Statistical parametric mapping for event-related potentials: I. Generic considerations

doi:10.1016/j.neuroimage.2004.02.012

NeuroImage

Volume 22, Issue 2, June 2004, Pages 492-502

https://doi.org/10.1016/j.neuroimage.2004.02.012 Get rights and content

Abstract

In this paper, we frame the strategy and motivations behind developments in statistical parametric mapping (SPM) for the analysis of electroencephalogram (EEG) data. This work deals specifically with SPM procedures for the analysis of event-related potentials (ERP). We place these developments in the larger context of integrating electrophysiological and hemodynamic measurements of evoked brain responses through the fusion of EEG and fMRI data. In this paper, we consider some fundamental issues when selecting an appropriate statistical model that enables diverse questions to be asked of the data and at the same time retains maximum sensitivity. The three key issues addressed in this paper are as follows: (i) should multivariate or mass univariate analyses be adopted, (ii) should time be treated as an experimental factor or as a dimension of the measured response variable, and (iii) how to form appropriate explanatory variables in a hierarchical observation model. We review the relative merits of the different options and explain the rationale for our choices. In brief, we motivate a mass univariate approach in terms of sensitivity to region-specific responses. This involves modeling responses at each voxel or space bin separately. In contradistinction, we treat time as an experimental factor to enable inferences about temporally distributed responses that encompass multiple time bins.

In a companion paper, we develop statistical models of ERPs in the time domain that follow from the heuristics established here and illustrate the approach using simulated and real data.

Introduction

There is a clear consensus that the most promising applications of neuroimaging rest upon the integration of different modalities. It has been shown recently that multimodal data acquisition and fusion are useful for gaining additional insight into the neuronal causes of observed hemodynamic and electrophysiological brain responses Czisch et al., 2002, Goldman et al., 2002, Lemieux et al., 2001, Salek-Haddadi et al., 2002, Salek-Haddadi et al., 2003, Trujillo-Barreto et al., 2001.

In this paper, we work towards one aspect of a particular combination of modalities, electroencephalography (EEG), and functional magnetic resonance imaging (fMRI). An integration of these modalities is promising because of the complementary spatiotemporal resolution of fMRI and EEG. Furthermore, both techniques are the most accessible modalities in research and clinics. An integration of EEG and fMRI is not only potentially useful from a theoretical point of view, but also in practice. One important component of any integrative initiative is the ability to model both types of data in the same mathematical framework to make inferences that are mutually informed. This first component can be seen as a prelude to a full data fusion based on an integrated model for fMRI and EEG.

A candidate for a such a framework is statistical parametric mapping (SPM) (Friston, 2004), which is a mass univariate approach to modeling spatiotemporal neuroimaging data. SPM was originally developed to deal with metabolic or hemodynamic imaging time series, that is, PET, SPECT, and fMRI data. A similar spatiotemporal model can be derived for EEG data. Other groups have already illustrated the usefulness of SPM techniques through applications of SPM to EEG data. For example, Bosch-Bayard et al. (2001) have described an SPM approach to source reconstructed Fourier transformed EEG data. Park et al. (2002) have implemented a procedure that produces statistical parametric maps with source reconstructed EEG data. Barnes and Hillebrand (2003) have applied SPM to source reconstructed MEG data. These developments demonstrate the applicability of SPM to many kinds of neuroimaging data.

This paper deals specifically with the characterization of event-related potentials (ERPs) as measured with the EEG using the same SPM concepts developed for metabolic imaging Friston, 2004, Friston et al., 2002b. The extension of SPM procedures, to cover ERPs, entails a number of critical choices, which are the subject of this paper. In a companion paper, we describe a temporal model for averaged ERPs, which can be used to test hypotheses about localized effects in peristimulus time or in the peristimulus time or frequency domain. These hypotheses can be tested using the same model. Inference is made in a classical sense based on the t or F statistic. A future communication will extend the model to deal with spatiotemporal data based on the principles described below.

