Multi Image Resolution Diffusion MRI Dataset

Ex Vivo Monkey Brain: Multi Image Resolution Diffusion Mri Dataset

dig monkeyThe multi image resolution diffusion MRI data set is obtained from one long scanning session (19 days) of a perfusion fixated Vervet monkey brain using a 4.7 T preclinical Agilent MRI scanner as described in Dyrby et al 2014, Neuroimage

The original work used the multi image resolution MRI dataset to invistigate the effect of interpolating the raw diffussion MRI data to a higher image resolution before fibre reconstruction (i.e. DTI) (see figure below when 7th order B-spline is used (figure 5 in Dyrby et al 2014)). Results of using various interpolation methods are investiagted and compared with data acquired in the true high image resolution. Indeed inteprolation of raw diffusion MRI dataset do reveal finer details but the choice of interpolation method matters and many challenges remains such as inteprolation artefacts - please read the paper it is free for dowload.

Here we have made the high quality ex vivo multi image resolution diffusion MRI datasets from Dyrby et al 2014 freely available so that you can try out your own interpolation or super-resolution methodologies as well as reconstruction methods. 

Histology of matched Vervet monkey brain can be found here.

Facts about the dataset:

The dataset include a slap of a one hemisphere coverage of axial slices with isotropic voxel sizes of 13, 0.83, 0.63, 0.53 and 0.43 mm3, (no gap) and a high resolution coronal dataset with 0.25×0.25 mm2 in plane resolution and slice thickness of 0.5 mm and 2 mm gaps. The number of repetitions was 1, 1, 2, 4, 8 and 8 respectively. The repeated magnitude images were averaged off line.

All datasets were acquired with the parameters TR = 5500 ms (but 3200 ms for the 0.25x0.25 mm2 data set), TE = 47 ms, and each repetition included 3 non-diffusion weighted and 61 uniformly distributed diffusion weighted directions. A b-value of 4056 s/mm2 was selected (gradient strength G = 180 mT/m, pulse width δ = 8 ms and time between pulse onset Δ = 30 ms) as found optimal for ex vivo imaging in Dyrby et al. (2011). For more details please see Dyrby et al., 2014.

Please cite the two following works in when using this dataset:

1) Describing the mullti image resolution diffusion MRI dataset: 
Dyrby TB, Lundell H, Burke MW, Reislev NL, Paulson OB, Ptito M, Siebner HR, Interpolation of diffusion weighted imaging datasets, NeuroImage, 103, 202–213, 2014 

2) Describing how the data was setup and acquired:  
Dyrby TB, Baaré WF, Alexander DC, Jelsing J, Garde E, Søgaard LV, An ex vivo imaging pipeline for producing high-quality and high-resolution diffusion weighted imaging datasets, Human Brain mapping, in-press, April, 2011

Download the multi image-resolution diffusion MRI data set and its schemefiles here 

Observed problems or artifacts: None 


Lundbeck Foundation and The EU founded CONNECT consortium under the FET program