Several species have disapeared over the last centuries. Some of them disapeared because of human activity or climatic changes. Famous example of extinct species linked to human activity are Steller's sea cow or Raphus cucullatus (Dodo). More recently, Thylacine, Quagga, Mexican grizzly, Javan tiger of Japanese Sea Lion are some of the numerous examples that are now classified as extinct species.

Fortunately, several museums like the Museum National d'Histoire Naturelle in France or the British Museum in England kept some of these recently disappeared specimens in their fluid collection. As a part of the Brain Catalogue project¹, we intend to provide to the community high quality MRI data sets of these extinct, rare and endangered specimens.

Despite the evident usefulness of getting high-resolution MRI of extinct species, this would give a new onset in neuroanatomy to understand how variability is expressed during brain evolution.

Here we show an example of one of the achievement of the Brain Catalogue project. We provide high quality MRI data set of the Thylacinus Cynocephalus and some examples of metrics that were derived from tools that will be available online with the Brain Catalogue.

The specimen used within this study was identified as a female Thylacine. This specimen was registered in Paris' zoo since the 9^th of February 1891. Thylacine's brain was kept in fluid after death.

MRI was performed with a 11.7T Biospec 117/16 (Bruker, Germany), equipped with a 72-mm transceiver. A 3D Flash sequence was used to acquire a 75 µm isotropic anatomical data set. Parameters were: FOV: 5.76*5.28*4.8 cm; Mtx: 768*704*640; TR/TE=25/9.2 ms; Nex=4; Bandwidth=50kHz; Scan Time : 12.5 hours. Raw data were saved and magnitude images were reconstructed using an in-house software running on Matlab (Mathworks, USA). The specimen was kept in his vial during the whole experiment.

For image processing, the brain was aligned to the sagittal/axial and coronal planes using an affine transformation and trilinear interpolation. The volume was cropped to isolate the brain from the container. Image intensity gradients were corrected using the N4 algorithm. Brain tissue was segmented using a combination of thresholding, topology-preserving mathematical morphology operators and manual edition using StereotaxicRAMON. (http://dx.doi.org/10.6084/m9.figshare.963147). We used the surface nets algorithm to build a mesh of the external surface of the brain². The topology of the resulting mesh was corrected to ensure manifoldness using a combination of automatic and manual topology correction (http://github.com/r03ert0/meshgeometry, http://github.com/r03ert0/MeshSurgery). The mesh was decimated using the edge collapse algorithm implemented in the VCG library, and smoothed using Taubin's lambda-mu algorithm³. Thus, brain volume, cortical surface, absolute folding index and cortical thickness were derived from the processed images.

Figure 1 shows some slices of the Thylacine's brain. Numerous dark spot are clearly indentified within these images and are linked to conservation issues.

A surface mesh reconstruction is presented in figure 2 and shows the overall shape of the Thylacine's brain.

A Brain volume of ~ 17500 mm³ was estimated, brain surface was estimated as ~ 5800 mm². An absolute folding index of 1.88 (i.e.1.88-times more surface than a sphere of the same volume) was derived from MRI data. Cortical thickness of gyri and sulci were estimated to 2.7 mm and 1.5 mm respectively.

The aim of the Brain Catalogue is to develop a Citizen Science platform to provide access to this unique dataset of high-quality data to collaboratively process, segment, reconstruct, label and in general study the evolution and phylogeny of the vertebrate brain. One of the major objective of this project was to provide MRI of rare specimens to the community. It is indeed sometimes forbidden to work on certain specimens, to get fresh tissue or specimens could have simply disappeared from earth's surface. Here the example of Thylacine shows the outstanding possibilities of such projects of collaborative Science. We hope that other Museum would open door and let access to some of their very unique specimens.