Histology-informed microstructural diffusion simulations for MRI cancer characterisation-the Histo-μSim framework.

Area of Science:

• Biomedical Imaging
• Computational Biology
• Oncology

Background:

• Diffusion Magnetic Resonance Imaging (dMRI) simulations traditionally focus on brain imaging.
• There is a need for novel non-invasive biomarkers in body cancer imaging.
• Microscopic tissue complexity is key for developing advanced dMRI biomarkers.

Purpose of the Study:

• To introduce Histo-μSim, a Monte Carlo simulation framework for histology-informed dMRI in body cancer imaging.
• To generate synthetic dMRI signals from virtual cancer environments linked to tissue properties.
• To enable data-driven estimation of microstructural properties like diffusivity and cell size.

Main Methods:

• Reconstruction of virtual cancer environments from human liver biopsy stains (hematoxylin-eosin).
• Generation of synthetic dMRI signal dictionaries coupled with tissue properties.
• Comparison of Histo-μSim metrics with analytical multi-compartment models in silico, ex vivo, and in vivo.

Main Results:

• Histo-μSim is feasible for clinical settings.
• The framework provides metrics that more accurately reflect histology compared to analytical models.
• Histo-μSim successfully estimated properties like extracellular diffusivity, cell size, and membrane permeability.

Conclusions:

• Histo-μSim offers histologically-meaningful tissue descriptors for dMRI.
• This approach can increase the specificity of dMRI for cancer detection.
• Histo-μSim has the potential to significantly contribute to precision oncology.