This paper is structured as follows. We will first review, briefly, multimodality integration and outline the overall strategy that we are pursuing. The second section focuses on the different observation and statistical models that could be used for the analysis of ERP data in the light of two key distinctions. These distinctions are between multivariate and mass univariate analyses over space and between treating time as a continuous dimension of the response variable versus a discrete replication factor. The implications of these different approaches for estimation and inference will be described and the motivation for the choices we have made is presented. In the third and final section, we describe the general analysis procedures that ensue. These procedures are based on a hierarchical linear observation model. In a companion paper, the specific operational details and model assumptions for ERP data are presented. These are then applied to synthetic and real data to establish their construct validity in relation to established approaches.

Section snippets

Integration of fMRI and EEG data

Over the past years, there has been an enormous interest in the fusion or integration of electrophysiological and hemodynamical measurements of evoked neuronal responses. The integration of these data (usually fMRI and EEG or MEG) can be classified into three sorts:

•
integration through temporal prediction;
•
integration through spatial constraints; and
•
integration through fusion.

The simplest approach, integration through temporal prediction, is to use one modality to predict the other. A significant

Statistical models for source reconstructed EEG time series

There are some fundamentally different alternatives that present themselves when choosing an appropriate statistical model for the analysis of EEG time series. We restrict ourselves to the analysis of (averaged) ERP data (Fig. 1). By ERP data, we mean averaged event-related time courses (Rugg and Coles, 1995), where each of these time courses has been averaged within subject and trial type (condition) to provide one peristimulus time series for each trial type and each subject.

Reconstruction

The temporal dimension

Having established the utility of an SPM-like approach to the analysis of each voxel time series, we now have to consider whether time is a fourth dimension of the response variable or a discrete series of observations over time bins. The RFT correction has been generalized to any arbitrary number of dimensions by Worsley et al. (1996). Many interesting applications of high-dimensional SPMs exist, for example, augmenting space with scale–space dimensions (Siegmund and Worsley, 1995) or even

Conclusion

In this paper, we have set out the choices guiding the development of analytic procedures for ERP data using the statistical parametric mapping framework. We have focussed on motivation and justification, particularly in relation to the different sorts of statistical models and analyses that could have been used. Guided largely by the sorts of questions that are asked of the data, we conclude that a mass univariate approach is appropriate. In the temporal domain, the linear two-level

Acknowledgements

The Wellcome Trust funded this work. We would like to thank Marcia Bennett for help in preparing the manuscript and Rik Henson for valuable discussions.

References (45)

E. Basar et al.
Oscillatory brain dynamics, wavelet analysis, and cognition
Brain Lang.
(1999)
M. Czisch et al.
Altered processing of acoustic stimuli during sleep: reduced auditory activation and visual deactivation detected by a combined fMRI/EEG study
NeuroImage
(2002)
A. Dale et al.
Dynamic statistical parametric mapping: combining fMRI and MEG for high-resolution imaging of cortical activity
Neuron
(2000)
O. David et al.
A neural mass model for MEG/EEG: coupling and neuronal dynamics
NeuroImage
(2003)
K.J. Friston et al.
Classical and Bayesian inference in neuroimaging: applications
NeuroImage
(2002)
K.J. Friston et al.
Classical and Bayesian inference in neuroimaging: theory
NeuroImage
(2002)
S. Kiebel et al.
Anatomically informed basis functions
NeuroImage
(2000)
L. Lemieux et al.
Event-related fMRI with simultaneous and continuous EEG: description of the method and initial case report
NeuroImage
(2001)
W. Penny et al.
Variational Bayes inference for fMRI time series
NeuroImage
(2003)
C. Phillips et al.
Anatomically informed basis functions for EEG source localization: combining functional and anatomical constraints
NeuroImage
(2002)

C. Phillips et al.

Systematic regularization of linear inverse solutions of the EEG source localization problem

NeuroImage

(2002)

A. Salek-Haddadi et al.

Simultaneous EEG correlated ictal fMRI

NeuroImage

(2002)

K. Toma et al.

Generators of movement-related cortical potentials: fMRI constrained EEG dipole source analysis

NeuroImage

(2002)

L.J. Trejo et al.

