Quantitative multimodal anisotropy imaging enables automated fibrosis assessment of H&E-stained tissue

Non-alcoholic steatohepatitis (NASH) disease severity is graded by Clinical Research Network (CRN) histologic scoring using two differently stained tissue sections per tissue block. To visualize and grade steatosis, inflammation, and ballooning, one tissue section must be stained with hematoxylin and eosin (H&E) and, to stage fibrosis, the other tissue section must be stained with Masson’s Trichrome (MT). The necessity for two differently-stained tissue samples to assess each patient at each time point is cumbersome, requires more tissue and resources than a single section, and may introduce interpretability challenges due to variability in staining and intra-biopsy sample heterogeneity.

Here, we report the development of quantitative multimodal anisotropy imaging (QMAI) which can highlight structured substances like collagen in tissue and apply this to visualize fibrosis in H&E-stained liver tissue sections from patients with NASH. QMAI performance was evaluated by comparison to MT-stain color-based fibrosis quantification.