Online Program Overview Session: SG-04

Content

Brain atlases are of utmost importance for any brain imaging approach as soon as a certain structure, a ROI, has to be structurally identified. Until recently anatomical brain atlases have been derived from two dimensional histological sections, which were labelled by experts e.g. the very well-known and highly appreciated atlases for mouse and rat by Paxinos et al.. These histological atlases provide microscopic, i.e. cellular resolution, which was and still is of importance to sufficiently delineate the different substructures in the brain. At this cellular level histochemistry provided specific marker like Nissl staining or ACh-esterase staining serving as basic information for brain parcellation. This approach goes back to the times of Broadmans’ parcellation of the human brain. However, all histological atlases suffer, even due to technical advances, to a different degree from cutting and preparation/staining artefacts and most importantly, the full three dimensional context is lost. Several approaches have been worked out to cope with these constrains. Methodologically inherent in 3D imaging techniques like MRI, SPECT, PET or CT is the 3D context and the reduced, if not even artefact-free, data quality, but unfortunately with a lower spatial resolution. Technical improvements particular in the MRI field lead to much higher resolutions (groups of Johnson (Duke) and Henkelmann (Toronto) in the order of ten’s of microns. Based on these resolutions and improved imaging contrasts MRI based fully 3D brain atlases became available. Of note, after 3D imaging the brain in the skull most of the classical histological atlas generation approaches can be applied in addition. Therefore, multimodal brain atlases were generated combining the best from the two worlds like the Waxholm datasets as only one example. At this point, another methodological progress is striking: non-affine registration or warping approaches for building probabilistic atlases. Probabilistic atlases due to averaging across many individual subjects (after appropriate registration) provide much higher SNR in the datasets but even more important allow for a new branch of anatomical analyses, namely voxel-wise group statistics, which were introduced under the term “voxel based morphometry”. This has led to fundamentally new insights into neurobiological research and continues to impact particularly the field of brain pathologies, transgenic mouse research as well as neurodegeneration but also studies of brain plasticity. In addition and somewhat parallel was the usage of brain atlases in the context of functional imaging studies were activation biomarkers like 2DG in autoradiography, 2FDG in PET or fMRI BOLD activation markers were intended to be analyzed in a brain structure specific manner and compared between groups of subjects. To obtain a “match” between functional data, in general at a (much) lower resolution, and the brain atlases used different registration approaches are applied. Here the obtained quality of the “match” is of great impact for the final results. Automated functional analysis pipelines, which are used in a “black-box” manner, starting from the raw data as input and provide statistical differences at the output should be much more scrutinized or better falsified than is currently the case. Further mention should be made of the fact that more and more high resolution atlases are being made available for single brain structures / areas like the cerebral cortex (Amunts/Zilles group) or the rat hippocampus (Bjaalie group). Of note, from more and more species (digital) brain atlases are provided from insects (drosophila, honey bee), fishes, Gerbil to monkeys using state of the art imaging technologies (see e.g. Bakkers https://scalablebrainatlas.incf.org/ for an overview at lease for mammalian species). It is noteworthy that almost all these atlases are freely accessible to the scientific community (cf. http://www.nitrc.org). Finally and very recently additional data entities led to an enormous boost in dimensionality of digital brain atlases to name only a few topics: connectivity by diffusion tensor imaging (MRI DTI) or trace injections as Allen Brain Atlas, OMICS approaches like ViBrism and again the ABA or development (emap and ABA). Another new endeavor is to perform new multimodal parcellation schemes e.g. incorporate functional data like resting state fMRI data to obtain parcellations of brain structures at a much higher level fusing anatomical/histological and functional information. As a conclusion, it should be noted that the data on digital brain atlases, especially in the mouse through the flagship project Allen brain atlas, already allows in-silico studies on the anatomy and function of the brain. This is very effective, is cost-saving, reduces the number of laboratory animals and allows completely new research questions (cf. Ganglberger et al., NI 2017).

Content

Registration to a standard brain atlas is the prerequisite for modern analysis of neuroimaging data, e.g. connectomics or voxel-based group statistics. Tools and pipelines for human neuroimaging data are quite prevalent but common agreements and workflows in the small animal imaging community are lacking. To date, the Allen brain atlas is the most comprehensive database of mouse brain anatomy and connectivity. We have developed MATLAB toolboxes¹ for registration of mouse MR and histology images to this atlas based on the freely available SPMMouse² and elastix³ packages. Examples will be given of how atlas registration can help to generate hypotheses in basic research of functional recovery after stroke in the mouse based on techniques that correlate data from behavioral testing with imaging data, e.g. voxel-based lesion symptom mapping⁴. To stimulate a discussion, some personal insight will be given into technical pitfalls of image registration, advantages and drawbacks of the Allen database, and challenges when using atlas registration-based techniques in models of disease.

