Abstract
Purpose :
Steady-State Visually Evoked Potentials (SSVEPs) are the brain's response to periodical optical stimuli. They are great for analyzing and investigating the function of the whole visual system. No conscious patient feedback is required.
We looked at variations of conventional checkerboard patterns to determine how onset/reversal, superimposed noise, or pictures of faces can improve the elicited signals' magnitude. The resulting comparisons may be used to improve existing SSVEP-based procedures using novel types of stimuli.
Methods :
Using our custom system, we designed two stimuli sets. Set 1 using checker and faces (7s) and set 2 using dartboard, checker, and faces (6s), both at 7.5Hz. The sets were shown to two groups of 5 participants each (age: 20-30). The resulting reactions to the SSVEP were recorded using an 8-channel EEG (500 Hz) with electrodes positioned on the occiput [P07,O1,Oz,O2,P08,P03,P04,Fz]. The data was analyzed using the Canonical Correlation Analysis (CCA) calculated over the stimulation window using the stimulation frequency and harmonics. Plots represent the mean value of all subjects in the set for the CCA.
Results :
Set 1 shows a significant increase in evocation when using pattern onset instead of reversal. Showing faces instead of checkers lead to a slightly increased response. Set 2 indicates that cartoon characters elicit a stronger response than a checker pattern. Through the addition of noise, the response increased at low levels but decreased at higher levels.
Conclusions :
SSVEP-based assessment of neuro-visual functions has immense potential in ophthalmology. However, checkerboard stimulations do not necessarily provide the best responses. The addition of low-level noise to the stimuli might improve the generation of SSVEPs in participants. Higher-level stimulation patterns like faces could improve the responses even further.
We suggest that Researchers in Neuro-Ophthalmology consider those or similar stimulation paradigms to improve SSVEP-based objective assessment methods. Especially stimulation with more complex structures like faces could lead to promising results.
This is a 2021 ARVO Annual Meeting abstract.