I am a PhD student studying diffusion MRI. One topic in the diffusion MRI community is to estimate the brain cytoarchitecture in vivo using the measured diffusion MRI signals. My thesis focuses on estimating the neuronal cell body/soma volume fraction distribution in the human brain, especially in the gray matter. The volume fraction is defined as the ratio between the volume of somas and the volume of the brain voxel. In diffusion MRI, the voxel size is around 1mm^3. One problem I am facing is the missing of a good reference for the distribution of the soma volume fraction over the human brain cortex. I think BigBrain could serve as a gold standard for this, thanks to the effort of the entire BigBrain team.
I download the volume data at 400um resolution from LORIS. If I understand correctly, the value in a voxel of the volume data represents the staining intensity in a 400^3 um^3 brain voxel. High staining intensity (dark coloring) means more or large somas in that brain voxel. Can I further interpret the staining intensity as an indicator of the soma volume fraction? Does high staining intensity mean that somas occupy a large portion of the volume in that voxel?
I am looking forward to your response! Thank you very much!
You can certainly interpret the staining intensity as an indicator for soma volume fraction, but I wouldn’t go as far as saying that it is equivalent. There are some other factors that also influence the intensity, such as the presence of blood vessels, and as you already point out, the relationship is likely different depending on the size of somas. If you want to get a more precise understanding of the intensities, I would recommend to compare BigBrain 20 micron intensities with results of cell segmentations from the corresponding 1 micron sections. Some of these high-resolution image data are available, and we have recently released datasets for several cortical brain regions with precise cell segmentations of 1 micron image patches of the BigBrain dataset, such as this one: EBRAINS - Knowledge Graph Search (a full list is here: EBRAINS - Knowledge Graph Search). You might want to have a look at the data descriptor and pdf reports (“report_all.pdf”) in those datasets.
There was a presentation at the OHBM conference in Glasgow a few weeks ago on this matter, using the above mentioned datasets. You find the abstract here: OHBM