The validation of computed tomography derived radiodensity measurements of bone healing using histopathology

Jack Stewart Davey, Randi Rotne, Glenn Edwards

Research output: Contribution to journalArticlepeer-review

1 Citation (Scopus)

Abstract

Background: The early identification of delayed bone healing or a non-union is vital for prompt treatment and superior patient outcomes. Current techniques rely heavily on operator skill for interpretation and hence their reliability and repeatability may be inconsistent. This study assessed the application of computed tomography (CT) derived densiometric measurements as a quantitative tool for the assessment of bone healing. Methods: This prospective, longitudinal, method comparison study was performed using a recognised sheep tibial ostectomy model. Secondary bone healing was assessed at 2, 4, 6, and 8 weeks after the ostectomy was performed. CT densiometric measures of bone healing (Hounsfield units) were taken of the cis, trans, cranial and caudal cortices relative to the bone plate, with histological measurements (percentage of ossification) sourced from the same areas. Cis cortical densiometric data points were excluded from analysis due to significant beam hardening artefact from the bone plate (P < 0.001). A univariable linear regression was performed on the remaining data using averaged radiodensity (independent variable) and histomorphometric (dependent variable) measurements. Results: The two measurements were significantly correlated (R2 = 0.623, P = 0.020) with a clear positive trend identified. Conclusion: This study suggests that radiodensity measurements may be a useful diagnostic and management tool for the monitoring of indirect bone healing.

Original languageEnglish
Article number111543
Number of pages6
JournalInjury
Volume55
Issue number6
DOIs
Publication statusPublished - Jun 2024

Fingerprint

Dive into the research topics of 'The validation of computed tomography derived radiodensity measurements of bone healing using histopathology'. Together they form a unique fingerprint.

Cite this