Elsevier

NeuroImage

Volume 49, Issue 2, 15 January 2010, Pages 1171-1179
NeuroImage

Comparison of functional and cytoarchitectonic maps of human visual areas V1, V2, V3d, V3v, and V4(v)

https://doi.org/10.1016/j.neuroimage.2009.09.063Get rights and content

Abstract

Cytoarchitectonic maps of human striate and extrastriate visual cortex based upon post-mortem brains can be correlated with functionally defined cortical areas using, for example, fMRI. We here assess the correspondence of anatomical maps of the visual cortex with functionally defined in vivo visual areas using retinotopic mapping. To this end, anatomical maximum probability maps (aMPM) derived from individual cytoarchitectonic maps of striate and extrastriate visual areas were compared with functional localisers for the early visual areas. Using fMRI, we delineated dorsal and ventral human retinotopic areas V1, V2, and V3, as well as a quarter-field visual field representation lateral to V3v, V4(v), in 24 healthy subjects. Based on these individual definitions, a functional maximum probability map (fMPM) was then computed in analogy to the aMPM. Functional and anatomical MPMs were highly correlated at group level: 78.5% of activated voxels in the fMPM were correctly assigned by the aMPM. The group aMPM was less effective in predicting functional retinotopic areas in the individual brain due to the large inter-individual variability in the location and extent of visual areas (mean overlap 32–69%). We conclude that cytoarchitectonic maps of striate and extrastriate visual areas may provide a valuable method for assigning functional group activations and thus add valuable a priori knowledge to the analysis of functional imaging data of the visual cortex.

Introduction

Retinotopic mapping has become the current gold standard for the in vivo delineation of early visual cortical areas in the human brain using non-invasive functional imaging. The principle behind this approach is to delineate the borders of retinotopically organised areas by finding the visual field meridian representations on cortical surface representations of individual brains (Sereno et al., 1995, DeYoe et al., 1996, Engel et al., 1997). The collection of retinotopic functional imaging data together with high-resolution structural data needed for cortical surface reconstruction is, however, time consuming and may thus limit the applicability of this approach, e.g., when examining patients who do not tolerate being scanned for a long time. Moreover excentric visual field localisations are technically difficult to access by retinotopic mapping due to limitations in the visual angle covered by standard stimulation devices. In practice, this approach is hence often restricted to the delineation of cortical regions representing the more central visual field. Finally, it is hardly feasible to obtain functional localisers for all areas involved in complex cortical networks, even in the visual cortex where higher areas become more and more tuned to particular stimulus features.

Anatomical a priori information on the location and extent of cortical areas in the visual system may help to overcome such limitations by providing a complementary frame of reference. Such anatomical maps based on the probabilistic, observer-independent mapping of cortical areas (Schleicher et al., 2005) have already been applied as references for the localisation of functional activations in fMRI and MEG studies of visual processing (Larsson et al., 2002, Barnikol et al., 2006, Dammers et al., 2007, Wohlschläger et al., 2005). Given that the location and extent of visual areas can also be delineated in vivo by retinotopic mapping as described above, the question obviously arises, how closely the maps derived from in vivo functional imaging and post-mortem anatomical analysis correspond to each other.

This question is addressed here directly by correlating architectonic and functional delineations of the early visual areas V1, V2, V3d, V3v, and V4(v). Retinotopic visual areas were delineated on surface reconstructions of the cortical grey–white matter boundary, which has emerged as the method of choice as it allows a definition of the meridian representation in spite of the convoluted cortical surface. Using this approach, the anatomical correlates of areas V1 (BA17) and V2 (BA18) (Amunts et al., 2000), which have already been compared to functional data (Wohlschläger et al., 2005) using a volume-based delineation, were re-analysed. We then quantitatively compared functionally defined dorsal V3 (V3d), ventral V3 (V3v), as well as a quarter-field representation lateral to V3v, V4(v), with their cytoarchitectonically defined putative anatomical correlate areas hOc3d (human occipital area 3, dorsal; Kujovic et al., 2007), hOc3v (human occipital area 3, ventral; Rottschy et al., 2007), and hOc4v (human occipital area 4, ventral; Rottschy et al., 2007).

Section snippets

Subjects and experimental paradigm

Twenty-four healthy subjects aged 21–38 years (mean age 26.4 ± 4.5 years, 10 female) participated in the study. All subjects were healthy without any history of neurological or psychiatric diseases, and had normal visual acuity. Informed written consent was obtained from all subjects. The study was approved by the ethics committee of the University Hospital Aachen, Germany.

