Elsevier

NeuroImage

Volume 18, Issue 3, March 2003, Pages 595-609
NeuroImage

Regular article
Consistent and precise localization of brain activity in human primary visual cortex by MEG and fMRI

https://doi.org/10.1016/S1053-8119(02)00053-8Get rights and content

Abstract

The tomographic localization of activity within human primary visual cortex (striate cortex or V1) was examined using whole-head magnetoencephalography (MEG) and 4-T functional magnetic resonance imaging (fMRI) in four subjects. Circular checkerboard pattern stimuli with radii from 1.8 to 5.2° were presented at eccentricity of 8° and angular position of 45° in the lower quadrant of the visual field to excite the dorsal part of V1 which is distant from the V1/V2 border and from the fundus of the calcarine sulcus. Both fMRI and MEG identified spatially well-overlapped activity within the targeted area in each subject. For MEG, in three subjects a very precise activation in V1 was identified at 42 ms for at least one of the two larger stimulus sizes (radii 4.5 and 5.2°). When this V1 activity was present, it marked the beginning of a weak wave of excitations in striate and extrastriate areas which ended at 50 ms (M50). The beginning of the next wave of activations (M70) was also marked by a brief V1 activation, mainly between 50 and 60 ms. The mean separation between V1 activation centers identified by fMRI and the earliest MEG activation was 3–5 mm.

Introduction

Most of our data about localization and retinotopic organization of the primary visual cortex (striate cortex or V1) in humans come from lesion studies (Horton and Hoyt, 1991), functional imaging with positron emission tomography (PET) (Fox et al., 1987), and functional magnetic resonance imaging (fMRI) DeYoe et al 1996, Engel et al 1997, Sereno et al 1995. Localization accuracy and resolution of fMRI underwent rapid advancement in the past decade. Recent studies aimed successfully at submillimeter structures such as ocular dominance columns in humans Cheng et al 2001, Menon et al 1997, Menon and Goodyear 1999 and orientation columns in cat’s visual cortex (Kim et al., 2000). However, the temporal resolution of PET and fMRI and other similar techniques that depend on slow metabolic changes that accompany neuronal activity rather than the neuronal activity directly is not enough to study the dynamics that govern the activation of different brain areas (Ogawa et al., 1998). Magnetoencephalography (MEG) is a noninvasive technique that directly measures mass postsynaptic neuronal potentials (Hamalainen et al., 1993) with excellent native temporal resolution. However, the localization of MEG has been the subject of intense study and, at times, controversy, partly because no “gold standard” for comparison is available in normal human brains.

Primary visual cortex offers a precise testing ground to study the limits of each technique and to compare electromagnetic and metabolic sources with respect to the known retinotopic organization of V1. The standard cruciform model Okada 1983, Onofrj et al 1995 was only partially supported by source estimation from visually evoked electromagnetographic (EEG) data Ahlfors et al 1992, Maclin et al 1983, Nakamura et al 1997, Slotnick et al 1999. The results were often inconsistent with the known retinotopic organization Aine et al 1996, Onofrj et al 1995. Direct comparison between electromagnetic and metabolic methods provides, at a first glance, an ideal combination. Many such recent studies met with limited success, reporting dipole sources estimated from either EEG or MEG measurements to be one to several centimeters away from the fMRI activation loci in V1 Beisteiner et al 1997, Disbrow et al 2000, Eulitz et al 1994, Gratton et al 1997, Roberts et al 2000, Stippich et al 1998. The discrepancies were attributed either to technical problems with MEG or fMRI, or to the ambiguity of biomagnetic inverse solutions, but they may simply reflect fundamentally different hemodynamic and electrophysiological processes Bonmassar et al 2001, Nunez and Silberstein 2000 and the fact that each probes very different time scales (Ioannides, 1999). On the other hand, increased neuronal activity that occurs with reduced firing synchrony may not produce detectable scalp electric or magnetic fields. Another possible contribution to the difference is the way solutions are calculated. The current dipole model is often used for MEG and EEG, thus reducing the activity to one or a few point-like sources computed from average data. We have recently demonstrated that localization with accurate retinotopy is easier to achieve with distributed source analysis and specifically magnetic field tomography (MFT) Ioannides 1994, Ioannides et al 1990 rather than with equivalent current dipoles because of the presence of early but labile activity in extrastriate areas (Tzelepi et al., 2001). The presence of early activation in extrastriate areas, was also demonstrated in another MEG study using distributed source analysis, but no comparison with current dipole solutions was presented (Vanni et al., 2001). In both studies activity was identified in a number of extrastriate areas, and prominently in two such areas. The first area had temporo-occipital location, consistent with the V5/MT area. The second area was in the parieto-occipital sulcus and its activation was interpreted both in the two recent MEG studies and in an earlier MEG study using current dipole analysis (Portin et al., 1999) as the human homologue of the monkey area V6 (Gallettii et al., 1999). For ease of reference we will label this area putative V6.

