MATLAB-Based Algorithm and Software for Analysis of Wavy Collagen Fibers.

Area of Science:

• Biomedical Engineering
• Materials Science
• Computational Biology

Background:

• Accurate characterization of soft tissue fiber structure is crucial for understanding tissue physiology and pathology.
• Existing methods struggle to reliably separate fiber waviness from global orientation, especially in complex biological samples.

Purpose of the Study:

• To develop and validate an automatic algorithm for simultaneously estimating fiber global orientation, abundance, and waviness in 2D images.
• To provide a reliable tool for biological tissue characterization by overcoming limitations of previous methods.

Main Methods:

• An algorithm based on incremental movement of local regions of interest (ROI) was developed.
• Pixels belonging to fibers were identified, and ROI movement was determined by local fiber orientation.
• Validation was performed using artificial images and porcine trachea samples, with manual fiber tracking for verification.

Main Results:

• The algorithm accurately estimates fiber waviness, curve length, and global orientation simultaneously.
• Accuracy for curve length was below 1%, with high R2 values for waviness (0.92) and global orientation (0.96).
• The software processes one fiber in approximately 1.1 seconds on a standard computer.

Conclusions:

• The developed algorithm reliably and accurately quantifies soft tissue fiber waviness and orientation.
• This method offers a significant advancement for biological tissue characterization and mechanical modeling.
• The ability to separate waviness from orientation is a key feature for analyzing complex fiber structures.