Related Experiment Video
Updated: Aug 28, 2025
06:09
Measuring 3D In-vivo Shoulder Kinematics using Biplanar Videoradiography
Published on: March 12, 2021
3.2K
Reliability of a human pose tracking algorithm for measuring upper limb joints: comparison with photography-based
Jingyuan Fan1, Fanbin Gu1, Lulu Lv1
1Department of Microsurgery, Orthopedic Trauma and Hand Surgery, The First Affiliated Hospital, Sun Yat-Sen University, Guangzhou, 510080, China.
BMC Musculoskeletal Disorders
|September 21, 2022
Summary
Pose estimation using RGB images reliably measures upper limb range of motion, offering a convenient alternative to traditional goniometry for joint angle assessment.
Area of Science:
- Biomedical Engineering
- Kinesiology
- Computer Vision
Background:
- Range of motion (ROM) is crucial for diagnosing upper extremity conditions.
- Clinical goniometry is standard but time-consuming and requires expertise.
- Automated angle measurement from RGB images offers a potential alternative.
Purpose of the Study:
- To evaluate the reliability of an automated pose tracking algorithm for measuring upper limb joint angles.
- To compare algorithm-based measurements with results from human raters.
Main Methods:
- Thirty healthy adults performed six upper limb movements (shoulder, elbow, wrist).
- Movements were captured via digital cameras and analyzed using the OpenPose algorithm.
- Algorithm measurements were compared to surgeon measurements using mean differences, Pearson correlation, and intra-class correlation coefficients.
Main Results:
- Mean differences were under 3 degrees for most motions, except wrist flexion.
- Intra-class correlation coefficients exceeded 0.60, indicating good reliability.
- High correlations (p < 0.001) were found between algorithm and manual measurements.
Conclusions:
- Pose estimation is a reliable method for measuring shoulder and elbow joint angles from RGB images.
- This technology supports the use of digital cameras for joint ROM assessment.
- Patients may be able to self-assess their ROM using photos from digital cameras.