Feature extraction of event-related potentials using wavelets: an application to human performance monitoring

Brain Lang.

(1999)

M.W. Woolrich et al.

Temporal autocorrelation in univariate linear modeling of fFMRI data

NeuroImage

(2001)

K.J. Worsley et al.

Analysis of fMRI time-series revisited—Again

NeuroImage

(1995)

K.J. Worsley et al.

Characterizing the response of PET and fMRI data using multivariate linear models

NeuroImage

(1997)

F. Abramovich et al.

Wavelet analysis and its statistical applications

Statistician

(2000)

S. Baillet et al.

A Bayesian approach to introducing anatomo-functional priors in the EEG/MEG inverse problem

IEEE Trans. Biomed. Eng.

(1997)

G. Barnes et al.

Statistical flattening of MEG beamformer images

Hum. Brain Mapp.

(2003)

J. Bosch-Bayard et al.

3D statistical parametric mapping of EEG source spectra by means of variable resolution electromagnetic tomography (VARETA)

Clin. EEG Electroencephalogr.

(2001)

M. Breakspear

Nonlinear phase desynchronization in human electroencephalograhic data

Hum. Brain Mapp.

(2002)

Cited by (102)

Aberrant Hierarchical Prediction Errors Are Associated With Transition to Psychosis: A Computational Single-Trial Analysis of the Mismatch Negativity
2023, Biological Psychiatry: Cognitive Neuroscience and Neuroimaging
Mismatch negativity reductions are among the most reliable biomarkers for schizophrenia and have been associated with increased risk for conversion to psychosis in individuals who are at clinical high risk for psychosis (CHR-P). Here, we adopted a computational approach to develop a mechanistic model of mismatch negativity reductions in CHR-P individuals and patients early in the course of schizophrenia.
Electroencephalography was recorded in 38 CHR-P individuals (15 converters), 19 patients early in the course of schizophrenia (≤5 years), and 44 healthy control participants during three different auditory oddball mismatch negativity paradigms including 10% duration, frequency, or double deviants, respectively. We modeled sensory learning with the hierarchical Gaussian filter and extracted precision-weighted prediction error trajectories from the model to assess how the expression of hierarchical prediction errors modulated electroencephalography amplitudes over sensor space and time.
Both low-level sensory and high-level volatility precision-weighted prediction errors were altered in CHR-P individuals and patients early in the course of schizophrenia compared with healthy control participants. Moreover, low-level precision-weighted prediction errors were significantly different in CHR-P individuals who later converted to psychosis compared with nonconverters.
Our results implicate altered processing of hierarchical prediction errors as a computational mechanism in early psychosis consistent with predictive coding accounts of psychosis. This computational model seems to capture pathophysiological mechanisms that are relevant to early psychosis and the risk for future psychosis in CHR-P individuals and may serve as predictive biomarkers and mechanistic targets for the development of novel treatments.
Impact of face outline, parafoveal feature number and feature type on early face perception in a gaze-contingent paradigm: A mass-univariate re-analysis of ERP data
2022, Neuroimage: Reports
Recent ERP research using a gaze-contingent paradigm suggests the face-sensitive N170 component is modulated by the presence of a face outline, the number of parafoveal facial features, and the type of feature in parafovea (Parkington and Itier, 2019). The present study re-analyzed these data using robust mass univariate statistics available through the LIMO toolbox, allowing the examination of the ERP signal across all electrodes and time points. We replicated the finding that the presence of a face outline significantly reduced ERP latencies and amplitudes, suggesting it is an important cue to the prototypical face template. However, we found that this effect began around 114 ms, and was maximal during the P1-N170 and N170-P2 intervals. The number of features present in parafovea also impacted the entire waveform, with systematic reductions in amplitude and latency as the number of features increased. This effect was maximal around 120 ms during the P1-N170 interval and around 170 ms between the N170 and P2. The ERP response was also modulated by feature type; contrary to previous findings this effect was maximal around 200 ms and the P2 peak. Although we provide partial replication of the previous results on the N170, the effects were more temporally distributed in the present analysis. These effects were generally maximal before and after the N170 and were the weakest at the N170 peak itself. This re-analysis demonstrates that classical ERP analysis can obscure important aspects of face processing beyond the N170 peak, and that tools like mass univariate statistics are needed to shed light on the whole time-course of face processing.
