In vivo diffusion MRI of optic pathways in 18 healthy mice

The dataset contains manganese enhanced T1- and diffusion-weighted images of optic pathways in 18 healthy mice. All raw (Bruker) and a subset of processed (NIfTI) files are stored.

Animals
Our study was approved by the Institutional Animal Care and Use Committee (IACUC) at NYU Langone Health. All n=18 mice were females, genotype C57BL/6, 7-8 weeks old at baseline and sacrificed within 24 hours after completing the imaging.
 
Data
We acquired in vivo MRI using a preclinical wide bore 7T scanner with a four-channel phased array cryogenically cooled receive-only coil (Bruker BioSpin MRI GmbH, Ettlingen, Germany). Prior to imaging, the mice were anesthetized with inhaled isoflurane (3% induction, 1.0-1.5% maintenance). The acquisition protocol consisted of FLASH T1-weighted images at 100µm isotropic resolution (TE/TR = 4.5/30ms) and multishell DWIs at 200µm isotropic resolution (multi-shot multi-slice EPI with 3 readout segments, TE/TR = 27.6/2781ms, L/R phase encoding) sampled in 60 directions distributed on radial lines at b ∈ {250, 1000, 2250, 4000} [s/mm2] interleaved with 5 images at b=0. 

To minimize motion artifacts, we fixed the mice to the holder with a bite bar and a pair of ear bars. Additionally, we used a TTL (Transistor-Transistor Logic) trigger receiving inputs from a pressure pad to account for breathing-related motion artifacts. The expected scan time for the dMRI protocol was 37min. 24sec. The effective scan time with the trigger varied between 60 and 70 minutes.

Immediately after the baseline scans, while the mice remained under isoflurane-induced general anesthesia, we administered intravitreally manganese chloride (MnCl2) solution to the left (nleft=8) or the right eye (nright=10). We began by injecting 1.0µl of 0.1M MnCl2 (Mice 1–2), which we eventually lowered to 0.5µl (Mice 3–18) without any observable loss in contrast enhancement. Finally, 21±4 hours after injection, we repeated the exact same FLASH T1 acquisition to obtain follow-up MEMRI scans. 

Image processing
In our DWI postprocessing pipeline, we executed the MRtrix3 implementations of the Marchenko-Pastur PCA denoising (dwidenoise), Gibbs ringing (mrdegibbs) and B1 field inhomogeneity  artifacts removal (dwibiascorrect ants). After that, we interpolated the DWIs to 100µm isotropic resolution such that they would match the voxel size of the anatomical images. Then, we registered the follow-up MEMRI scans to the corresponding b=0 images using 3-dimensional rigid body transformation implemented in: antsRegistrationSyN.sh 

Acknowledgements
The authors would like to thank Orin Mishkit, Zakia Ben Youss Gironda, Orlando Aristizabal, and Youssef Wadghiri from the NYU Langone Health Preclinical Imaging Laboratory for their invaluable help.

This project was supported in part by the National Institutes of Health (NIH, R01 EB028774, R01 EB029306, and R01 NS082436). It was performed under the rubric of the Center for Advanced Imaging Innovation and Research (CAI2R, https://www.cai2r.net), a NIBIB Biomedical Technology Resource Center (NIH P41 EB017183), and at the NYU Langone Health Preclinical Imaging Laboratory, a shared resource partially supported by the NIH/SIG 1S10OD018337-01, the Laura and Isaac Perlmutter Cancer Center Support Grant NIH/NCI 5P30CA016087.