Application of rapid invisible frequency tagging for brain computer interfaces

Brickwedde, Marion; Bezsudnova, Yulia; Kowalczyk, Anna; Jensen, Ole; Zhigalov, Alexander

doi:10.1016/j.jneumeth.2022.109726

Cited by 14 publications

(9 citation statements)

References 41 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Indeed, the classification accuracy for the twoclass problem reached 93.7% and 81.1% for the control and periliminal conditions respectively. This classification performance surpasses the 60% classification accuracy achieved using high-frequency (using 56 and 60 Hz flickers) flickers displayed with high-refresh-rate projectors to elicit SSVEP responses recorded using MEG (Brickwedde et al, 2022). The improved user experience, coupled with the high classification performance achieved by periliminal flickers, has the potential to enhance initial engagement and user retention in future SSVEP-based applications.…”

Section: Discussionmentioning

confidence: 86%

Frequency-tagging of spatial attention using periliminal flickers

Ladouce,

Dehais

2024

Preprint

View full text Add to dashboard Cite

Steady-State Visually Evoked Potentials (SSVEP) manifest as a sustained rhythmic activity that can be observed in surface electroencephalography (EEG) in response to periodic visual stimuli, commonly referred to as flickers. SSVEPs are widely used in fundamental cognitive neuroscience paradigms and Brain-Computer Interfaces (BCI) due to their robust and rapid onset. However, they have drawbacks related to the intrusive saliency of flickering visual stimuli, which may induce eye strain, cognitive fatigue, and biases in visual exploration. Previous findings highlighted the potential of altering features of flicker stimuli to improve user experience. In this study, we propose to reduce the amplitude modulation depth of flickering stimuli down to the individuals’ perceptual visibility threshold (periliminal) and below (subliminal). The stimulus amplitude modulation depth represents the contrast difference between the two alternating states of a flicker. A simple visual attention task where participants responded to the presentation of spatially-cued target stimuli (left and right) was used to assess the validity of such periliminal and subliminal frequency-tagging probes to capture spatial attention. The left and right sides of the screen, where target stimuli were presented, were covered by large flickers (13 and 15 Hz respectively). The amplitude modulation depth of these flickers was manipulated across three conditions: control, periliminal, and subliminal. The latter two levels of flickers amplitude modulation depth were defined through a perceptual visibility threshold protocol on a single-subject basis. Subjective feedback indicated that the use of periliminal and subliminal flickers substantially improved user experience. The present study demonstrates that periliminal and subliminal flickers evoked SSVEP responses that can be used to derive spatial attention in frequency-tagging paradigms. The single-trial classification of attended space (left versus right) based on SSVEP response reached an average accuracy of 81.1% for the periliminal and 58% for the subliminal conditions. These findings reveal the promises held by the application of inconspicuous flickers to both cognitive neuroscience research and BCI development.HighlightsFrequency-tagging of spatial attention can be achieved through the presentation of flickering visual stimuli (flickers) whose contrast is reduced down to the individual’s perceptual visibility threshold revealing the potential of periliminal flickers as reliable frequency-tagging probes of spatial attentionBelow this perceptual visibility threshold, the signal-to-noise ratio of SSVEP responses was not sufficient to reliably distinguish the field upon which participants directed their attentionThe subliminal and periliminal flickers ameliorated the overall user experience and represent effective solutions to reduce bottom-up distraction, eye strain, and fatigue related to the presentation of flickering stimulationThe present findings have implications for the design of minimally intrusive frequency-tagging probes used within the frame of both fundamental cognitive neuroscience research and Brain Computer Interface

show abstract

Section: Discussionmentioning

confidence: 86%

Frequency-tagging of spatial attention using periliminal flickers

Ladouce,

Dehais

2024

Preprint

View full text Add to dashboard Cite

show abstract

“…Though it may be possible to evoke a stronger RIFT response from higher frequencies in future tasks as described above, currently we only see usable responses to 60Hz and 64Hz tagging. However, given that previous RIFT work has also made use of 56Hz (Brickwedde et al, 2022) without any resulting concerns over perceptibility of the flicker, and that different choices during the analysis stage (narrower bandpass filtering for coherence) may allow increased frequency resolution for tagging, we infer that RIFT-EEG is capable of uniquely tagging at least two individual stimuli/regions, with a strong likelihood of being able to tag at least three.…”

