Elsevier

NeuroImage

Volume 49, Issue 1, 1 January 2010, Pages 849-864
NeuroImage

EEG signatures of auditory activity correlate with simultaneously recorded fMRI responses in humans

https://doi.org/10.1016/j.neuroimage.2009.06.080Get rights and content

Abstract

We recorded auditory-evoked potentials (AEPs) during simultaneous, continuous fMRI and identified trial-to-trial correlations between the amplitude of electrophysiological responses, characterised in the time domain and the time–frequency domain, and the hemodynamic BOLD response. Cortical AEPs were recorded from 30 EEG channels within the 3 T MRI scanner with and without the collection of simultaneous BOLD fMRI. Focussing on the Cz (vertex) EEG response, single-trial AEP responses were measured from time-domain waveforms. Furthermore, a novel method was used to characterise the single-trial AEP response within three regions of interest in the time–frequency domain (TF-ROIs). The latency and amplitude values of the N1 and P2 AEP peaks during fMRI scanning were not significantly different from the Control session (p > 0.16). BOLD fMRI responses to the auditory stimulation were observed in bilateral secondary auditory cortices as well as in the right precentral and postcentral gyri, anterior cingulate cortex (ACC) and supplementary motor cortex (SMC). Significant single-trial correlations were observed with a voxel-wise analysis, between (1) the magnitude of the EEG TF-ROI1 (70–800 ms post-stimulus, 1–5 Hz) and the BOLD response in right primary (Heschl's gyrus) and secondary (STG, planum temporale) auditory cortex; and (2) the amplitude of the P2 peak and the BOLD response in left pre- and postcentral gyri, the ACC and SMC. No correlation was observed with single-trial N1 amplitude on a voxel-wise basis. An fMRI-ROI analysis of functionally-identified auditory responsive regions identified further single-trial correlations of BOLD and EEG responses. The TF amplitudes in TF-ROI1 and TF-ROI2 (20–400 ms post-stimulus, 5–15 Hz) were significantly correlated with the BOLD response in all bilateral auditory areas investigated (planum temporale, superior temporal gyrus and Heschl's gyrus). However the N1 and P2 peak amplitudes, occurring at similar latencies did not show a correlation in these regions. N1 and P2 peak amplitude did correlate with the BOLD response in bilateral precentral and postcentral gyri and the SMC. Additionally P2 and TF-ROI1 both correlated with the ACC. TF-ROI3 (400–900 ms post-stimulus, 4–10 Hz) correlations were only observed in the ACC and SMC. Across the group, the subject-mean N1 peak amplitude correlated with the BOLD response amplitude in the primary and secondary auditory cortices bilaterally, as well as the right precentral gyrus and SMC. We confirm that auditory-evoked EEG responses can be recorded during continuous and simultaneous fMRI. We have presented further evidence of an empirical single-trial coupling between the EEG and BOLD fMRI responses, and show that a time–frequency decomposition of EEG signals can yield additional BOLD fMRI correlates, predominantly in auditory cortices, beyond those found using the evoked response amplitude alone.

Introduction

EEG and fMRI are two neuroimaging techniques that are used to sample brain activity with complementary spatial and temporal resolution. Their simultaneous recording is especially desirable when variations in a subject's cognitive or physiological conditions mean that the brain activity under investigation cannot be guaranteed as consistent between separate experimental sessions. It is increasingly used to study epilepsy where the EEG recording is used to indicate when epileptic discharges occurred during the fMRI acquisition (Salek-Haddadi et al., 2002, Gotman, 2008), and has been successfully applied to investigate spontaneous brain rhythms (Goldmann et al., 2002, Laufs et al., 2003, Moosmann et al., 2003), and sleep (Schabus et al., 2007). Simultaneous EEG–fMRI is especially important for investigating the possible causes of variability in the brain activity measured by EEG and fMRI during time-dependent effects such as habituation, attention or learning; and in drug studies, where between-session variability may be large (Iannetti and Wise, 2007).

EEG–fMRI allows the relationship between the trial-to-trial variations of electrophysiological and fMRI responses to be investigated with high temporal and spatial resolution (Debener et al., 2007b). The integration of EEG and fMRI data may reveal information not easily extracted from either technique alone. Assuming a linear neurovascular coupling relationship between the hemodynamic response, the local-field potential and the scalp EEG, the “integration by prediction” approach models the fMRI signal as a function of the EEG convolved with a canonical hemodynamic response function. This approach has proven successful for the treatment of single-trial evoked responses to establish correlations with the BOLD response using the auditory oddball P300 (Eichele et al., 2005, Benar et al., 2007) and the error-related negativity (Debener et al., 2005) as well as interictal epileptic activity (Salek-Haddadi et al., 2002) and resting alpha rhythms (Goldmann et al., 2002, Laufs et al., 2003).

