Evidence for microscopic kurtosis in neural tissue revealed by correlation tensor MRI

Three sample double diffusion encoding (DDE) datasets for three female rat brains acquired using a 9.4T Bruker Biospec scanner equipped with an 86 mm quadrature transmission coil and four-element array reception cryocoil (Rat1_invivo_cti_data.nii, Rat2_invivo_cti_data.nii, Rat3_invivo_cti_data.nii).

All animal experiments for the collection of these datasets were preapproved by the institutional and national authorities and carried out according to European Directive 2010/63.

DDE data were acquired for 636 pairs of DDE b-values (bvals1.bval and bvals2.bval) and gradient directions (bvecs1.bvec and bvecs2.bvec) according to the DDE protocol for Correlation Tensor MRI (CTI) fitting described by Henriques et al. (Magn Reson Med 2021, doi:10.1002/mrm.28938).

For each rat brain dataset, thermal noise from each cryocoil channel was suppressed using the threshold-based PCA denoising algorithm (Henriques et al., bioRxiv 2023, doi:10.1101/2023.03.29.534707). The denoised data was subsequently corrected for Gibbs ringing using the sub-voxel shift algorithm (Kellner et al. Magn Reson Med 2016, doi: 10.1002/mrm.26054) which is implemented in DIPY (Garyfallidis et al., Frontiers in Neuroinformatics 2014, doi: 10.3389/fninf.2014.00008).  The processed diffusion-weighted signals from the four channels were then combined using sum-of-squares. Combined data for different gradient direction pairs were then aligned along the different b-values and directions using a sub-pixel registration technique (Guizar-Sicairos et al., Opt Lett. 2008, doi: 10.1364/ol.33.000156).

Full description of these datasets acquisition and preprocessing can be found in its original manuscript (Henriques et al., Magn Reson Med 2021, doi:10.1002/mrm.28938).