Reading direct speech quotes increases theta phase-locking: Evidence for cortical tracking of inner speech?
2021, NeuroImage
Citation Excerpt :
Topographic images were stacked to create a 3D space-time image volume. The volume was smoothed with a Gaussian kernel at FWHM=[16 mm 16 mm 16 ms], about 3 times of voxel size, which is in accordance with the assumptions of Random Field Theory (Kiebel and Friston, 2004; Worsley et al., 1996). To verify that the observed results reflect theta-specific rather than broad-band phase-locked responses, a full time-frequency analysis was conducted.
Growing evidence shows that theta-band (4–7 Hz) activity in the auditory cortex phase-locks to rhythms of overt speech. Does theta activity also encode the rhythmic dynamics of inner speech? Previous research established that silent reading of direct speech quotes (e.g., Mary said: “This dress is lovely!”) elicits more vivid inner speech than indirect speech quotes (e.g., Mary said that the dress was lovely). As we cannot directly track the phase alignment between theta activity and inner speech over time, we used EEG to measure the brain's phase-locked responses to the onset of speech quote reading. We found that direct (vs. indirect) quote reading was associated with increased theta phase synchrony over trials at 250–500 ms post-reading onset, with sources of the evoked activity estimated in the speech processing network. An eye-tracking control experiment confirmed that increased theta phase synchrony in direct quote reading was not driven by eye movement patterns, and more likely reflects synchronous phase resetting at the onset of inner speech. These findings suggest a functional role of theta phase modulation in reading-induced inner speech.
Probability waves: Adaptive cluster-based correction by convolution of p-value series from mass univariate analysis
2021, Journal of Neuroscience Methods
Methods for p-value correction are criticized for either increasing Type II error or improperly reducing Type I error in large exploratory data analysis. This text considers patterns in probability vectors resulting from mass univariate analysis to correct p-values, where clusters of significant p-values may indicate true H0 rejection.
We used ERP experimental data from control and ADHD boys to test the method. The Log10 of p-vector was convolved with a Gaussian window whose length was set as the shortest lag above which autocorrelation of each ERP wave may be assumed to have vanished. We realized Monte-Carlo simulations (MC) to (1) evaluate confidence intervals of rejected and non-rejected areas of our data, (2) to evaluate differences between corrected and uncorrected p-vectors or simulated ones in terms of distribution of significant p-values, and (3) to empirically verify the type-I error rate (comparing 10,000 pairs of mixed samples whit control and ADHD subjects).
The differences between simulation or raw p-vector and corrected p-vectors were, respectively, minimal and maximal for window length set by autocorrelation in p-vector convolution.
Our method was less conservative while FDR methods rejected basically all significant p-values.The MC simulations presented 2.78 ± 4.83% of difference (20 channels) from corrected p-vector, while difference from raw p-vector was 596 ± 5.00% (p = 0.0003).
As a cluster-based correction, the present new method seems to be biological and statistically suitable to correct p-values in mass univariate analysis of ERP waves, which adopts adaptive parameters to correction.
Using multilevel models for the analysis of event-related potentials
2021, International Journal of Psychophysiology
