Tensor-valued diffusion MRI of human acute stroke
(2024) In Magnetic Resonance in Medicine 91(5). p.2126-2141- Abstract
Purpose: Tensor-valued diffusion encoding can disentangle orientation dispersion and subvoxel anisotropy, potentially offering insight into microstructural changes after cerebral ischemia. The purpose was to evaluate tensor-valued diffusion MRI in human acute ischemic stroke, assess potential confounders from diffusion time dependencies, and compare to Monte Carlo diffusion simulations of axon beading. Methods: Linear (LTE) and spherical (STE) b-tensor encoding with inherently different effective diffusion times were acquired in 21 acute ischemic stroke patients between 3 and 57 h post-onset at 3 T in 2.5 min. In an additional 10 patients, STE with 2 LTE yielding different effective diffusion times were acquired for comparison.... (More)
Purpose: Tensor-valued diffusion encoding can disentangle orientation dispersion and subvoxel anisotropy, potentially offering insight into microstructural changes after cerebral ischemia. The purpose was to evaluate tensor-valued diffusion MRI in human acute ischemic stroke, assess potential confounders from diffusion time dependencies, and compare to Monte Carlo diffusion simulations of axon beading. Methods: Linear (LTE) and spherical (STE) b-tensor encoding with inherently different effective diffusion times were acquired in 21 acute ischemic stroke patients between 3 and 57 h post-onset at 3 T in 2.5 min. In an additional 10 patients, STE with 2 LTE yielding different effective diffusion times were acquired for comparison. Diffusional variance decomposition (DIVIDE) was used to estimate microscopic anisotropy (μFA), as well as anisotropic, isotropic, and total diffusional variance (MKA, MKI, MKT). DIVIDE parameters, and diffusion tensor imaging (DTI)-derived mean diffusivity and fractional anisotropy (FA) were compared in lesion versus contralateral white matter. Monte Carlo diffusion simulations of various cylindrical geometries for all b-tensor protocols were used to interpret parameter measurements. Results: MD was ˜40% lower in lesions for all LTE/STE protocols. The DIVIDE parameters varied with effective diffusion time: higher μFA and MKA in lesion versus contralateral white matter for STE with longer effective diffusion time LTE, whereas the shorter effective diffusion time LTE protocol yielded lower μFA and MKA in lesions. Both protocols, regardless of diffusion time, were consistent with simulations of greater beading amplitude and intracellular volume fraction. Conclusion: DIVIDE parameters depend on diffusion time in acute stroke but consistently indicate neurite beading and larger intracellular volume fraction.
(Less)
- author
- organization
-
- eSSENCE: The e-Science Collaboration
- MR Physics (research group)
- Multidimensional microstructure imaging (research group)
- Medical Radiation Physics, Lund
- Diagnostic Radiology, (Lund)
- LTH Profile Area: Engineering Health
- LUCC: Lund University Cancer Centre
- MultiPark: Multidisciplinary research focused on Parkinson´s disease
- Neuroradiology (research group)
- Lund University Bioimaging Center
- publishing date
- 2024-05
- type
- Contribution to journal
- publication status
- published
- subject
- keywords
- b-tensor, diffusion time, Monte Carlo diffusion simulation, neurite beading, stroke, tensor-valued diffusion encoding
- in
- Magnetic Resonance in Medicine
- volume
- 91
- issue
- 5
- pages
- 16 pages
- publisher
- John Wiley & Sons Inc.
- external identifiers
-
- scopus:85181197209
- pmid:38156813
- ISSN
- 0740-3194
- DOI
- 10.1002/mrm.29975
- language
- English
- LU publication?
- yes
- id
- c1b7bc8c-2623-4769-9e2f-6349ac848ebf
- date added to LUP
- 2024-07-09 14:31:15
- date last changed
- 2024-09-17 22:21:21
@article{c1b7bc8c-2623-4769-9e2f-6349ac848ebf, abstract = {{<p>Purpose: Tensor-valued diffusion encoding can disentangle orientation dispersion and subvoxel anisotropy, potentially offering insight into microstructural changes after cerebral ischemia. The purpose was to evaluate tensor-valued diffusion MRI in human acute ischemic stroke, assess potential confounders from diffusion time dependencies, and compare to Monte Carlo diffusion simulations of axon beading. Methods: Linear (LTE) and spherical (STE) b-tensor encoding with inherently different effective diffusion times were acquired in 21 acute ischemic stroke patients between 3 and 57 h post-onset at 3 T in 2.5 min. In an additional 10 patients, STE with 2 LTE yielding different effective diffusion times were acquired for comparison. Diffusional variance decomposition (DIVIDE) was used to estimate microscopic anisotropy (μFA), as well as anisotropic, isotropic, and total diffusional variance (MK<sub>A</sub>, MK<sub>I</sub>, MK<sub>T</sub>). DIVIDE parameters, and diffusion tensor imaging (DTI)-derived mean diffusivity and fractional anisotropy (FA) were compared in lesion versus contralateral white matter. Monte Carlo diffusion simulations of various cylindrical geometries for all b-tensor protocols were used to interpret parameter measurements. Results: MD was ˜40% lower in lesions for all LTE/STE protocols. The DIVIDE parameters varied with effective diffusion time: higher μFA and MK<sub>A</sub> in lesion versus contralateral white matter for STE with longer effective diffusion time LTE, whereas the shorter effective diffusion time LTE protocol yielded lower μFA and MK<sub>A</sub> in lesions. Both protocols, regardless of diffusion time, were consistent with simulations of greater beading amplitude and intracellular volume fraction. Conclusion: DIVIDE parameters depend on diffusion time in acute stroke but consistently indicate neurite beading and larger intracellular volume fraction.</p>}}, author = {{Zhou, Mi and Stobbe, Robert and Szczepankiewicz, Filip and Budde, Matthew and Buck, Brian and Kate, Mahesh and Lloret, Mar and Fairall, Paige and Butcher, Ken and Shuaib, Ashfaq and Emery, Derek and Nilsson, Markus and Westin, Carl Fredrik and Beaulieu, Christian}}, issn = {{0740-3194}}, keywords = {{b-tensor; diffusion time; Monte Carlo diffusion simulation; neurite beading; stroke; tensor-valued diffusion encoding}}, language = {{eng}}, number = {{5}}, pages = {{2126--2141}}, publisher = {{John Wiley & Sons Inc.}}, series = {{Magnetic Resonance in Medicine}}, title = {{Tensor-valued diffusion MRI of human acute stroke}}, url = {{http://dx.doi.org/10.1002/mrm.29975}}, doi = {{10.1002/mrm.29975}}, volume = {{91}}, year = {{2024}}, }