Related Experiment Videos
Reliability in multi-site structural MRI studies: effects of gradient non-linearity correction on phantom and human
Jorge Jovicich1, Silvester Czanner, Douglas Greve
1MGH/MIT/HMS Athinoula A. Martinos Center for Biomedical Imaging, Building 149, 13th Street, Radiology/CNY149-Room 2301, Charlestown, MA 02129, USA. jovicich@nmr.mgh.harvard.edu
Neuroimage
|November 23, 2005
Summary
Correcting imaging gradient non-linearity significantly improves the reproducibility of multi-site human magnetic resonance imaging (MRI) data. This method enhances precision for quantitative, platform-independent, multi-site evaluations in research.
Area of Science:
- Medical Imaging
- Neuroimaging
- Biophysics
Background:
- Reproducibility in multi-site structural MRI studies is crucial for accurate, quantitative, and platform-independent evaluations.
- Technological variances, particularly imaging gradient non-linearity, can limit the reliability of high-resolution MRI data across different sites and platforms.
- Addressing these variances is essential for longitudinal and multi-site clinical research.
Purpose of the Study:
- To investigate the impact of imaging gradient non-linearity on the reproducibility of multi-site human MRI.
- To develop and validate an image distortion correction method for improving MRI data consistency.
- To assess the effectiveness of the correction method in enhancing voxel-based image intensity reproducibility.
Main Methods:
- Applied a novel image distortion correction technique utilizing a spherical harmonics description of imaging gradients.
- Validated the correction method's accuracy using controlled phantom data.
- Implemented the correction on human brain MRI data from subjects scanned twice across multiple 1.5 T platforms with varying gradient characteristics.
Main Results:
- The image distortion correction method significantly improved the image intensity reproducibility of human brain MRI data.
- Within-site and across-site variability in voxel-based image intensity was substantially reduced post-correction.
- Phantom data verification confirmed the accuracy and efficacy of the spherical harmonics-based distortion correction.
Conclusions:
- Gradient non-linearity correction is a vital step for enhancing the reproducibility of multi-site structural MRI.
- The developed method offers a promising solution for achieving precise, quantitative, and platform-independent multi-site MRI evaluations.
- This approach has the potential to improve the reliability of morphometry studies and other quantitative analyses in neuroimaging research.