Automatic analysis of cerebral shape asymmetry in 3D MRI
-
1
Tampere University of Technology, Finland
-
2
VTT Technical Research Centre of Finland, Finland
-
3
University of Turku & Turku PET Center, Finland
MRI based brain asymmetry studies have reported that the magnitude and direction of the structural asymmetry of human brain could be affected by gender, handedness and mental diseases. However, the previous studies always needed manual interaction. In this work, we developed a fully automatic method to analyze the shape asymmetry between cerebral hemispheres in 3D MRI. This method consists of four major steps. First, the cerebral hemispheres are extracted and segmented in MR brain images using an accurate method based on partial volume effect information and partial differential equations. Since no image registration is involved in the segmentation, the original shapes of the cerebral hemispheres are properly preserved [1,2]. Next, all segmented cerebrum volumes of different subjects are aligned into a MNI152 space in SPM5 using affine registration in order to remove scale differences and minimize position variabilities. In each aligned cerebral volume, the right cerebral hemisphere is mirrored to match the left hemisphere with respect to the middle longitudinal plane of the MNI152 space. Then the Dual Surface Minimization based Deformable Model [3] is used to extract smoothed surfaces from the left and mirrored right hemispheres of each aligned cerebrum volume. In each surface, the shape of the corresponding cerebral hemisphere is represented with meshes consisting of 5120 triangles and 2562 vertexes. Because of the alignment and mirroring, the same vertex represents the same location in surfaces of different cerebral hemispheres. In the last stage, one average surface is obtained out of surfaces of the left and mirrored right cerebral hemispheres of all studied subjects. For an individual subject, the shape difference between each cerebral hemisphere surface and the average surface is calculated as the difference of normals at every vertex and by then projecting it on the surface normal of the average surface. The Wilcoxon two sided rank sum test is utilized to evaluate the statistical significance of the variability of the normal difference from the average surface at each vertex between left and right cerebral hemispheres. The proposed method was applied to study the shape asymmetry of cerebrum in schizophrenia with a dataset containing T1 weighted MR images of 19 healthy controls (12 males, 7 females) and 18 schizophrenic subjects (11 males, 7 females). All subjects were right-handed, and schizophrenic subjects were never medicated. We found, in schizophrenia, reduction of shape asymmetry in the superior frontal lobe in females and occipital lobe in males, and increase of shape asymmetry in the occipital lobe in females.
References
1. L. Zhao and J. Tohka, EMBC `08, Vancouver, Canada, 2008, submitted
2. L. Zhao et al., Lecture Notes in Computer Science 4522: 581-590, 2007.
3. J. Tohka, International Journal of Image and Graphics (IJIG) 4(3): 405??432, 2004
Conference:
Neuroinformatics 2008, Stockholm, Sweden, 7 Sep - 9 Sep, 2008.
Presentation Type:
Poster Presentation
Topic:
Neuroimaging
Citation:
Pepe
A,
Zhao
L,
Koikkalainen
J,
Tohka
J,
Hietala
J and
Ruotsalainen
U
(2008). Automatic analysis of cerebral shape asymmetry in 3D MRI.
Front. Neuroinform.
Conference Abstract:
Neuroinformatics 2008.
doi: 10.3389/conf.neuro.11.2008.01.066
Copyright:
The abstracts in this collection have not been subject to any Frontiers peer review or checks, and are not endorsed by Frontiers.
They are made available through the Frontiers publishing platform as a service to conference organizers and presenters.
The copyright in the individual abstracts is owned by the author of each abstract or his/her employer unless otherwise stated.
Each abstract, as well as the collection of abstracts, are published under a Creative Commons CC-BY 4.0 (attribution) licence (https://creativecommons.org/licenses/by/4.0/) and may thus be reproduced, translated, adapted and be the subject of derivative works provided the authors and Frontiers are attributed.
For Frontiers’ terms and conditions please see https://www.frontiersin.org/legal/terms-and-conditions.
Received:
28 Jul 2008;
Published Online:
28 Jul 2008.
*
Correspondence:
Antonietta Pepe, Tampere University of Technology, Tampere, Finland, antonietta.pepe@tut.fi