Automated Multi-Scale Morphometry Analysis of Fibrous Networks.

Fibrous materials are ubiquitous in modern technology and essential to diverse industrial applications spanning medicine, energy storage, advanced textiles, and structural composites. Quantitative morphometric analysis of these fibrous structures from microscopy images, however, remains challenging due to the complexity of discriminating the elongated and intertwined fibers, particularly in heterogeneous materials with varying fiber diameters and particulate matter. Traditional image analysis methods often fail to accurately characterize these mixed fiber-particle networks, and manual measurement approaches remain subjective, time-consuming, and prone to operator-dependent variability. We present a novel automated framework, the interactive Fiber Analyzer GUI, with free access to any interested parties, for multiscale morphometry analysis that integrates several key innovations: (1) a robust dual-pipeline discriminator that uses a filter-based approach for pure fiber networks and morphological filtering with aspect-ratio thresholding to effectively segregate components in mixed-morphology systems; and (2) Adaptive Contour-based Diameter Estimation (ACDE), utilizing active contour models for precise fiber diameter measurement at multiple skeletal points. The integrated pipeline demonstrates superior performance, achieving efficient implementation times (19-25 s per ROI) and significantly reduced error rates across validation samples (mean error 7.17% ± 1.29%, range 5.13-8.57%) for both pure fiber and complex mixed-morphology samples when combined with empirically optimized correction factors that compensate for systematic measurement bias in automated analysis. This framework advances automated morphometric analysis by providing accurate, reproducible quantification of complex fibrous networks.