Related Experiment Video
Updated: Nov 7, 2025
09:33
Diffusion Tensor Magnetic Resonance Imaging in the Analysis of Neurodegenerative Diseases
Published on: July 28, 2013
28.8K
Automated cerebral cortex segmentation based solely on diffusion tensor imaging for investigating cortical
Graham Little1, Christian Beaulieu1
1Department of Biomedical Engineering, University of Alberta, 1098 Research Transition Facility, 8308-114 Street, Edmonton, Alberta T6G 2V2, Canada.
Neuroimage
|May 2, 2021
Summary
A new method automatically segments the human cortex using Diffusion Tensor Imaging (DTI) directly, eliminating the need for T1-weighted scans. This approach enables accurate cortical analysis in native DTI space, simplifying neuroimaging studies.
Area of Science:
- Neuroimaging
- Medical Physics
- Radiology
Background:
- Cortical analysis using Diffusion Tensor Imaging (DTI) typically requires segmenting inner and outer cortical boundaries from high-resolution 3D-T1-weighted images.
- This conventional approach necessitates an additional structural scan, which may not be practical for all imaging studies.
Purpose of the Study:
- To develop and validate an automatic cortical boundary segmentation method that operates directly on native DTI images.
- To assess the feasibility and accuracy of extracting DTI parameters from the human cortex without additional structural imaging.
Main Methods:
- Developed an automatic segmentation method using fractional anisotropy (FA) maps and mean diffusion weighted images (DWI) from native DTI data.
- Compared the proposed method's segmentations with conventional segmentations derived from high-resolution T1 structural images in 5 participants.
- Applied the method to 15 healthy young adults to measure cortical DTI parameters (FA, MD, radiality) on 1.5 mm isotropic whole-brain images.
Main Results:
- The proposed method achieved reasonable cortical boundary segmentations, with over 85% agreement within ±1 mm of conventional T1-based segmentations.
- Extracted DTI parameters like FA and mean diffusivity (MD) showed relative stability across the cortex, with higher FA in specific regions (e.g., central sulcus, insula).
- Primary eigenvector orientations were predominantly radial to the cortical surface, with tangential orientations observed in sulcal regions.
Conclusions:
- The developed automatic segmentation method is feasible and accurate for cortical analysis directly in native DTI space.
- This approach eliminates the need for additional 3D T1-weighted scans, simplifying neuroimaging workflows and potentially enabling broader DTI-based cortical studies.
