Reliability and validity of pose estimation algorithm for measurement of knee range of motion after total knee arthroplasty

Abstract

AimsWe aimed to assess the reliability and validity of OpenPose, a posture estimation algorithm, for measurement of knee range of motion after total knee arthroplasty (TKA), in comparison to radiography and goniometry.MethodsIn this prospective observational study, we analyzed 35 primary TKAs (24 patients) for knee osteoarthritis. We measured the knee angles in flexion and extension using OpenPose, radiography, and goniometry. We assessed the test-retest reliability of each method using intraclass correlation coefficient (1,1). We evaluated the ability to estimate other measurement values from the OpenPose value using linear regression analysis. We used intraclass correlation coefficients (2,1) and Bland–Altman analyses to evaluate the agreement and error between radiography and the other measurements.ResultsOpenPose had excellent test-retest reliability (intraclass correlation coefficient (1,1) = 1.000). The R2 of all regression models indicated large correlations (0.747 to 0.927). In the flexion position, the intraclass correlation coefficients (2,1) of OpenPose indicated excellent agreement (0.953) with radiography. In the extension position, the intraclass correlation coefficients (2,1) indicated good agreement of OpenPose and radiography (0.815) and moderate agreement of goniometry with radiography (0.593). OpenPose had no systematic error in the flexion position, and a 2.3° fixed error in the extension position, compared to radiography.ConclusionOpenPose is a reliable and valid tool for measuring flexion and extension positions after TKA. It has better accuracy than goniometry, especially in the extension position. Accurate measurement values can be obtained with low error, high reproducibility, and no contact, independent of the examiner’s skills.Cite this article: Bone Joint Res 2023;12(5):313–320.

Biomechanical Posture Analysis in Healthy Adults with Machine Learning: Applicability and Reliability

“…For example, Ota et al [58] assessed the reliability using a motion capture system during bilateral squat movements and found ICC scores ranging between 0.92 and 0.96. Saiki et al [38] conducted two studies on the reliability of their method in patients with knee osteoarthritis. They initially demonstrated excellent test-retest reliability (ICC (1, 1) = 1.000) and substan-tial agreement with radiography (ICC (2, 1) = 0.915) for measuring the hip-knee-ankle angle.…”

“…The significant potential of these ML approaches makes their transition into clinical settings crucial, especially considering the limited number of studies that have attempted this to date [19,38]. Therefore, the aim of the current study is two-fold.…”