Reproducibility and validity of knee joint position estimation using markerless motion capture

Abstract

The purpose of this study was to examine the reproducibility and validity of using a markerless motion capture system to measure femorotibial angle (FTA). The experimental units included 36 legs of 18 healthy young adult males (mean age 20.4 ± 0.6 years). We measured FTA using markerless motion capture and optical motion capture. Reproducibility was assessed by the intraclass correlation coefficient using the testretest method, and validity was evaluated using the Pearson correlation coefficient. The results of the analysis indicated that FTA measured using markerless motion capture and optical motion capture had high reproducibility（ICC： 0.74，0.97）. Additionally, significant strong positive correlations were found between FTA measured using optical motion capture and markerless motion capture（r=0.74，p ＜0.01）. These results indicate that FTA measured by markerless motion capture has excellent reproducibility and high validity as an estimation of knee joint position in the frontal plane.