Related Experiment Video
Updated: May 23, 2026
13:26
Measuring Connectivity in the Primary Visual Pathway in Human Albinism Using Diffusion Tensor Imaging and Tractography
Published on: August 11, 2016
Tract specific reproducibility of tractography based morphology and diffusion metrics
René M H Besseling1, Jacobus F A Jansen, Geke M Overvliet
1Department of Radiology, Maastricht University Medical Center, Maastricht, The Netherlands.
Plos One
|April 10, 2012
Summary
Tractography reproducibility varies by white matter tract. Fractional anisotropy (FA) and apparent diffusion coefficient (ADC) show good reproducibility, suggesting clinical potential for these diffusion tensor imaging (DTI) metrics. Tract Density Imaging (TDI) demonstrated poor reproducibility.
Area of Science:
- Neuroimaging
- Biomedical Engineering
- Medical Physics
Background:
- Tractography reproducibility is crucial for detecting pathological changes.
- Tract morphology reproducibility and Tract Density Imaging (TDI) at the tract-specific level remain understudied.
- Assessing the reliability of neuroimaging techniques is vital for clinical applications.
Purpose of the Study:
- To systematically evaluate the inter-session reproducibility of tractography morphology and key metrics.
- To assess the tract-specific reproducibility of Tract Density Imaging (TDI).
- To compare the reproducibility of different diffusion tensor imaging (DTI) metrics including FA, ADC, volume, and TDI.
Main Methods:
- Diffusion tensor imaging (DTI) and probabilistic constrained spherical deconvolution (CSD) tractography were performed twice in 9 healthy subjects.
- Tractography was conducted using common space seed and target regions for major white matter tracts.
- Reproducibility of tract morphology (Dice Similarity Coefficient) and metrics (COV, ICC) for FA, ADC, volume, and TDI were analyzed for proximal and extended segmentations.
Main Results:
- Tract morphology reproducibility (Dice Similarity Coefficient) ranged from 0.70-0.92 for proximal segmentations, decreasing by 5-10% for extended segmentations.
- Fractional anisotropy (FA) and apparent diffusion coefficient (ADC) showed low coefficient of variation (COV) and high intraclass correlation coefficient (ICC) values.
- Tract Density Imaging (TDI) exhibited high COV (8-31%) and low ICC (0.48-0.70), indicating unfavorable reproducibility.
Conclusions:
- Reproducibility of diffusion tensor imaging (DTI) metrics is tract-dependent.
- FA and ADC demonstrate high reproducibility, suggesting significant clinical potential for these diffusion metrics.
- While tract volume shows higher variability, its moderate to high ICC indicates subject-differentiating power; Tract Density Imaging (TDI) reproducibility is unfavorable for clinical use.