Stimuli were presented via fMRI compatible goggles (subjects S1–S10: Silent Vision™, Avotec, FL; S11–S24: VisuaStim™,

Results

Using retinotopic mapping of polar angle and subsequent delineation of visual areas based on the representation of the vertical meridians, the following areas were defined separately for the left and right hemispheres of all 24 subjects: The dorsal and ventral parts of visual areas V1, V2, and V3, as well as a quarter-field visual field representation lateral of ventral V3, i.e., area V4(v). Similar to previous descriptions of anatomical data, also the functional probability maps for these

Discussion

The cortical parcellation defined by quantitative cytoarchitectonic analysis and summarised by the maximum probability map is not only supported by other histological modalities such as receptor autoradiography (for a review see Zilles and Amunts, 2009) but also by in vivo approaches. Most of the work comparing histological borders and functional data has been performed by comparing the location of significant fMRI activations to cytoarchitectonic areas. Using such an approach, it has been

Conclusion

Our data's most important implication for fMRI studies will pertain to situations where retinotopic maps have not been acquired for the individual subject, but detailed information about the activated visual areas is nevertheless useful. While retinotopic mapping is now routinely possible, this situation may nevertheless arise in experiments where scanning time is very limited (for example in the investigation of patients) or where the main focus rested on other functional systems. In such

Acknowledgments

We are grateful to our colleagues from the MR group of the Institute for Neurosciences and Medicine for their assistance in acquiring the fMRI data. G.R.F. was supported by the Deutsche Forschungsgemeinschaft. S.B.E. was supported by the Human Brain Project (R01-MH074457-01A1) and the Helmholtz Initiative on Systems-Biology “The Human Brain Model.” This work was part of a Human Brain Project/Neuroinformatics Research Grant funded by the National Institute of Biomedical Imaging and

References (52)

  • WandellB.A. et al.

    Visual field maps in human cortex

    Neuron

    (2007)
  • WarnkingJ. et al.

    fMRI retinotopic mapping—step by step

    NeuroImage

    (2002)
  • WohlschlägerA.M. et al.

    Linking retinotopic fMRI mapping and anatomical probability maps of human occipital areas V1 and V2

    NeuroImage

    (2005)
  • AndrewsT.J. et al.

    Correlated size variations in human visual cortex, lateral geniculate nucleus, and optic tract

    J. Neurosci.

    (1997)
  • AnwanderA. et al.

    Connectivity-based parcellation of Broca's area

    Cereb. Cortex

    (2007)
  • ArgallB.D. et al.

    Simplified intersubject averaging on the cortical surface using SUMA

    Hum. Brain Mapp.

    (2006)
  • BrewerA.A. et al.

    Visual field maps and stimulus selectivity in human ventral occipital cortex

    Nat. Neurosci.

    (2005)
  • BrückeE.

    Über den Nutzeffect intermittierender Netzhautreizungen. Sitzungsbericht der k. Akad. Wissenschaften, math.-naturwiss

    Klasse, Wien

    (1864)
  • DeYoeE.A. et al.

    Mapping striate and extrastriate visual areas in human Cereb. Cortex

    Proc. Natl. Acad. Sci. USA

    (1996)
  • DoughertyR.F. et al.

    Visual field representations and locations of visual areas V1/2/3 in human visual cortex

    J. Vis.

    (2003)
  • EickhoffS.B. et al.

    Laminar distribution and co-distribution of neurotransmitter receptors in early human visual cortex

    Brain Struct. Funct.

    (2007)
  • EickhoffS.B. et al.

    The somatotopic organization of cytoarchitectonic areas on the human parietal operculum

    Cereb. Cortex

    (2007)
  • EickhoffS.B. et al.

    Organizational principles of human visual cortex revealed by receptor mapping

    Cereb. Cortex

    (2008)
  • EngelS.A. et al.

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

    Cereb. Cortex

    (1997)
  • FellemanD.J. et al.

    Distributed hierarchical processing in the primate Cereb. Cortex

    Cereb. Cortex

    (1991)
  • FischlB. et al.

    High-resolution intersubject averaging and a coordinate system for the cortical surface

    Hum. Brain Mapp.

    (1999)
  • Cited by (0)

    View full text