used in combinations of three different modulating frequencies and two different carrier frequencies in the offline experiment, and two additional targets were added with one additional modulating and one carrier frequency in online experiments.
Results: In the AM-SSVEP spectra, seven harmonic components were identified at 2f(c), 2f(m), f(c) +/- f(m), f(c) +/- 3f(m), and 2f(c) 4f(m). Using an optimized combination of the harmonic frequencies, online experiments demonstrated that the accuracy of the AM-SSVEP was equivalent to that of the low-frequency SSVEP. Furthermore, subject evaluation indicated that an AM stimulus caused lower eye fatigue and less sensing of flickering than a low-frequency stimulus, in a manner similar to a high-frequency stimulus.
Conclusions: The actual stimulus frequencies of AM-SSVEPs are in the high-frequency band, resulting in reduced eye fatigue. Furthermore, AM-SSVEPs can utilize both fundamental stimulus frequencies and non-integer harmonic frequencies including low frequencies for SSVEP recognition. The feasibility of AM-SSVEP with high BCI performance and low eye fatigue was confirmed through offline and online experiments.
Significance: AM-SSVEPs combine the advantages of both low- and high-frequency SSVEPs - high power and low eye fatigue, respectively. AM-SSVEP-based BCI systems exploit these advantages, making them promising for application in practical BCI systems.

• 出版日期2014-7