Non-invasive imaging of CSF-mediated brain clearance pathways via assessment of perivascular fluid movement with diffusion tensor MRI
Figures
![](https://iiif.elifesciences.org/lax/34028%2Felife-34028-fig1-v2.tif/full/617,/0/default.jpg)
Non-invasive MRI of perivascular channels.
(A) Example b0 MRI image. The position and orientation of the imaging slice is adjusted to optimally visualize the perivascular space (PVS) around both branches of the MCA. Bright signal can be observed from fluid filled compartments: CSF in the subarachnoid space around the Circle of Willis (CoW); fluid in the perivascular space that surrounds the MCA; the ventral aspect of the third ventricle. (B) Photograph of the ventral aspect of the rat brain surface illustrating a putative PVS surrounding the middle cerebral artery (MCA) (reproduced with permission from Lochhead et al., 2015).
![](https://iiif.elifesciences.org/lax/34028%2Felife-34028-fig2-v2.tif/full/617,/0/default.jpg)
Multiple direction diffusion weighted imaging of subarachnoid and perivasular fluid movement.
(A) Example bO and ‘diffusion weighted’ images acquired with the motion probing gradients applied in 3 orthogonal directions respectively. (B) The mean D* calculated within ROIs [see insert] in the right perivascular space (red), left perivascular space (blue), subarachnoid space (green) with the motion probing gradients applied in three orthogonal directions (±SEM).
![](https://iiif.elifesciences.org/lax/34028%2Felife-34028-fig2-figsupp1-v2.tif/full/617,/0/default.jpg)
The individual animal D* calculated within ROIs in the right perivascular space.
(A), left perivascular space (B), subarachnoid space (C) with the motion probing gradients applied in three orthogonal directions (x-axis). Each line represents an individual animal (n = 10).
![](https://iiif.elifesciences.org/lax/34028%2Felife-34028-fig3-v2.tif/full/617,/0/default.jpg)
Diffusion tensor imaging of subarachnoid and perivascular fluid movement.
(A) bO image with ROIs in the right and left PVS and subarachnoid space highlighted in blue, red and green respectively. The mean pseudo-diffusion tensor ellipsoid within the subarachnoid space ROI (B) and right (C) and left (D) PVS respectively across the six rats. The pseudo-diffusion tensors for each individual animal are shown in Figure 3—figure supplement 1. (E) Example map of pseudo-diffusion tensor ellipsoids with corresponding b0 image (insert).
![](https://iiif.elifesciences.org/lax/34028%2Felife-34028-fig3-figsupp1-v2.tif/full/617,/0/default.jpg)
D* tensors within the left and right perivascular space (PVS) and subarachnoid space ROIs for each of the individual subjects imaged in part ii).
The corresponding D* tensor for the dead brain is also shown.
![](https://iiif.elifesciences.org/lax/34028%2Felife-34028-fig4-v2.tif/full/617,/0/default.jpg)
Cerebral arterial pulsation drives non-invasive measures of perivascular fluid movement.
(A) b0 image (first column) and D* maps during arterial pulsation (second and third column) and during diastole (fourth and fifth column) from a single animal [the white arrows represent the direction of the applied MPGs]. (B) The mean D* during arterial pulsation and diastole respectively within the three ROIs for MPGs applied parallel (black line) and perpendicular to (grey dashed line) PVS orientation. (C) The mean D* at baseline and after dobutamine (black line) or vehicle (grey dashed line) within the same ROIs (non-gated).
![](https://iiif.elifesciences.org/lax/34028%2Felife-34028-fig4-figsupp1-v2.tif/full/617,/0/default.jpg)
(A) The individual animal D* during arterial pulsation and diastole respectively within the three ROIs for MPGs applied parallel (black line) and perpendicular to (grey dashed line) PVS orientation. (B) The mean D* at baseline and after dobutamine (black line) or vehicle (grey dashed line) within the same ROIs (non-gated). Each line represents an individual animal.
![](https://iiif.elifesciences.org/lax/34028%2Felife-34028-resp-fig1-v2.tif/full/617,/0/default.jpg)
Comparison to contrast enhanced MRI following cisterna magna injection of gadolinium
A. b0 image at the ventral aspect of the rat brain captured using the noninvasive approach with an ultra-long TE. B. T1 weighted image at the matched ventral slice at baseline, prior to intra-cisternal gadolinium infusion. T1 weighted images 12 minutes (C) and 24 minutes (D) after intra-cisternal gadolinium infusion.
Additional files
-
Transparent reporting form
- https://doi.org/10.7554/eLife.34028.010