Section: Dependence On Tagged Frequencymentioning

confidence: 97%

“…While the existing framework has clearly delivered novel insights for cognitive neuroscience as described above, the low number of MEG setups worldwide combined with both the active cost of running MEG experiments as well as the initial capital required encourages the exploration of viable, more accessible alternatives. Additionally, as the potential of RIFT as a communication medium for Brain-Computer Interfaces begins to be explored (Brickwedde et al, 2022) as an improvement upon existing SSVEP Brain-Computer Interfaces (Zhu et al, 2010), the need to assess the possibility of a more portable RIFT framework rises further. Electroencephalography (EEG) has been shown to be sensitive to periodic stimulation in the RIFT frequency range using flickering LEDs (Gulbinaite et al, 2019;Herrmann, 2001).…”

Section: Introductionmentioning

confidence: 99%

“…At this stage, such early studies have already answered certain fundamental questions about the utility of RIFT across various applications. It has been shown that RIFT is a powerful tool to capture spatial shifts of attention both to tagged stimuli (Bouwkamp et al, 2023, Preprint), as well as to tagged, but perceptually indistinguishable regions of visual space (Brickwedde et al, 2022). Over the last few years RIFT has been applied within the visual domains of reading (Pan et al, 2021), distractor suppression (Ferrante et al, 2023), and visual search (Spaak et al, 2023, Preprint) among others, establishing that its robustness is comparable to that of the more conventional SSVEP.…”

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

Rapid Invisible Frequency Tagging (RIFT) in a novel setup with EEG

Arora,

Gayet,

Kenemans

et al. 2024

Preprint

View full text Add to dashboard Cite

Steady-State Visual Evoked Potentials (SSVEPs) provide a report-free and continuous measure of neural processing. Recent progress in display technology has allowed for the tagging of multiple stimuli simultaneously at >60Hz frequencies - high enough to evade perceptibility, while still evoking an oscillatory neural response. Known as Rapid Invisible Frequency Tagging (RIFT), this technique has currently only been used in combination with Magnetoencephalography (MEG), which is less accessible compared to Electroencephalography (EEG). Although responses to LEDs flickering at similar frequencies have been shown in EEG, it is currently unclear whether RIFT, using a more conventional stimulus display, can sufficiently evoke a response in EEG, and therefore whether it is worth adding the RIFT-EEG pairing to the cognitive neuroscientist's toolkit. Here, we successfully implement the first RIFT-EEG setup. We show that the oscillatory input is measurable in the EEG trace, what its topographical spread is, a rough range of applicable frequencies, and that this response is comparable to that evoked in MEG.

show abstract

“…To address this limitation, we developed the rapid invisible frequency tagging (RIFT) technique, which involves flickering visual stimuli at a frequency above 60 Hz, making it invisible and non-disruptive to the ongoing task. Responses to RIFT have been shown to increase with the allocation of attention to the stimulus bearing the visual flicker (Brickwedde et al, 2022; Drijvers et al, 2021; Duecker et al, 2021; Ferrante et al, 2023; Gutteling et al, 2022; Zhigalov et al, 2021, 2019; Zhigalov and Jensen, 2022, 2020). In our previous study, we adapted RIFT to a natural reading task and found temporally-precise evidence for parafoveal processing at the lexical level (Pan et al, 2021).…”

Section: Introductionmentioning

confidence: 99%

Early parafoveal semantic integration in natural reading

Pan

Frisson

Federmeier

et al. 2022

Preprint

Self Cite

View full text Add to dashboard Cite

While humans have a remarkable ability to read and comprehend text, the neuronal mechanisms supporting natural reading remain elusive. It is debated whether upcoming words in the parafovea are processed semantically. Here we show that the brain can indeed detect that a word is contextually implausible even before fixating on it. We demonstrated this by imperceivably flickering target words at 60 Hz that either were congruent or incongruent with the sentence context. While participants read these sentences, we used MEG to detect the flicker-response prior to saccades to the target words. We found a smaller tagging response for words that were incongruent with the sentence context; this reduction was related to the individual reading speed. We conclude that parafoveal semantic integration occurs during natural reading.

show abstract

Application of rapid invisible frequency tagging for brain computer interfaces

Cited by 14 publications

References 41 publications

Frequency-tagging of spatial attention using periliminal flickers

Frequency-tagging of spatial attention using periliminal flickers

Rapid Invisible Frequency Tagging (RIFT) in a novel setup with EEG

Early parafoveal semantic integration in natural reading

Contact Info

Product

Resources

About