Several authors have reported the successful recording of truly continuous and simultaneous EEG and fMRI responses to visual and somatosensory stimuli (Becker et al., 2005, Iannetti et al., 2005). However, simultaneous EEG–fMRI recordings of auditory stimulation are challenging due to the MR-acoustic environment, in addition to the necessary subtraction of the MRI artifacts from the EEG signal. Several previous attempts at recording auditory-evoked potentials (AEPs) during fMRI scanning have used an interleaved experimental paradigm with quiet periods for stimulus delivery (Liebenthal et al., 2003, Scarff et al., 2004a, Mulert et al., 2005, Debener et al., 2005, Eichele et al., 2005, Debener et al., 2007a) to avoid the potential confound of the MR-acoustic noise interfering with the auditory stimulus-evoked brain activity. Interleaving comes with the additional benefit of allowing recording of the EEG signal without contamination by MRI gradient artifacts. No direct comparisons between AEPs during fMRI to a Control measure without scanning were made, though Mulert et al. (2005), reported significant sound-level dependent increases in the amplitude of AEPs recorded outside the scanner that were not observed inside the fMRI scanner. These studies reported a close correspondence between ERP equivalent dipole source localisations and LORETA current source distributions compared with the extent of fMRI activation localised in the temporal cortex (Scarff et al., 2004a, Mulert et al., 2005). Interleaving EEG and fMRI acquisition has important practical and theoretical limitations, including inefficient sampling of the neural activity and the consequent hemodynamic response, and a reduction in the flexibility of the stimulus presentation paradigm.

The sound of the MRI scanner can induce a BOLD response in the cortical areas responsible for auditory processing, but is largely restricted to primary auditory cortex (Bandettini et al., 1998, Talavage et al., 1999, Hall et al., 2000, Scarff et al., 2004b). Several studies have shown that MR-related acoustic noise may actually interfere with the activation of the auditory cortex measured with fMRI during auditory stimulation (Cho et al., 1998) and phonetic discrimination (Shah et al., 1999). It has also been shown that auditory activation elicited by modulated tones in a high noise condition can result in a smaller number of activated voxels than a low noise condition (Elliot et al., 1998). The high-intensity noise generated by EPI imaging sequences can reduce auditory-related brain activity through psychophysiological masking and habituation effects (Edmister et al., 1999). It can also have an effect via the higher level cognitive load of discriminating between two different sounds presented at the same time (Belin et al., 1999). The response to the background scanner noise will lead to an elevated baseline activation level (Di Salle et al., 2003, Langers et al., 2005) and to a reduction in the dynamic range of the responses that experimental stimuli will induce (Edmister et al., 1999). Under the constant acoustic noise conditions of an fMRI study, a partial saturation of neuronal-evoked regional hemodynamics may occur which can reduce the BOLD signal amplitude in response to auditory stimuli (Talavage and Edminster, 1998, Langers et al., 2005). The effect of acoustic MRI noise on the neuronal response to auditory stimulation itself has been less extensively investigated. Novitski et al. (2001) using recorded EPI noise at 54 dB, showed no significant difference in peak-to-peak amplitude of the N1 and P2 auditory-evoked potential (AEP) peaks evoked by 57 dB pure tones and chords.

As an alternative to the conventional measurement of event-related potential (ERP) peak amplitudes in the time-domain, time–frequency transforms provide a rich representation of ERP activity. In addition to the phase-locked activity that comprises the ERP waveform, a stimulus can induce event-related oscillations that are time-locked but not phase-locked to the stimulus and represent frequency specific changes in the ongoing EEG. These specific changes may consist of either increases or decreases in power in given frequency bands relative to a pre-stimulus period and are known as event-related synchronisation (ERS) and event-related desynchronisation (ERD) respectively. Studies using analysis in the frequency domain have revealed that EEG oscillations and ERP peaks in different frequency ranges are functionally related to information processing and behaviour (Pantev et al., 1994, Pfurtscheller et al., 1997, Basar et al., 1999).

