Surface extraction can provide a reference for micro-CT analysis of retrieved total knee implants

Abstract

Background: Quantitative measurements of damage and wear in orthopaedic components retrieved from patients during revision surgery can provide valuable information. However, to perform these measurements there needs to be an estimate of the original, unworn geometry of the component, often requiring multiple scans of the various sizes of components that have been retrieved. The objective of this study was to determine whether the articular and backside surfaces could be independently segmented from a micro-CT reconstruction of a tibial insert, such that a tibial insert of one thickness could be used as a reference for a tibial insert of a different thickness.

Methods: New tibial inserts of a single width but with six different thicknesses were obtained and scanned with micro-CT. An automated method was developed to computationally segment the articular and backside surfaces of the components. Variability between intact and extracted components was determined.

Results: The deviations between the comparisons of the extracted surfaces (range, 0.0004 to 0.010 mm) were less (p<0.001) than the baseline deviation between the intact surfaces (range, 0.0002 to 0.053 mm).

Conclusions: An extracted surface from one insert thickness could be used to accurately represent the surface of an insert of a different thickness. This greatly enhances the feasibility of performing retrieval studies using micro-CT as a quantitative tool, by reducing the costs and time associated with acquiring, scanning, and reconstructing multiple reference tibial insert geometries.

Clinical relevance: This will add greater detail to studies of retrieved implants, to better establish how implants are functioning in vivo.

Keywords: Micro-computed tomography; Polyethylene wear; Retrieval analysis; Total knee arthroplasty.