Journal: bioRxiv
Article Title: Label-free quantitative 3D mapping of collagen architecture by holotomography
doi: 10.64898/2026.05.05.722893
Figure Lengend Snippet: HT reconstructs volumetric RI distributions of fibrillar collagen without exogenous labels, enabling automated fiber-level quantification. a , Representative 3D RI tomogram (axial extent ∼30 μm) visualized as orthogonal slices, volumetric rendering, maximum-intensity projection (MIP), and depth-colored MIP, resolving continuous fibrillar architecture. b , Co-registered HT and SHG images with corresponding line profiles, confirming spatial correspondence of individual fibrils. c , Volumetric RI renderings of separately prepared type I and type III collagen networks, highlighting distinct fibrillar organization. d , Quantitative analysis pipeline: HT construction followed by curvelet transform-based fiber segmentation (CT-FIRE) applied to representative axial sections, generating fiber masks and centerlines. e , Representative concentration-matched (0.8 mg/mL) sections of type I and type III collagen with extracted fibers and effective-width-weighted masks used for per-fiber quantification. f , RI probability density istributions for fiber-associated and background regions, showing distinct higher-RI populations corresponding to fibrillar collagen. g , Concentration-dependent distributions of effective fiber width and per-fiber dry mass for type I and type III collagen, demonstrating systematic shifts with increasing concentration and subtype-specific differences.
Article Snippet: Type I collagen (rat tail, 4 mg/mL; Advanced BioMatrix, Cat. No. 5153) and type III collagen (human, 1 mg/mL; Advanced BioMatrix, Cat. No. 5021) were kept on ice before use.
Techniques: Concentration Assay