Anatomy and Embryology

, Volume 210, Issue 5, pp 373–386

Quantitative architectural analysis: a new approach to cortical mapping

Authors

    • C. and O. Vogt Brain Research InstituteUniversity of Düsseldorf
  • N. Palomero-Gallagher
    • Research Center Jülich
  • P. Morosan
    • Research Center Jülich
  • S. B. Eickhoff
    • Research Center Jülich
  • T. Kowalski
    • Research Center Jülich
  • K. de Vos
    • Netherlands Institute for Brain Research
  • K. Amunts
    • Research Center Jülich
  • K. Zilles
    • C. and O. Vogt Brain Research InstituteUniversity of Düsseldorf
    • Research Center Jülich
    • Brain Imaging Center West
Original Article

DOI: 10.1007/s00429-005-0028-2

Cite this article as:
Schleicher, A., Palomero-Gallagher, N., Morosan, P. et al. Anat Embryol (2005) 210: 373. doi:10.1007/s00429-005-0028-2

Abstract

Recent progress in anatomical and functional MRI has revived the demand for a reliable, topographic map of the human cerebral cortex. Till date, interpretations of specific activations found in functional imaging studies and their topographical analysis in a spatial reference system are, often, still based on classical architectonic maps. The most commonly used reference atlas is that of Brodmann and his successors, despite its severe inherent drawbacks. One obvious weakness in traditional, architectural mapping is the subjective nature of localising borders between cortical areas, by means of a purely visual, microscopical examination of histological specimens. To overcome this limitation, more objective, quantitative mapping procedures have been established in the past years. The quantification of the neocortical, laminar pattern by defining intensity line profiles across the cortical layers, has a long tradition. During the last years, this method has been extended to enable a reliable, reproducible mapping of the cortex based on image analysis and multivariate statistics. Methodological approaches to such algorithm-based, cortical mapping were published for various architectural modalities. In our contribution, principles of algorithm-based mapping are described for cyto- and receptorarchitecture. In a cytoarchitectural parcellation of the human auditory cortex, using a sliding window procedure, the classical areal pattern of the human superior temporal gyrus was modified by a replacing of Brodmann’s areas 41, 42, 22 and parts of area 21, with a novel, more detailed map. An extension and optimisation of the sliding window procedure to the specific requirements of receptorarchitectonic mapping, is also described using the macaque central sulcus and adjacent superior parietal lobule as a second, biologically independent example. Algorithm-based mapping procedures, however, are not limited to these two architectural modalities, but can be applied to all images in which a laminar cortical pattern can be detected and quantified, e.g. myeloarchitectonic and in vivo high resolution MR imaging. Defining cortical borders, based on changes in cortical lamination in high resolution, in vivo structural MR images will result in a rapid increase of our knowledge on the structural parcellation of the human cerebral cortex.

Keywords

CytoarchitectureReceptorarchitectureAuditory cortexCentral sulcusMultivariate statistics

Abbreviations

2D

two-dimensional

3D

three-dimensional

AChE

acetylcholinesterase

BA

Brodmann’s area

CL

cluster analysis

Cs

central sulcus

d

cortical depth

GLI

Grey level index

HG

Heschl’s gyrus

HS

Heschl’s sulcus

ips

intraparietal sulcus

l

length of feature vector

MD

Mahalanobis distance

MRI

magnet resonance imaging

MTG

middle temporal gyrus

n

number of profiles in a cortical sector

P

level of significance

ROI

region of interest

STG

superior temporal gyrus

STS

superior temporal sulcus

SW

sliding window

TP

temporal plane

w

width of a cortical layer

Copyright information

© Springer-Verlag 2005