< Back to previous page

Project

Time and Numbers, or the when and where of attentional processing: A psychophysiological approach using the SSVEP

Daily activities require our brain to selectively attend to aspects of the environment that are most relevant to guide and monitor our behavior and to avoid undesirable outcomes of our and other’s actions. Because prioritizing and processing relevant information may be highly demanding, and thus difficult to maintain at all times, our brain needs to properly select and filter different features (e.g., based on location, physical features such as color or shape, or time of occurrence) and adjust our cognitive control system, among others, so that our behavioral performance corresponds to the requirements of each particular situation.

To understand how our brain selects and processes information to generate a proper response is one of the goals of cognitive neuroscience. To this end, several techniques have been implemented for studying cognitive processes, such as attentional processing, cognitive control and temporal predictability, among which we note behavioral testing and noninvasive electrophysiological recording. Because cognitive processing can be deployed and can change rapidly (within a fraction of a second), brain-imaging techniques with high temporal resolution are particularly useful for studying the aforementioned cognitive processes. Electroencephalography (EEG), with a temporal resolution in the order of milliseconds, is one of the most widely used brain-imaging tools for studying human cognition among which the steady-state visual evoked potential (SSVEP) is a commonly used technique. 

The SSVEP technique involves repeatedly flickering visual stimuli while measuring their EEG response on the scalp. As the visual stimulus flickers (like a strobe light), the activity of large neural populations in the brain’s early visual processing areas will resonate at the same temporal frequency, and leads to a sinusoidal EEG response called SSVEP (Regan, 1989). The SSVEP has been extensively used to study visual attention because its amplitude correlates with the amount of attention paid to the flickering stimulus.  Although changes in SSVEP amplitude are generally related to the effects of attention, Ding et al. (2006) suggested that the attention modulation may depend on stimulation frequency and that for frequencies in the lower alpha band (8 – 10 Hz) an increase in SSVEP amplitude can be found when the flickering frequency is peripheral to the attended target.

In the present thesis, I report on the SSVEP to study temporal attention and the relation between spatial attention and mental representation of numbers.

First, because understanding the cognitive and biological aspects of cognition in healthy as well as in clinical populations would extend our knowledge about such processes, and in order to aid in development of solutions that could benefit clinical populations, I presented in chapter 2 a comprehensive review of Brain-Computer Interfaces (BCIs). BCI systems enable a user to communicate with the external world by directly translating his/her brain activity into (computer) commands without relying on the brain’s normal output pathways. The review is presented in the context of language model applications that supplement non-invasive BCI-based communication systems. The review offers an overview of the advantages and challenges when implementing language models combined with different techniques, including the use of SSVEP, in BCI-based communication systems when implemented in conjunction with other Ambient Assisted Living (AAL) technologies.

To the best of my knowledge, SSVEP has not been used to study temporal attention. This is important because temporal predictability has been shown to improve attention and sensory processing across different domains (Woodrow, 1914).  The relevance of SSVEP to study temporal attention is important to demonstrate for fundamental scientific reasons but also for future diagnostic ones since temporal attention and the correct use of temporal cues to guide behavior are impaired in several clinical disorders including Parkinson's disease (Rochester et al., 2005) and Schizophrenia (Silberstein et al., 2000). In the research lab, temporal attention is often studied as attention to the moment a particular stimulus appears or disappears from the screen (Babiloni et al., 2004). Several ERPs (e.g.,  Miniussi, Wilding, Coull, & Nobre, 1999) and oscillatory studies (e.g., Klimesch, 2012) have been implemented to study temporal attention. Neural oscillations in the alpha band (~10 Hz) seem to be part of the substrate of temporal attention.  One of the theories about temporal attention proposes the existence of internal oscillators whose attentional pulses can be entrained to environmental rhythmicity to enhance stimuli processing at predictable times (Mathewson, Fabiani, Monica, et al., 2010), where temporal expectancies are phase locked to ongoing brain oscillations to improve target recognition when time intervals are predictable (Schroeder & Lakatos, 2010).

We designed a first experiment to evaluate whether SSVEP stimulation and its temporal dynamics can be used to study temporal attention, and to evaluate whether attention-modulation of SSVEP is similar for three different frequencies (Chapter 3). Because temporal expectancies allow maximal target recognition at predictable times to have an effect on the participant’s response, we manipulated the Inter stimulus interval (ISI; constant versus variable) to evaluate whether there is an effect of SSVEP stimulation on behavioral performance. The obtained behavioral results replicated previous studies confirming the benefits of temporal expectations on performance for trials with constant ISI. EEG analyses revealed robust SSVEP amplitudes for all flicker frequencies, although a main effect of temporal expectations on SSVEP amplitude was not significant. Additional analyses revealed temporal predictability-related modulations of SSVEP amplitude at 10 Hz and its second harmonic (20 Hz), and for the 6 Hz flicker but not for 15 Hz, for any ISI condition. These results provide some evidence for the feasibility of the SSVEP technique to study temporal attention for stimuli with flicker frequencies around the alpha band.

