Integration of multimodal neuroanatomical data of gray short-tailed opossum
Piotr Majka (Nencki Institute of Experimental Biology), Natalia Chlodzinska (Nencki Institute of Experimental Biology), Tomasz Banasik (H. Niewodniczański Institute of Nuclear Physics PAN), Ruzanna L. Djavadian (Nencki Institute of Experimental Biology), Władysław P. Węglarz (H. Niewodniczański Institute of Nuclear Physics PAN), Krzysztof Turlejski (Nencki Institute of Experimental Biology), Daniel K. Wójcik (Nencki Institute of Experimental Biology)
Digital brain atlases are very different from their traditional book predecessors. Conventional 2D atlases are based on collections of delineated microphotographs of series of stained brain slices. Digital three-dimensional brain atlases allow unconstrained navigation through brain volume and reslicing at arbitrary angles. Furthermore, combining data from different specimens or modalities, localizing and analyzing data within the context of brain volume e.g. performing morphometric analysis or assessing intergroup variances quantitatively, is far more natural in the 3D context.
The aim of this study is to integrate multimodal neuroanatomical data of gray short-tailed opossum. The data include: (1) MRI images collected using Bruker BioSpec 9.4T imaging system: T2 weighted, 100um isotropic in vivo and T1/T2* weighted, 50um isotropic taken after 48 hours and 30 days post mortem. (2) Photographs of the tissue block taken during cryosectioning of the brains ('blockface' images). (3) Microphotographs of slices stained with histological methods including Nissl and Acetylocholinesterase.
The collected data were processed to merge different modalities. In this process the blockface volume, as an intermediate modality between MRI and histology, was the key element. In the workflow, the first step is to reconstruct and segment blockface images. Then, to eliminate global deformations related to brain fixation and its extraction from skull, in vivo and ex vivo MRI images were brought into blockface volume. Afterwards, images of stained sections and corresponding blockface images were nonlinearly registered eliminating slice-specific deformations due to staining procedures. These steps fulfilled the task of converging multimodal data into a single volumetric template. The integrated data will be used for identification and delineation of anatomical structures eventually forming a digital atlas.
The project is partly supported by an infrastructural grant from the Polish Ministry of Regional Development POIG.02.03.00-00-003/09.