Human MR Evaluation of Cortical Thickness Using CIVET-2.1
Stand-By Time
Wednesday, June 28, 2017: 12:45 PM - 2:45 PMSubmission No:
4166
Submission Type:
Abstract Submission
On Display:
Wednesday, June 28 & Thursday, June 29
Authors:
Claude Lepage1, Lindsay Lewis1, Seun Jeon1, Patrick Bermudez1, Najmeh Khalili-Mahani1, Mona Omidyeganeh1, Alex Zijdenbos2, Robert Vincent1, Reza Adalat1, Alan Evans1
Institutions:
1McGill Centre for Integrative Neuroscience, Montreal Neurological Institute, McGill University, Montreal, Quebec, Canada, 2Biospective, Inc., Montreal, Quebec, Canada
First Author:
Claude Lepage, Ph.D.
-
Lecture Information
|
Contact Me
McGill Centre for Integrative Neuroscience, Montreal Neurological Institute, McGill University
Montreal, Quebec, Canada
McGill Centre for Integrative Neuroscience, Montreal Neurological Institute, McGill University
Montreal, Quebec, Canada
Introduction:
We introduce CIVET-2.1, a fully-automated pipeline developed at the MNI for extraction of cortical surfaces and evaluation of cortical thickness from in-vivo MR images of the human brain (Kim 2005). CIVET is available via the CBRAIN platform (Sherif 2014, https://github.com/aces/cbrain/), which is linked to the computing resources from Compute Canada (https://www.computecanada.ca), and on Cloud computing facilities (Glatard, 2015) for large scale processing of thousands of scans.The current work provides an overview of CIVET's main features and of the major changes in the latest release.
Methods:
History
CIVET has evolved over the last decade. Versions up to 1.1.12 used a deformable ellipsoid model for the extraction of the white surface (MacDonald 2000). This model was replaced by a highly accurate marching-cubes algorithm in version 2.0 and further refined in version 2.1.
Description (Kim 2005)
The pipeline takes as input a t1w image (in-vivo human brain MRI) and transforms it to stereotaxic MNI space. The image is then corrected for scanner inhomogeneities and masked for the brain, in stereotaxic space. The image is classified into the main tissue classes: white matter (WM), cortical (GM) and sub-cortical (SC) gray matter, cerebro-spinal fluid (CSF), and background (BG). A white surface with spherical topology is fitted to the classified white matter mask (per hemisphere) using a marching-cubes algorithm. This white surface is further fitted locally to the maximum t1w intensity gradient to account for shortcomings in tissue classification due to residual field inhomogeneities. The gray surface is obtained by expanding the white surface to the pial boundary. Cortical thickness is calculated as the distance between the cortical surfaces and is smoothed on the surface to improve signal-to-noise ratio. The surfaces are registered based on morphological landmarks (sulcal patterns) and resampled at predefined sampling points for a basis of group comparison. Group statistical analyses are performed to regress cortical thickness and other corticometric maps to behavioral data.
Main Features
-Processing at the isotropic voxel size of 0.50mm (even for t1w input of lower resolution, such as 1mm), with consequently reduced discretization error in the tissue classification and refined anatomical delineations.
-Estimation of partial volumes by iteratively updating the tissue class thresholds, which, at convergence, become invariant to input priors for training of the classifier.
-Surfaces extraction at low resolution (40962 vertices) or high resolution (163842 vertices).
-Accurate extraction of the white surface using a spherical topology preserving marching-cubes algorithm.
-T1w intensity gradient correction on the white surface and incorporation of GM partial volumes in the determination of the pial surface.
-Surface parcellations such as DKT and AAL for regional analyses.
-Surface registration for corticometric group comparisons.
-Thorough and efficient quality control tools for assessing the quality of the results against noisy input scans and processing malfunctions.
CIVET has evolved over the last decade. Versions up to 1.1.12 used a deformable ellipsoid model for the extraction of the white surface (MacDonald 2000). This model was replaced by a highly accurate marching-cubes algorithm in version 2.0 and further refined in version 2.1.