References

1. Koch S et al. Atlas registration for edema-corrected MRI lesion volume in mouse stroke models. J Cereb Blood Flow Metab. 2017;0271678X1772663. Epub ahead of print.

2. Sawiak SJ et al. SPMMouse: A New Toolbox for SPM in the Animal Brain. Proc Int’l Soc Mag Res Med. 2009. p. 1086.

3. Klein S et al. elastix: a toolbox for intensity-based medical image registration. IEEE Trans Med Imaging. 2010;29(1):196–205.

4. Bates E et al. Voxel-based lesion-symptom mapping. Nat Neurosci. 2003;6(5):448–450.

Acknowledgement

This work was supported by the Federal Ministry of Education and Research (BMBF) (grant number 01EO0801, Center for Stroke Research Berlin) and Deutsche Forschungsgemeinschaft (DFG) (Excellence Cluster NeuroCure).

Content

The cerebral vasculature has a key supporting role by supplying the brain with oxygen, nutrients and removing brain metabolites. Any aberrations in this vascular system can have severe consequences on normal brain functioning. As seen in stroke, occlusion of vessels can lead to ischemia and eventually neuronal death. Furthermore, structural vascular abnormalities have been observed in many neuropathological diseases such as Alzheimer’s disease, Huntington’s disease, multiple sclerosis and brain tumors[1]. Therefore, it is essential to have knowledge of the whole brain’s vascular architecture.

Recent advances in preclinical imaging methods allow detection of whole brain vasculature on different imaging scales. Ex vivo histological imaging techniques allow the visualization of whole brain vasculature at resolutions < 3 µm. As the high resolution datasets are generally large, the processing of the data is time consuming. The datasets, however, allow precise investigation of the topology of the micro-vasculature and provide information on regional capillary densities[2]–[4]. In contrast to histological imaging, non-invasive in vivo imaging techniques have the advantage to assess the development of macro-vasculature and venous sinuses within the same subject over time. This is essential in studies investigating the effect of progressing neuropathology on cerebral vasculature [5].

Vascular imaging data usually have a high specificity allowing for extraction of the vasculature after noise reduction and filtering. Further processing can be undertaken to extract information on vascular branching and vascular density. Additionally, vascular imaging data can be co-registered to the Allen Brain Atlas, which gives access to and allows comparison with anatomical, genetic and connectivity information.

References

[1]        C. C. V Chen, Y. C. Chen, H. Y. Hsiao, C. Chang, and Y. Chern, “Neurovascular abnormalities in brain disorders: Highlights with angiogenesis and magnetic resonance imaging studies,” J. Biomed. Sci., vol. 20, no. 1, p. 1, 2013.

[2]        A. Paolo et al., “Whole-Brain Vasculature Reconstruction At the Single Capillary Level,” pp. 1–25, 2017.

[3]        B. Xiong et al., “Precise Cerebral Vascular Atlas in Stereotaxic Coordinates of Whole Mouse Brain,” Front. Neuroanat., vol. 11, no. December, pp. 1–17, 2017.

[4]        S. Xue et al., “Indian-ink perfusion based method for reconstructing continuous vascular networks in whole mouse brain,” PLoS One, vol. 9, no. 1, pp. 1–7, 2014.

[5]        N. Beckmann et al., “Age-dependent cerebrovascular abnormalities and blood flow disturbances in APP23 mice modeling Alzheimer’s disease.,” J. Neurosci., vol. 23, no. 24, pp. 8453–8459, 2003.

Acknowledgement

This talk was made possible by the Bio-imaging lab and Laboratory of Cell Biology and Histology - University of Antwerp and was supported by Molecular Imaging of Brain Pathophysiology (BRAINPATH) under grant agreement number 612360 within the Marie Curie Actions-Industry-Academia Partnerships and Pathways (IAPP) program, by the Fund of Scientific Research Flanders (FWO G048917N), Flagship ERA-NET (FLAG-ERA) FUSIMICE (grant agreement G.0D7651N) and by the European Union’s Seventh Framework Programme (FP7/2007–2013) under grant agreement number 278850 (INMiND).