# Concurrent Validity of Three Photogrammetric Methods for Assessing Knee Alignment in Sagittal Plane

## TL;DR

This study tested three photogrammetric methods for measuring knee alignment in the sagittal plane and found that correcting for thigh and leg rotations improves accuracy.

## Contribution

The study introduces and validates three photogrammetric methods for knee alignment assessment with corrections for rotational deviations.

## Key findings

- Correcting for thigh and leg rotations significantly improved validity metrics for all photogrammetric methods.
- The corrected condylar angle (CAcorr) and corrected four-point angle (4PAcorr) showed the best performance with high correlation and low RMSE.
- Proportional errors in measurements suggest the need for better tools to assess rotational changes in lower limbs.

## Abstract

Background: Evidence supporting the validity of photogrammetry for assessing body segment alignment remains limited, with most studies focusing on spinal evaluation. Thus, there is a lack of robust research examining its use for other body segments such as the lower limbs. Objective: This study aimed to evaluate the concurrent validity of three photogrammetric methods for measuring knee alignment in the sagittal plane with and without corrections for potential rotational deviations in the participant’s thigh and leg. Methods: A total of 21 adults underwent sequential evaluations involving panoramic radiography of the lower limbs and photogrammetry at a private radiology clinic. Photogrammetric analysis involved identifying the following anatomical landmarks: the greater trochanter of the femur (GTF), the lateral condyle of the femur (LCF), the head of the fibula (HF), and lateral malleolus (LM). Three photogrammetric methods were employed: (1) the condylar angle (CA) defined by the GTF, LCF, and LM points; (2) the fibula head angle (FHA) defined by the GTF, HF, and LM points; and (3) the four-point angle (4PA) incorporating the GTF, LCF, HF, and LM. Concurrent validity was assessed using correlation analysis, agreement with radiographic measurements, and the root mean square error (RMSE). Each photogrammetric method was tested using raw (CA, FHA, and 4PA) and corrected (CAcorr, FHAcorr, and 4PAcorr) values, accounting for thigh and/or leg rotational deviations. Results: Correcting for thigh and leg rotations significantly improved the validity metrics for all methods. The best performance was observed with the corrected condylar angle (CAcorr: r = 0.746; adjusted r2 = 0.533; RMSE = 2.9°) and the corrected four-point angle (4PAcorr: r = 0.733; adjusted r2 = 0.513; RMSE = 3.0°); however, the measurements presented proportional errors, possible due the method of assessment of rotations. Conclusions: The findings validate the evaluated photogrammetric methods for assessing sagittal knee alignment. Accounting for thigh and leg rotational deviations is critical for achieving accurate measurements, raising the need of accurate tools for measuring rotational changes in the lower limbs to avoid errors.