Prestimulus alpha power predicts fidelity of sensory encoding in perceptual decision making
Introduction
Perceptual decision making is often described as the simplest form of a cognitive process, in that it involves transforming sensory evidence into a decision and behavioral response (Gold and Shadlen, 2007, Heekeren et al., 2004, Smith and Ratcliff, 2004). Substantial work has looked to identify and characterize the neural processes underlying perceptual decision making, with a focus on the neural correlates of processes that occur post-stimulus. For instance, experiments in non-human primates have shown neurons in the lateral intraparietal area (LIP) demonstrate activity indicative of evidence accumulation (Gold and Shadlen, 2007, Liu and Pleskac, 2011, Shadlen and Newsome, 2001). Analogous studies using neuroimaging in humans have focused on, among other areas, dorsal lateral prefrontal cortex (dLPFC) functioning as a comparator of decision alternatives (Heekeren et al., 2004, Heekeren et al., 2006, Ostwald et al., 2012, Philiastides et al., 2011).
Not all aspects of a perceptual decision are characterized by the post-stimulus activity. The state of the subject prior to stimulus presentation is also a factor in understanding how the perceptual decision evolves. Several groups have measured pre-stimulus oscillatory activity as a way to index the state of the subject prior to the presentation of the stimulus. Pre-stimulus oscillations in the α band (8–12 Hz) have been shown to correlate with visual discrimination performance (Babiloni et al., 2006, Hanslmayr et al., 2007, Hanslmayr et al., 2011, Thut et al., 2006, Van Dijk et al., 2008). Pre-stimulus α power is hypothesized to reflect top-down control of attention (Wyart and Tallon-Baudry, 2009) with increased pre-stimulus α power representing a low attentional state resulting in reduced decision accuracy. Recent studies have shown a correlation between pre-stimulus α power and subjective rating of attention toward a visual discrimination task (Macdonald et al., 2011). Pre-stimulus α phase has also shown to be predictive of visual awareness and perception (Busch et al., 2009, Mathewson et al., 2009).
Studies using electroencephalography (EEG) and magnetoencephalography (MEG) investigating pre-stimulus α within the context of perceptual decision making typically analyze data with respect to behavioral responses — e.g. segregating correct and error trials and characterizing the difference in the power spectrum or phase distributions (Busch et al., 2009, Van Dijk et al., 2008). Relatively little work has been done to investigate the variation of pre-stimulus α power when there is no difference in behavioral decision performance or when stimuli are nominally identical. It is possible that constituent neural processes are affected by pre-stimulus attentional state, though by the time the decision is made this relation is not observable in behavior or is confounded by other factors.
In this paper we investigate the relationship between pre-stimulus α power and post-stimulus discriminating components in a two alternative forced choice (2-AFC) decision making task. Unique to our approach is that we do not use behavioral data to separate trials for conducting our analyses, instead we investigate how pre-stimulus α power relates to post-stimulus neural components for cases in which the decisions are correct and the stimuli nominally identical.
Section snippets
Subjects, experimental design and data acquisition
Previous work by our group has used EEG to identify the timing of neural components which reflect the constituent processes of perceptual decision making (Philiastides and Sajda, 2006b, Philiastides and Sajda, 2007, Philiastides et al., 2006, Ratcliff et al., 2009, Sajda et al., 2009). In this previous work we identified a set of post-stimulus neural components that reflected, among other processes, an early perceptual component, correlating with sensory evidence, and a late discriminating
Results
We first analyzed the behavioral data to check whether the manipulation of phase coherence in the images significantly affected subjects' accuracy. All subjects were able to correctly identify more than 90% of images in the easiest trials (45% coherence) but performed at approximately chance for the most difficult trials (20% coherence). Results showed that phase coherence level was positively correlated with detection accuracy (p = 1.79 × 10− 11, t(46) = 8.84), and negatively correlated with reaction
Discussion
Uncovering the neural correlates of a perceptual decision is likely to lead to a better understanding of the neural processes underlying more complex decision making. In this paper we used a simple decision making task to investigate how pre-stimulus activity varies relative to post-stimulus activity which is discriminative of stimulus category. Specifically we found that pre-stimulus α power is reduced for trials with high magnitude (i.e. high absolute value) of the early discriminating
Acknowledgments
This work has been supported by NIH grants MH085092 and EB004730 and ARO grant W911NF-11-1-0219.
References (70)
- et al.
Paradox lost? Exploring the role of alpha oscillations during externally vs. internally directed attention and the implications for idling and inhibition hypotheses
Int. J. Psychophysiol.
(2003) - et al.
What causes nonmonotonic tuning of fMRI response to noisy images?
Curr. Biol.
(2002) - et al.
EEGLAB: an open source toolbox for analysis of single-trial EEG dynamics including independent component analysis
J. Neurosci. Methods
(2004) - et al.
Buildup of choice-predictive activity in human motor cortex during perceptual decision making
Curr. Biol.
(2009) - et al.
Alpha rhythm of the EEG modulates visual detection performance in humans
Cogn. Brain Res.
(2004) - et al.
Neural “ignition”: enhanced activation linked to perceptual awareness in human ventral stream visual cortex
Neuron
(2009) - et al.
Prestimulus oscillations predict visual perception performance between and within subjects
Neuroimage
(2007) - et al.