PET and fMRI solutions are expressed tomographically as statistical parametric maps. What is required for a better comparison is to add statistical analysis to the MFT solutions and to compare the results with established fMRI statistical tomographic solutions. It is further desirable to have a third, even approximate but independent measure of where the expected V1 activation should be, so that it is clearly discriminated from activations in areas beyond V1, such as V2/V3. This work targeted these goals, following logically from two previous investigations where we studied completely independently the localization of fMRI (Cheng et al., 2001) and MEG (Tzelepi et al., 2001). Identical reversing checkerboard patterns were used in both techniques, but with the flickering frequency optimized for each technique based on signal-to-noise ratio (SNR) measurements from a pilot experiment and other studies Singh et al 2000, Thomas and Menon 1998. For the analysis, we used similar techniques to extract tomographic statistical estimates from fMRI and MEG data. The results were compared to each other, and to the expected location derived from a separate fMRI retinotopy experiment, without any assumptions about mechanisms of electromagnetic and metabolic activation or the nature and number of generators. Finally we repeated the fMRI main experiment with one of the subjects using the two flickering frequencies in the respective MEG and fMRI experiments, confirming that the fMRI localization in V1 was nearly identical for the two flickering frequencies.

Section snippets

Subjects and stimuli

Four male subjects (ages 27–35) from RIKEN Brain Science Institute with normal or corrected to normal visual acuity participated in the experiments, which were approved by the RIKEN functional MRI safety and institutional ethical committees. The procedure and scientific background for the experiments were fully explained to each subject and a signed agreement for participation was obtained prior to the experiments. Stimuli consisted of flickering (reversing) checkerboard patterns on a

fMRI results

In all four subjects, the V1/V2 borders were identified by mapping the representation of the vertical meridian (Fig. 1 ). The representation of the horizontal meridian was located in the depth of the calcarine sulcus. Most of the V1 was buried inside the calcarine sulcus. Checkerboard rings of four different sizes were used to trace the activation at different eccentricities (Figs. 2A–2C ). For each subject, the changes in activity with eccentricity were consistent with the known retinotopic

Discussion

In our earlier MEG study (Tzelepi et al., 2001) we demonstrated that using MFT localization with accurate retinotopy is possible from relatively small number of trials with stimuli large sinusoidal spatial gratings with MFT. To test for higher resolution we needed to use smaller stimuli and a more precise comparison standard. Our earlier fMRI studies showed that such a standard could be provided by high-field fMRI (Cheng et al., 2001). In designing the experiment we were guided by

References (62)

  • A. Tzelepi et al.

    Early (N70m) neuromagnetic signal topography and striate and extrastriate generators following pattern onset quadrant stimulation

    NeuroImage

    (2001)
  • C.J. Aine et al.

    Retinotopic organization of human visual cortex: departures from the classical model

    Cereb. Cortex

    (1996)
  • R. Beisteiner et al.

    Magnetoencephalography may help to improve functional MRI brain mapping

    Eur. J. Neurosci.

    (1997)
  • H. Buchner et al.

    Fast visual evoked potential input into human area V5

    NeuroReport

    (1997)
  • V.P. Clark et al.

    Identification of early visual evoked potential generators by retinotopic and topographic analysis

    Human Brain Mapping

    (1995)
  • E.A. DeYoe et al.

    Mapping straite and extrastriate visual areas in human cerebral cortex

    Proc. Natl. Acad. Sci. USA

    (1996)
  • E.A. Disbrow et al.

    Functional MRI at 1.5 tesla: a comparison of the blood oxygenation level-dependent signal and electrophysiology

    Proc. Natl. Acad. Sci. USA

    (2000)
  • S.A. Engel et al.

    Retinotopic organization in human visual cortex and the spatial precision of functional MRI

    Cereb. Cortex

    (1997)
  • C. Eulitz et al.

    Comparison of magnetic and metabolic brain activity during a verb generation task

    NeuroReport

    (1994)
  • D.H. Ffytche et al.

    The parallel visual motion inputs into areas V1 and V5 of human cerebral cortex

    Brain

    (1995)
  • P.T. Fox et al.

    Retinotopic organization of human visual cortex mapped with positron-emission tomography

    J. Neurosci.

    (1987)
  • J.J. Foxe et al.

    Flow of activation from V1 to frontal cortex in humans. A framework for defining early visual processing

    Exp. Brain. Res.

    (2002)
  • C. Galletti et al.

    The cortical visual area V6: brain location and visual topography

    Eur. J. Neurosci.

    (1999)
  • J.S. Gati et al.

    Experimental determination of the BOLD field strength dependence in vessels and tissue

    Magn. Reson. Med.

    (1997)
  • M. Hamalainen et al.

    Magnetoencephalography—theory, instrumentation, and applications to noninvasive studies of the working human brain

    Rev. Mod. Phys.

    (1993)
  • S.A. Hillyard et al.

    Event-related brain potentials in the study of visual selective attention

    Proc. Natl. Acad. Sci. USA

    (1998)
  • J.C. Horton et al.

    The representation of the visual field in human striate cortex. A revision of the classic Holmes map

    Arch. Ophthalmol.

    (1991)
  • X. Hu et al.

    Reduction of signal fluctuation in functional MRI using navigator echoes

    Magn. Reson. Med.

    (1994)
  • X. Hu et al.

    Retrospective estimation and correction of physiological fluctuation in functional MRI

    Magn. Reson. Med.

    (1995)
  • A.A. Ioannides

    Estimation of brain activity using magnetic field tomography and large scale communication within the brain

  • A.A. Ioannides

    Problems associated with the combination of MEG and fMRI data: theoretical basis and results in practice

  • Cited by (113)

    View all citing articles on Scopus
    View full text