Signal processing techniques such as the wavelet transform facilitate the decomposition of the evoked potential into its constituent oscillatory and non-stationary signal components, and allow the analysis of event-related data in both the time and frequency domains. Consequently a time–frequency analysis is perceived as having the potential to reveal neurophysiological information not available to analyses restricted to the time domain (Demiralp et al., 1999, Mouraux et al., 2003, Mouraux and Iannetti, 2008). Recently, wavelet transform techniques have been used to assess non-phase-locked activity in auditory- (Demiralp et al., 1999, Makinen et al., 2004), visual- (Makeig et al., 2002) and laser-evoked potentials (Mouraux et al., 2003) as well as the unpredictable properties of epileptiform discharges (Adeli et al., 2003) and the fluctuations in the power of alpha and gamma frequency bands (Yordanova et al., 2002, Moosmann et al., 2003, Busch et al., 2004).

In the present study we characterise and compare single-trial measures of AEP response amplitude in both the time- and the time–frequency domains as electrophysiological predictors of the BOLD response. We hypothesise that trial-to-trial variability in the AEP response is reflected in the BOLD time-series. We also investigate whether intersubject variability in the AEP amplitude is reflected in the BOLD signal.

Our investigation further evaluates whether robust and repeatable AEPs can be recorded during truly continuous, simultaneously acquired BOLD fMRI at 3 T. By comparing AEPs recorded during fMRI to recordings inside the scanner but without fMRI, we aim to assess whether AEP data quality is degraded by the combined effects of acoustic scanner noise and MRI gradient artifact removal.

Section snippets

Experimental paradigm

12 healthy volunteers (3 females, age range 23–32 years) participated in this study. All subjects gave their informed consent and the local ethics committee approved the procedures. For each subject, AEPs were recorded in two sessions conducted on the same day. In the ‘fMRI session’, AEPs were recorded during continuous fMRI at 3 T. In the ‘Control session’, AEPs were recorded within the MRI scanner but without fMRI acquisition. The order of the sessions was balanced across subjects.

The

Auditory EEG responses

All subjects reported after the experiment that they had heard the stimuli without difficulties. Auditory stimulation evoked clear and reproducible AEPs time-locked to the stimulus. Fig. 2 shows the group mean AEP waveforms, scalp maps of N1 and P2 peaks and group mean AEP TFM calculated from single-trial Cz-ref data recorded during both the fMRI and the Control sessions. The earliest identifiable response observed was the negative peak (N1). For N1, the amplitude and latency from the ERP grand

Discussion

We have successfully demonstrated the simultaneous acquisition of auditory-evoked potentials with continuous BOLD fMRI at 3 T. The primary negative and positive peaks (N1 and P2) of the vertex potential of the AEP were well preserved during fMRI compared to the Control recording sessions (without MRI gradient acoustic and electrical noise). While N1 appeared to be unaffected by fMRI, there was a trend towards a reduction in amplitude of the later P2 potential. We have used our 30 channel EEG

Acknowledgments

The authors would like to thank Dr. Andrew Bagshaw for very useful discussions during production of this manuscript. The authors would also like to acknowledge the support of The Engineering and Physical Sciences Research Council (EPSRC) (SM), The Royal Society (GDI) and The Medical Research Council (MRC) (SD, RN, RW).

References (81)

  • DebenerS. et al.

    Towards single-trial analysis in cognitive brain research

    Trends Cogn. Sci.

    (2007)
  • DemiralpT. et al.

    Time–frequency analysis of single-sweep event-related potentials by means of fast Fourier transform

    Brain Lang.

    (1999)
  • Di SalleF. et al.

    fMRI of the auditory system: understanding the basis of auditory gestalt

    Magn. Reson. Imaging.

    (2003)
  • GarreffaG. et al.

    Simultaneous EEG/fMRI acquisition, how far is it from being a standardised technique?

    Magn. Reson. Imaging

    (2004)
  • GiardM.H. et al.

    Dissociation of temporal and frontal components in the human auditory N1 wave, a scalp current density and dipole model analysis

    Electroencephalogr. Clin. Neurophysiol.

    (1994)
  • GiraudAL. et al.

    Endogenous cortical rhythms determine cerebral speech specialisation for speech perception and production

    Neuron

    (2007)
  • GodeyB. et al.

    Neuromagnetic source localization of auditory evoked fields and intracerebral evoked potentials, a comparison of data in the same patients

    Clin. Neurophysiol.

    (2001)
  • HartH.C. et al.

