P300 recordings during event-related fMRI: a feasibility study.

UMR 7004-Applications Biologiques et Médicales de la RMN et Génie Biologique et Médical, Université Louis Pasteur (ULP)/Centre National de Recherche Scientifique (CNRS)/IFR 37 de Neurosciences, Strasbourg Cedex-France.
Cognitive Brain Research (Impact Factor: 3.77). 06/2005; 23(2-3):306-15. DOI: 10.1016/j.cogbrainres.2004.10.017
Source: PubMed

ABSTRACT Analysis of combined event-related potentials (ERP) and functional magnetic resonance imaging (fMRI) can provide a high temporal and high spatial resolution to study functional cerebral processes. However, EEG data recorded inside an MR scanner is heavily distorted by artifacts. It is important in cognitive studies to ensure that recorded data reflect the same brain activity, and this is achieved through interleaved electroencephalographic (EEG) and fMRI measurements. Here, we demonstrate the feasibility of recording P300 ERPs during fMRI using a three-stimulus visual oddball task and involving a small number of trials for each stimulus. Ten EEG channels were acquired interleaved with fMRI images in five healthy subjects. The stimuli, including rare targets "X," frequent repetitive distractors "O," and rare distractors referred to as novels, were randomly presented every 2 +/- 1 s. The post hoc filter presented here was designed and applied to EEG data to remove the cardiac pulse artifact. Interleaved EEG/fMRI acquisition evidenced two P300 ERPs evoked at Fz, Cz, and Pz by targets and novels. Novel-related ERPs were of higher amplitude than their target-related counterparts. The fMRI maps acquired concurrently showed stronger BOLD response for target condition. We have shown that interleaved acquisition allows to obtain reliable P300 data and fMRI results, likely to shed light on the anatomical location of brain regions involved in cognitive ERPs relevant to many disorders affecting CNS functions. These noninvasive multimodal neuroimaging techniques can be used to explore and better understand processes underlying the functional brain organization.

  • [Show abstract] [Hide abstract]
    ABSTRACT: In on-axis lensless Fourier phase-shifting synthetic aperture digital holography, to compose all of the phase-shifting sub-holograms to a large synthetic aperture digital hologram effectively, firstly, the cross-correlation algorithm of the object waves is presented to correct the joint misplacement of the sub-holograms. Secondly, to make the phase-shifting synchronization matching of different sequence phase-shifting holograms, the cross-correlation algorithm of the phase-shifting holograms is employed. Compared with the traditional cross-correlation algorithm of the sub-holograms, the proposed approach makes the joint precision of the sub-holograms reach sub-pixel accuracy, and the resolution of the reconstructed image is improved significantly. In general, the proposed approach is effective in restraining the quality degradation of the synthetic reconstructed image that comes from the joint misplacement of the sub-holograms and the phase-shifting non-synchronization of the phase-shifting holograms.
    Optics Communications 05/2011; 284(9):2268-2272. · 1.44 Impact Factor
  • [Show abstract] [Hide abstract]
    ABSTRACT: In this study, techniques to measure EEG and fMRI simultaneously were investigated and P300 responses evoked by visual stimuli were examined. Event-related analysis was applied to combine the high temporal resolution of the EEG with the high spatial resolution of the fMRI, which may allow estimation of the temporal change in the activation of multiple cortical areas. A time scheme of stimulus presentation combined with an MRI scan was designed, taking into consideration the temporal delay between the generation of the P300 potential and the BOLD response. Three pattern oddball paradigm using standard, target and novel letter stimuli was performed, a technique in which subjects responded to rare target-letters but not to frequent standard- and rare novel-letters. Noises arising from the MRI scan and cardio-ballistic artifacts were removed from the raw data of the EEG by subtraction of the time-averaged waveforms of those artifacts. In comparing the grand average response of the EEG evoked by target events with those evoked by standard events, a significant difference was found in latency range from 280 to 450 ms (P
  • [Show abstract] [Hide abstract]
    ABSTRACT: While attempting to examine the physiological correlates of human cognitive functions, neuroscientists are restricted to noninvasive measures when dealing with healthy subjects. Correlates of cognitive brain processes are present in electromagnetic fields and haemodynamic responses that can be recorded with electroencephalography (EEG) and functional magnetic resonance imaging (fMRI), respectively. While EEG offers a temporal resolution on the millisecond timescale, intracranial sources of activity must be inferred from extracranial recordings—a phenomenon referred to as the inverse problem. fMRI offers spatial resolution on the millimetre scale but suffers from a suboptimal temporal resolution, since the blood oxygen level dependent (BOLD) signal is an indirect haemodynamic consequence of electrical brain activity. Combining EEG and fMRI is an approach that promises to integrate the good temporal resolution of EEG with the good spatial resolution of fMRI (for recent reviews, see Debener et al. 2006; Herrmann and Debener (2007); Menon and Crottaz-Herbette 2005). However, it should be noted that some authors have questioned the implicit assumption that both measures pick up more or less the same neural activity. A number of studies have demonstrated that EEG and BOLD responses do not reflect identical neural activity, resulting in the notion of EEG signals without fMRI correlates, and vice versa (Ritter and Villringer 2006).


Available from