Validating MRI-based in-vivo histology

• Applicant: Dr. Siawoosh Mohammadi
• Subject Area: Human Cognitive and Systems Neuroscience
• Term: since 2017
• Project identifier: Deutsche Forschungsgemeinschaft (DFG) - Project number 313644856

The established technique to measure microstructural properties in the human brain is ex-vivo histology, which, due to the limited tissue sample size (of microscopic volume) and its destructive, single measurement nature, is impractical for measuring progressive white matter changes across the entire brain. This greatly limits our ability to investigate the pathophysiological origin of aging and neurodegeneration. Advanced MRI technologies summarized under the term “quantitative MRI” (qMRI) can measure in-vivo changes associated with white-matter microstructure. In-vivo histology using MRI (hMRI) combines quantitative MRI (qMRI) with mathematical biophysical models to quantify microscopic tissue properties of the living human brain, such as myelin and axonal volume fractions, and the relative myelination of fiber pathways measured by the MR g-ratio. To do so, they have to solve the ill-posed problem of extracting metrics at the scale of microns from MRI voxels at millimeter scale and thus need thorough validation before application. Today, existing validation attempts have yielded insufficient insight because they were performed only qualitatively, or because the histological counterpart was measured in small tissue sections that covered only a fraction of an MRI voxel and thus was not representative of the hMRI metric extracted from an ensemble-average MR signal. Moreover, the comparison is typically based on ex-vivo MRI of fixed tissue that will differ from fresh tissue, e.g., due to cross-linking of proteins, water loss, or self-digestion (autolysis). The goal of the running project is to develop a pipeline that addresses the aforementioned limitations of existing validation attempts for hMRI models of the MR g-ratio and it’s constituents, i.e. myelin and axon volume fractions. To do so, we have acquired a worldwide unique multi-modal dataset, including quantitative multi-parameter mapping and multi-shell diffusion MRI data of 6 human post-mortem brains across the entire fixation period (with more than 300 time points), and additional high-resolution data, as well as ex vivo histology data (incl. electron and light microscopy) of the same brains. The current application for an extra year on this project aims at maximising the added value of these acquired data. An extra year for this Emmy Noether grant will facilitate the continuation of the unique validation pipeline developed in this project far beyond the grant duration because it will make the data openly available for forthcoming research projects. The availability of this data will support translation into the clinical setting because the published data allow to directly test pathology-specific modification of the biophysical models. Particularly its inherent white-matter myelin density and axon density measures have the potential to become important surrogate markers.

• DFG Programme: Independent Junior Research Groups