An image-based approach for advanced statistical quantification of architectural parameters and permeability in porous biological media.

Abstract

Load-bearing soft tissues-such as the meniscus, articular cartilage, and intervertebral discs-exhibit a multilayered architecture within their extracellular matrix (ECM). Variations in structural organization across these layers give rise to graded physical properties, which are critical for the tissues to perform their diverse biomechanical functions. In this study, we introduce an image-based unified framework for the statistical quantification and spatial evolution of architectural/topological, including Minkowski Functionals (MFs), porosity, pore connectivity, pore size, tortuosity, and permeability through Pore Network Modeling (PNM) of multilayered ECMs, enabling a deeper understanding of the interplay between structure and function. We demonstrate the effectiveness of our approach using two image datasets-VOI(1) and VOI(2)-obtained from micro-CT scans of meniscal tissue. We observe correlations between the evolution of spatial permeability within VOI(1) and VOI(2) and tortuosity, throat lengths, complex connectivity, and pore pressure zones. Despite VOI(2) containing a higher number of pores and throats, its more complex and tortuous pore network results in convoluted fluid pathways that reduce transport efficiency. Consequently, VOI(2) exhibits lower permeability compared to the simpler and more direct pore architecture of VOI(1). The proposed methodology enables image stack datasets to be used for rapid and advanced characterization of architecture, permeability, and pressure fields within a single, unified code. Matlab codes and two datasets of image stack of meniscal tissue are available at GitHub (https://github.com/olgaBARRERA/Biologic-porous-media-characterisation) and at ( 10.5281/zenodo.14018501).