    The sound-level-dependent growth in the extent of fMRI activation in Heschl's gyrus is different for low- and high-frequency tones

    Hearing Res.

    (2003)
  • IannettiG.D. et al.

    BOLD functional MRI in disease and pharmacological studies, room for improvement?

    Mag. Reson. Med.

    (2007)
  • IannettiG.D. et al.

    Simultaneous recording of laser-evoked brain potentials and continuous, high-field functional magnetic resonance imaging in humans

    NeuroImage

    (2005)
  • JenkinsonM. et al.

    A global optimization method for robust affine registration of brain images

    Med. Image Anal.

    (2001)
  • JenkinsonM. et al.

    Improved optimisation for the robust and accurate linear registration and motion correction of brain images

    NeuroImage

    (2002)
  • JuckelG. et al.

    Intensity dependance of auditory evoked potentials in behaving cats

    Electroencephalogr. Clin. Neurophysiol.

    (1996)
  • KruggelF. et al.

    Hemodynamic and electroencephalographic responses to illusory figures, recording of the evoked potentials during functional MRI

    NeuroImage

    (2001)
  • LaufsH. et al.

    EEG-correlated fMRI of human alpha activity

    NeuroImage

    (2003)
  • LiebenthalE. et al.

    Simultaneous ERP and fMRI of the auditory cortex in a passive oddball paradigm

    NeuroImage

    (2003)
  • MayhewS.D. et al.

    Automated single-trial measurement of amplitude and latency of laser-evoked potentials (LEPs) by multiple-linear regression

    Clin. Neurophysiol.

    (2006)
  • MoosmannM. et al.

    Correlates of alpha rhythm in functional magnetic resonance imaging and near infrared spectroscopy

    NeuroImage

    (2003)
  • MourauxA. et al.

    Across-trial averaging of event-related responses and beyond

    Magn. Reson. Imag.

    (2008)
  • MourauxA. et al.

    Non-phase locked electroencephalogram (EEG) response to CO2 laser skin stimulations may reflect central interactions between Aδ and C-fibre afferent volleys

    Clin. Neurophysiol.

    (2003)
  • MulertC. et al.

    Integration of fMRI and simultaneous EEG: towards a comprehensive understanding of localization and time-course of brain activity in target detection

    NeuroImage

    (2004)
  • MulertC. et al.

    Sound level dependence of the primary auditory cortex, simultaneous measurement with 61-channel EEG and MRI

    NeuroImage

    (2005)
  • NagaiY. et al.

    Brain activity relating to the contingent negative variation, an fMRI investigation

    NeuroImage

    (2004)
  • NiazyR.K.N. et al.

    Removal of fMRI-environment artefacts using optimal basis sets

    NeuroImage

    (2005)
  • NovitskiN. et al.

    Effects of acoustic gradient noise from functional magnetic resonance imaging on auditory processing as reflected by event-related brain potentials

    NeuroImage

    (2001)
  • PfurtschellerG. et al.

    Foot and hand area mu rhythms

    Int. J. Psychophysiology

    (1997)
  • PictonT.W. et al.

    Human auditory evoked potentials. I: Evaluation of components

    Electroenceph. Clin. Neurophysiol.

    (1974)
  • RitterP. et al.

    Evaluating gradient artifact correction of EEG data acquired simultaneously with fMRI

    Mag. Reson. Imaging

    (2007)
  • Salek-HaddadiA. et al.

    Simultaneous EEG-correlated ictal fMRI

    NeuroImage

    (2002)
  • ScarffC.J. et al.

    Simultaneous 3 T fMRI and high-density recording of human auditory evoked potentials

    NeuroImage

    (2004)
  • Cited by (67)

    • The spatiotemporal pattern of pure tone processing: A single-trial EEG-fMRI study

      2019, NeuroImage
      Citation Excerpt :

      EEG is a selective measure of current source activity which raises energy metabolism, whereas the haemodynamic fMRI signal is related to energy consumption of neural populations. Assuming a linear neurovascular coupling relationship (for healthy young adults) between the hemodynamic response, the LFP and the scalp EEG, the “integration by prediction” approach models the fMRI signal as a function of the EEG convolved with a canonical hemodynamic response function (Mayhew et al., 2010). This approach has proven successful for the treatment of single-trial evoked responses to establish correlations with the BOLD response using the auditory oddball P300 (Eichele et al., 2005) and task relevant ERP components (Walz et al., 2014).

    View all citing articles on Scopus
    View full text