Description (Kim 2005)
The pipeline takes as input a t1w image (in-vivo human brain MRI) and transforms it to stereotaxic MNI space. The image is then corrected for scanner inhomogeneities and masked for the brain, in stereotaxic space. The image is classified into the main tissue classes: white matter (WM), cortical (GM) and sub-cortical (SC) gray matter, cerebro-spinal fluid (CSF), and background (BG). A white surface with spherical topology is fitted to the classified white matter mask (per hemisphere) using a marching-cubes algorithm. This white surface is further fitted locally to the maximum t1w intensity gradient to account for shortcomings in tissue classification due to residual field inhomogeneities. The gray surface is obtained by expanding the white surface to the pial boundary. Cortical thickness is calculated as the distance between the cortical surfaces and is smoothed on the surface to improve signal-to-noise ratio. The surfaces are registered based on morphological landmarks (sulcal patterns) and resampled at predefined sampling points for a basis of group comparison. Group statistical analyses are performed to regress cortical thickness and other corticometric maps to behavioral data.
Main Features
-Processing at the isotropic voxel size of 0.50mm (even for t1w input of lower resolution, such as 1mm), with consequently reduced discretization error in the tissue classification and refined anatomical delineations.
-Estimation of partial volumes by iteratively updating the tissue class thresholds, which, at convergence, become invariant to input priors for training of the classifier.
-Surfaces extraction at low resolution (40962 vertices) or high resolution (163842 vertices).
-Accurate extraction of the white surface using a spherical topology preserving marching-cubes algorithm.
-T1w intensity gradient correction on the white surface and incorporation of GM partial volumes in the determination of the pial surface.
-Surface parcellations such as DKT and AAL for regional analyses.
-Surface registration for corticometric group comparisons.
-Thorough and efficient quality control tools for assessing the quality of the results against noisy input scans and processing malfunctions.
Results:
Results are shown to depict the recent major advances in CIVET-2.1. Figure 1 shows the action of the t1w intensity gradient correction, with the final placement of the white surface being invariant of the user-defined N3 spline distance for inhomogeneity field correction (Sled 1998). Figure 2 shows the benefits of processing the image at 0.50mm voxel size, both in terms of white and gray surface accuracy.
Conclusions:
From batch processing on CBRAIN to efficient quality control to corticometric analyses, the updated CIVET-2.1 pipeline provides an integrated environment for automated estimation of cortical measures from MR images.
These recent upgrades and particularly the 0.50mm volumetric template option place CIVET in an optimal position for the advent of hi-resolution 7T data.
These recent upgrades and particularly the 0.50mm volumetric template option place CIVET in an optimal position for the advent of hi-resolution 7T data.
Imaging Methods:
Anatomical MRI 2
Modeling and Analysis Methods:
Methods Development 1
Keywords:
Computational Neuroscience
Data analysis
Design and Analysis
MRI
STRUCTURAL MRI
1|2Indicates the priority used for review
Would you accept an oral presentation if your abstract is selected for an oral session?
I would be willing to discuss my abstract with members of the press should my abstract be marked newsworthy:
Please indicate below if your study was a "resting state" or "task-activation” study.
By submitting your proposal, you grant permission for the Organization for Human Brain Mapping (OHBM) to distribute the presentation in any format, including video, audio print and electronic text through OHBM OnDemand, social media channels or other electronic media and on the OHBM website.
Healthy subjects only or patients (note that patient studies may also involve healthy subjects):
Internal Review Board (IRB) or Animal Use and Care Committee (AUCC) Approval. Please indicate approval below. Please note: Failure to have IRB or AUCC approval, if applicable will lead to automatic rejection of abstract.
Please indicate which methods were used in your research:
For human MRI, what field strength scanner do you use?
Which processing packages did you use for your study?
Provide references in author date format
Kim, J.S. (2005), 'Automated 3-D extraction and evaluation of the inner and outer cortical surfaces using a Laplacian map and partial volume effect classification', NeuroImage 27(1): 210-221.
MacDonald, D. (2000), 'Automated 3-D Extraction of Inner and Outer Surfaces of Cerebral Cortex from MRI', NeuroImage 12: 340-356.
Sherif, T. (2014), 'CBRAIN: a web-based, distributed computing platform for collaborative neuroimaging research', Front. Neuroinform. 8:54.
Sled, J.G. (1998), 'A non-parametric method for automatic correction of intensity non-uniformity in MRI data', in 'IEEE Transactions on Medical Imaging', 17(1): 87-97.