Automated Tissue Strain Calculations Using Harris Corner Detection

Jake Elliott, Sujata Khandare, Ali A. Butt, Molly Smallcomb, Meghan E. Vidt, Julianna C. Simon

Research output: Contribution to journalArticlepeer-review

5 Scopus citations

Abstract

The elastic modulus, or slope of the stress-strain curve, is an important metric for evaluating tissue functionality, particularly for load-bearing tissues such as tendon. The applied force can be tracked directly from a mechanical testing system and converted to stress using the tissue cross-sectional area; however, strain can only be calculated in post-processing by tracking tissue displacement from video collected during mechanical testing. Manual tracking of Verhoeff stain lines pre-marked on the tissue is time-consuming and highly dependent upon the user. This paper details the development and testing of an automated processing method for strain calculations using Harris corner detection. The automated and manual methods were compared in a dataset consisting of 97 rat tendons (48 Achilles tendons, 49 supraspinatus tendons), divided into ten subgroups for evaluating the effects of different therapies on tendon mechanical properties. The comparison showed that average percent differences between the approaches were 0.89% and -2.10% for Achilles and supraspinatus tendons, respectively. The automated approach reduced processing time by 83% and produced similar results to the manual method when comparing the different subgroups. This automated approach to track tissue displacements and calculate elastic modulus improves post-processing time while simultaneously minimizing user dependency.

Original languageEnglish (US)
Pages (from-to)564-574
Number of pages11
JournalAnnals of Biomedical Engineering
Volume50
Issue number5
DOIs
StatePublished - May 2022

All Science Journal Classification (ASJC) codes

  • Biomedical Engineering

Fingerprint

Dive into the research topics of 'Automated Tissue Strain Calculations Using Harris Corner Detection'. Together they form a unique fingerprint.

Cite this