Related Experiment Video
Updated: Nov 29, 2025
17:06
Co-analysis of Brain Structure and Function using fMRI and Diffusion-weighted Imaging
Published on: November 8, 2012
26.7K
Empirical field mapping for gradient nonlinearity correction of multi-site diffusion weighted MRI
Colin B Hansen1, Baxter P Rogers2, Kurt G Schilling3
1Computer Science, Vanderbilt University, Nashville, TN, USA.
Magnetic Resonance Imaging
|November 22, 2020
Summary
This study introduces a new method to correct spatial variations in diffusion-weighted MRI (DW-MRI) data, improving the accuracy and reproducibility of diffusion metrics across different scanners. The approach enhances quantitative analysis in diffusion imaging pipelines.
Area of Science:
- Medical Imaging
- Biophysics
- Quantitative MRI
Background:
- Diffusion-weighted magnetic resonance imaging (DW-MRI) metrics face challenges in inter-site and inter-scanner reproducibility due to variations in acquisition protocols, analysis models, and hardware.
- Scanner-dependent spatial variations in diffusion weighting arise from magnetic field gradient nonlinearities, complicating accurate metric interpretation.
Purpose of the Study:
- To propose and validate an empirical approach for mapping and correcting gradient nonlinearities in DW-MRI.
- To overcome limitations of relying on manufacturer specifications for scanner nonlinearity correction.
- To develop a method compatible across major scanner vendors using standardized sequences.
Main Methods:
- An empirical method was developed to map gradient nonlinearities using supported sequences across vendors.
- A prospective observational study utilized a spherical isotropic diffusion phantom and a human control volunteer.
- Data acquired using a 3T Stejskal-Tanner spin echo sequence with b-values of 1000 and 2000 s/mm² and multiple diffusion gradient directions were analyzed.
Main Results:
- The proposed correction method significantly reduced variation in mean diffusivity in phantom data across sessions compared to uncorrected data (p < 0.05).
- In human data, the method also demonstrated a significant reduction in mean diffusivity variation across different scanners (p < 0.05).
- Comparison with prior methods, including ignoring nonlinearities and using manufacturer specifications, was performed.
Conclusions:
- The developed empirical method for correcting gradient nonlinearities is simple, fast, and applicable retroactively.
- Incorporating voxel-specific b-value and b-vector maps into DW-MRI harmonization preprocessing pipelines is recommended for improved quantitative accuracy.
- This approach enhances the reproducibility and quantitative accuracy of diffusion parameters in DW-MRI.