The role of alpha oscillations in temporal attention
Brain Res. Rev.
(2011) Categorical data analysis: away from ANOVAs (transformation or not) and towards logit mixed models
J. Mem. Lang.
(2008)- et al.
Visual masking and RSVP reveal neural competition
Trends Cogn. Sci.
(2002)
EEG alpha oscillations: the inhibition-timing hypothesis
Brain Res. Rev.
Temporal framing of thalamic relay-mode firing by phasic inhibition during the alpha rhythm
Neuron
New vistas for alpha-frequency band oscillations
Trends Neurosci.
Linear spatial integration for single-trial detection in encephalography
Neuroimage
Graphical display and statistical evaluation of event-related desynchronization (ERD)
Electroencephalogr. Clin. Neurophysiol.
Event-related EEG/MEG synchronization and desynchronization: basic principles
Clin. Neurophysiol.
Causal role of dorsolateral prefrontal cortex in human perceptual decision making
Curr. Biol.
Information processing algorithms in the brain
Trends Cogn. Sci.
Psychology and neurobiology of simple decisions
Trends Neurosci.
An objective method for measuring face detection thresholds using the sweep steady-state visual evoked response
J. Vis.
Prediction of response speed by anticipatory high-frequency (gamma band) oscillations in the human brain
Hum. Brain Mapp.
Pre- and poststimulus alpha rhythms are related to conscious visual perception: a high-resolution EEG study
Cereb. Cortex
Temporal characteristics of the influence of punishment on perceptual decision making in the human brain
J. Neurosci.
Responses of neurons in macaque MT to stochastic motion signals
Vis. Neurosci.
A relationship between behavioral choice and the visual responses of neurons in macaque MT
Vis. Neurosci.
The brain's default network: anatomy, function, and relevance to disease
Ann. N. Y. Acad. Sci.
The phase of ongoing EEG oscillations predicts visual perception
J. Neurosci.
The persistences of vision
Philos. Trans. R. Soc. Lond. B Biol. Sci.
Brain dynamics underlying the nonlinear threshold for access to consciousness
PLoS Biol.
Temporal characteristics of iconic memory
Nature
Dynamic predictions: oscillations and synchrony in top-down processing
Nat. Rev. Neurosci.
Parieto-occipital similar to 10 Hz activity reflects anticipatory state of visual attention mechanisms
Neuroreport
Modulation of oscillatory neuronal synchronization by selective visual attention
Science
The neural basis of decision making
Annu. Rev. Neurosci.
A general mechanism for perceptual decision-making in the human brain
Nature
Cited by (40)
Breathing in waves: Understanding respiratory-brain coupling as a gradient of predictive oscillations
2023, Neuroscience and Biobehavioral ReviewsEEG-based work experience prediction using hazard recognition
2022, Automation in ConstructionCitation Excerpt :This further indicates that the differences in brain activity before stimulation are related to individual differences among workers. This hypothesis is consistent with those of previous psychology studies suggesting that pre-stimulus EEG signals are related to the cognitive state [73] and decision making [74]. In this study, a machine learning algorithm was used to establish an experience prediction model based on the brain activity of workers performing a hazard recognition task.
Ongoing neural oscillations influence behavior and sensory representations by suppressing neuronal excitability
2022, NeuroImageCitation Excerpt :We found slower RTs following states of strong prestimulus alpha power estimated in the window before the task-relevant target stimulus (i.e., in syllable-locked epochs) across both periodic and aperiodic sites, including auditory and somatomotor regions. It is important to note that previous studies reported both positive (Bollimunta et al., 2008; Bompas et al., 2015; Kelly and O'Connell, 2013; Kirschfeld, 2008; Lou et al., 2014; Mazaheri et al., 2014; Min and Herrmann, 2007; Paoletti et al., 2019; van den Berg et al., 2016; Zhang et al., 2008), negative (Bollimunta et al., 2008; Del Percio et al., 2007; Zhang et al., 2008), and null relationships (Andino et al., 2005; van Dijk et al., 2008; Bays et al., 2015) between prestimulus alpha power and RTs. This mixed evidence may be due to whether power is estimated in a brain region processing task-relevant or -irrelevant information, as well as to task differences (e.g., whether accuracy or speed is emphasized), different laminar organization across regions (Bollimunta et al., 2008, 2011; Mo et al., 2011), or to long-range temporal dependencies in both alpha power and RT estimates, resulting in spurious positive and negative correlations (Schaworonkow et al., 2015).
Spatiotemporal characteristics and modulators of perceptual decision-making in the human brain
2017, Decision Neuroscience: An Integrative PerspectiveSpace-by-time decomposition for single-trial decoding of M/EEG activity
2016, NeuroImageCitation Excerpt :To illustrate the usefulness of the resulting space-by-time decomposition, we apply it to EEG activity recorded during a simple visual categorization task. As this EEG dataset has been studied thoroughly before (Lou et al., 2014; Philiastides et al., 2006a,b; Philiastides and Sajda, 2006; Ratcliff et al., 2009), the relevant EEG signatures are known and thus, it can be used as benchmark for testing and validating the proposed methodology. In the following, we first introduce the space-by-time decomposition and develop an algorithm based on NMF (Lee and Seung, 1999) that implements it.