Predicting the mechanical hip–knee–ankle angle accurately from standard knee radiographs: a cross-validation experiment in 100 patients

Authors

  • Willem Paul Gielis Department of Orthopedic Surgery, UMC Utrecht, Utrecht, The Netherlands
  • Hassan Rayegan Department of Mechanical Engineering, Faculty of Engineering, University of Birjand, Birjand, Iran
  • Vahid Arbabi Department of Orthopedic Surgery, UMC Utrecht, Utrecht, The Netherlands; Department of Mechanical Engineering, Faculty of Engineering, University of Birjand, Birjand, Iran; Department of Biomechanical Engineering, Faculty of Mechanical, Maritime, and Materials Engineering, Delft University of Technology (TU Delft), Delft, The Netherlands
  • Seyed Y Ahmadi Brooghani Department of Mechanical Engineering, Faculty of Engineering, University of Birjand, Birjand, Iran
  • Claudia Lindner Division of Informatics, Imaging & Data Sciences, University of Manchester, Manchester, UK
  • Tim F Cootes Division of Informatics, Imaging & Data Sciences, University of Manchester, Manchester, UK
  • Pim A de Jong Department of Radiology, UMC Utrecht and Utrecht University, Utrecht, The Netherlands
  • H Weinans Department of Orthopedic Surgery, UMC Utrecht, Utrecht, The Netherlands
  • Roel J H Custers Department of Orthopedic Surgery, UMC Utrecht, Utrecht, The Netherlands

DOI:

https://doi.org/10.1080/17453674.2020.1779516

Abstract

Background and purpose — Being able to predict the hip–knee–ankle angle (HKAA) from standard knee radiographs allows studies on malalignment in cohorts lacking full-limb radiography. We aimed to develop an automated image analysis pipeline to measure the femoro-tibial angle (FTA) from standard knee radiographs and test various FTA definitions to predict the HKAA.

Patients and methods — We included 110 pairs of standard knee and full-limb radiographs. Automatic search algorithms found anatomic landmarks on standard knee radiographs. Based on these landmarks, the FTA was automatically calculated according to 9 different definitions (6 described in the literature and 3 newly developed). Pearson and intra-class correlation coefficient [ICC]) were determined between the FTA and HKAA as measured on full-limb radiographs. Subsequently, the top 4 FTA definitions were used to predict the HKAA in a 5-fold cross-validation setting.

Results — Across all pairs of images, the Pearson correlations between FTA and HKAA ranged between 0.83 and 0.90. The ICC values from 0.83 to 0.90. In the crossvalidation experiments to predict the HKAA, these values decreased only minimally. The mean absolute error for the best method to predict the HKAA from standard knee radiographs was 1.8° (SD 1.3).

Interpretation — We showed that the HKAA can be automatically predicted from standard knee radiographs with fair accuracy and high correlation compared with the true HKAA. Therefore, this method enables research of the relationship between malalignment and knee pathology in large (epidemiological) studies lacking full-limb radiography.

Downloads

Download data is not yet available.

Downloads

Published

2020-06-22

How to Cite

Gielis, W. P. ., Rayegan, H. ., Arbabi, V. ., Brooghani, S. Y. A., Lindner, C. ., Cootes, T. F. ., … Custers, R. J. H. (2020). Predicting the mechanical hip–knee–ankle angle accurately from standard knee radiographs: a cross-validation experiment in 100 patients. Acta Orthopaedica, 91(6), 732–737. https://doi.org/10.1080/17453674.2020.1779516