Related Experiment Video
Updated: Jan 17, 2026
17:06
Co-analysis of Brain Structure and Function using fMRI and Diffusion-weighted Imaging
Published on: November 8, 2012
26.9K
Frequency-dependence in multidimensional diffusion-relaxation correlation MRI of the brain: Overfitting or meaningful
Maxime Yon1,2, Omar Narvaez2, Jan Martin1
1Department of Chemistry, Lund University, Lund, Sweden.
Imaging Neuroscience (Cambridge, Mass.)
|September 25, 2025
Summary
Frequency-dependent inversion in diffusion-relaxation correlation MRI improves accuracy for microimaging studies of brain tissue. However, for clinical in vivo human brain scans, frequency dependence has minimal impact on diffusion metrics.
Area of Science:
- Magnetic Resonance Imaging
- Biophysics
- Neuroscience
Background:
- Diffusion MRI provides insights into brain microstructure but is complicated by intravoxel heterogeneity.
- Multidimensional diffusion-relaxation correlation MRI with tensor-valued diffusion encoding characterizes this heterogeneity.
- Frequency-dependence has been incorporated to better resolve restricted diffusion effects.
Purpose of the Study:
- To investigate the impact of including or omitting frequency-dependence in data inversion for diffusion-relaxation correlation MRI.
- To assess the effects across various samples including liquids, tumor tissue, ex vivo mouse brain, and in vivo human brain.
Main Methods:
- Utilized tensor-valued diffusion encoding with oscillating gradient waveforms to explore frequency dependence.
- Performed data inversion with and without explicit consideration of frequency-dependence.
- Analyzed data from microimaging (35-320 Hz) and clinical scanners (5-11 Hz) at high b-values (up to 4·10^9 s/m^2).
Main Results:
- Inclusion of frequency-dependence significantly reduced fit residuals and bias in diffusion metrics for microimaging of tumor and brain tissues with micrometer structures.
- For in vivo human brain data acquired over a narrow frequency range (5-11 Hz), analyses with and without frequency-dependence yielded similar results.
- Fit residuals and diffusion metrics were comparable between the two analysis approaches in the clinical scanner scenario.
Conclusions:
- Frequency-dependent inversion is a valuable tool for diffusion-relaxation correlation MRI, particularly for microimaging applications with wide frequency ranges.
- The impact of frequency-dependence is less pronounced in clinical settings with limited frequency ranges.
- Frequency-dependent inversion can be generally applied to diffusion-relaxation correlation MRI data, regardless of observable restricted diffusion effects.