Multilevel modeling (MLM) is becoming increasingly accessible and popular in the analysis of event-related potentials (ERPs). In this article, we review the benefits of MLM for analyzing psychophysiological data, which often contains repeated observations within participants, and introduce some of the decision-making points in the analytic process, including how to set up the data set, specify the model, conduct hypothesis tests, and visualize the model estimates. We highlight how the use of MLM can extend the types of theoretical questions that can be answered using ERPs, including investigations of how ERPs vary meaningfully across trials within a testing session. We also address reporting practices and provide tools to calculate effect sizes and simulate power curves. Ultimately, we hope this review contributes to emerging best practices for the use of MLM with psychophysiological data.
State anxiety biases estimates of uncertainty and impairs reward learning in volatile environments
2021, NeuroImage
Clinical and subclinical (trait) anxiety impairs decision making and interferes with learning. Less understood are the effects of temporary anxious states on learning and decision making in healthy populations, and whether these can serve as a model for clinical anxiety. Here we test whether anxious states in healthy individuals elicit a pattern of aberrant behavioural, neural, and physiological responses comparable with those found in anxiety disorders—particularly when processing uncertainty in unstable environments. In our study, both a state anxious and a control group learned probabilistic stimulus-outcome mappings in a volatile task environment while we recorded their electrophysiological (EEG) signals. By using a hierarchical Bayesian model of inference and learning, we assessed the effect of state anxiety on Bayesian belief updating with a focus on uncertainty estimates. State anxiety was associated with an underestimation of environmental uncertainty, and informational uncertainty about the reward tendency. Anxious individuals’ beliefs about reward contingencies were more precise (had smaller uncertainty) and thus more resistant to updating, ultimately leading to impaired reward-based learning. State anxiety was also associated with greater uncertainty about volatility. We interpret this pattern as evidence that state anxious individuals are less tolerant to informational uncertainty about the contingencies governing their environment and more willing to be uncertain about the level of stability of the world itself. Further, we tracked the neural representation of belief update signals in the trial-by-trial EEG amplitudes. In control participants, lower-level precision-weighted prediction errors (pwPEs) about reward tendencies were represented in the ERP signals across central and parietal electrodes peaking at 496 ms, overlapping with the late P300 in classical ERP analysis. The state anxiety group did not exhibit a significant representation of low-level pwPEs, and there were no significant differences between the groups. Smaller variance in low-level pwPE about reward tendencies in state anxiety could partially account for the null results. Expanding previous computational work on trait anxiety, our findings establish that temporary anxious states in healthy individuals impair reward-based learning in volatile environments, primarily through changes in uncertainty estimates, which play a central role in current Bayesian accounts of perceptual inference and learning.

View all citing articles on Scopus

View full text

CommentaryStatistical parametric mapping for event-related potentials: I. Generic considerations

Abstract

Introduction

Section snippets

Integration of fMRI and EEG data

Statistical models for source reconstructed EEG time series

The temporal dimension

Conclusion

Acknowledgements

Brain Lang.

NeuroImage

Neuron

NeuroImage

NeuroImage

NeuroImage

NeuroImage

NeuroImage

NeuroImage

NeuroImage

NeuroImage

NeuroImage

NeuroImage

Brain Lang.

NeuroImage

NeuroImage

NeuroImage

Wavelet analysis and its statistical applications

Statistician

A Bayesian approach to introducing anatomo-functional priors in the EEG/MEG inverse problem

IEEE Trans. Biomed. Eng.

Statistical flattening of MEG beamformer images

Hum. Brain Mapp.

3D statistical parametric mapping of EEG source spectra by means of variable resolution electromagnetic tomography (VARETA)

Clin. EEG Electroencephalogr.

Nonlinear phase desynchronization in human electroencephalograhic data

Hum. Brain Mapp.

Commentary
Statistical parametric mapping for event-related potentials: I. Generic considerations