Another important component of attention is the ability to direct attention in the visual field to a particular stimulus while suppressing the processing of competing stimuli in other locations, called spatial attention. The representation of numbers in the brain is spatially organized on what is known as the mental number line. The Spatial-Numerical Association of Responses Codes (SNARC) effect describes the relationship between the spatial representation of numbers in the brain and the number magnitude (S Dehaene et al., 1993). It has been used to explain a participant’s reaction time in a parity judgment task. The magnitude of the number in the mental number line leads to faster responses when the digit’s position is congruent with the response hand. This interaction between mental representation of numbers and enhancement of behavior has been described in tasks where the number magnitude is relevant for the task (magnitude-judgment) but also when the magnitude information is not task-relevant (parity judgment; Fias & Fischer, 2004; Fias, Lauwereyns, & Lammertyn, 2001).

The studies on spatial attention and mental representation of numbers reported in this thesis involved modified versions of the SNARC task. The purpose was to use the SSVEP technique to provide direct evidence (or lack thereof) for the hypothesis that SNARC induces spatial shifts of attention. This is a novel approach for studying number processing and its effect on the allocation of spatial attention. Previous EEG studies on spatial attention during covert attention shifts provide evidence of changes in SSVEP amplitude at posterior and contralateral locations for target detection compared to non-attended flickering stimuli (M. M. Müller & Hillyard, 2000). Because attentional shifts can be achieved without moving the eyes −covert attention in which participants keep their gaze at the fixation point while the stimulus is displayed either to the left or to the right hemifield of the fixation point−, we implemented a modified Fischer’s et al. (2003) task. In a covert spatial attention task participants are asked to fixate the center of the screen while detecting a target displayed either to the left or the right side of the fixation point. For each trial, but before target presentation, one out of four digits were displayed and participants were informed that the numbers were not relevant to the task.

The results of the parity judgment task (classify numbers as odd or even while ignoring their magnitudes) described in chapter 4 replicated the expected behavioral patterns predicted by the SNARC effect. We also observed significant changes in the SSVEP amplitude with respect to the baseline for the left and right flickering stimuli, for both the congruent and incongruent conditions. Statistically significant differences between the congruent and incongruent conditions were larger for the congruent conditions for the flickering stimulus on the left. Taken together, these findings support the hypothesis that, when numbers are part of the task but their magnitude is not relevant, the SNARC-spatial attention effect is elicited as a cognitive effect resulting from the mental representation of numbers and its relation with the space representation, more so than merely a motor effect. Results of the target detection task (detection of peripheral presented targets while ignoring numbers that were not relevant for the task) in chapter 5 showed benefits for the interaction between large magnitude numbers preceding targets presented in the right hemifield, providing a partial replication of the SNARC behavioral effect. And EEG SSVEP results showed an attention modulation effect for the difference between congruent and incongruent conditions, showing larger SSVEP amplitudes for incongruent conditions (small numbers/targets on the right) for the flickering stimuli in the left hemifield and for the congruent conditions (large numbers/target on the right) for the flickering stimuli in the right hemifield. Our results demonstrate that passively observing numbers does not induce an automatic shift of spatial attention, or at least not an effect that could be measured electrophysiologically. They also demonstrate that even when numbers are irrelevant for the task, number semantic representation is activated.

Together, the results presented in this thesis provide, on the one hand, supporting evidence for language models within BCI systems, as a significant improvement in spelling performance can be achieved, opening new avenues for BCI integration in the AAL community and for rehabilitation. On the other hand, the results reported here contribute to a better understanding of the psychophysiological components of temporal and spatial attentional processing, the relation between flickering frequencies and endogenous brain oscillations, and the relation of these endogenous brain oscillations with information processing, cognitive control and behavioral performance. Additionally, for all studies presented in this thesis, I provide important evidence for a future implementation of the SSVEP technique to study temporal attention and the role of number inducing automatic shifts of spatial attention.

Overall, the studies presented here further our understanding not only of the interaction between number magnitude, the mental representation of space and the physiology behind the SNARC effect, but also how changes in temporal and spatial attentional processing affect SSVEP amplitude, and whether SSVEP flickering stimuli have an effect on endogenous brain oscillations affecting attentional processing. 

Date:1 Oct 2009 →  29 May 2018
Keywords:Attention, Psychophysiology, Steady State Visual-Evoked Potentials (SSVEP), SNARC effect, Temporal Attention, Response Conflict
Disciplines:Biological and physiological psychology, General psychology, Other psychology and cognitive sciences, Neurosciences, Cognitive science and intelligent systems, Developmental psychology and ageing
Project type:PhD project