Related Experiment Video
Updated: Jun 29, 2026
08:27
Novel Triple-Loop Technique for Suturing TFCC Injuries without Transosseous Tunnel
Published on: May 23, 2025
Identifying scapholunate ligamentous injury
Frederick W Werner1, Haoyu Wang, Walter H Short
1Department of Orthopedic Surgery, SUNY Upstate Medical University, 3214 Institute for Human Performance, 505 Irving Avenue, Syracuse, NY 13210, USA. wernerf@upstate.edu
Summary
This study developed a noninvasive tool using neural networks to predict wrist ligament injuries, achieving 93% accuracy in detecting damage and identifying specific injured ligament groups.
Area of Science:
- Orthopedics
- Biomechanics
- Medical Imaging
Background:
- Scapholunate instability often involves injury to the scapholunate interosseous ligament and secondary stabilizers.
- Accurate diagnosis of these ligament injuries is crucial for effective treatment.
- Current diagnostic methods may be invasive or lack precision in identifying specific ligament damage.
Purpose of the Study:
- To develop a noninvasive clinical tool for predicting scapholunate interosseous ligament and secondary ligament injuries.
- To determine which specific ligaments or ligament groups are injured in cases of suspected scapholunate instability.
- To validate the predictive capabilities of neural network models based on kinematic and 3D measurements.
Main Methods:
- Utilized kinematic and 3D measurements from 62 cadaver wrists in a motion simulator.
- Developed neural network models based on angular changes in scaphoid and lunate motion.
- Developed alternative models based on changes in scaphoid-lunate distance and 3D gap measurements.
- Tested models with simulated ligament sectioning to represent instability.
Main Results:
- Angular data-based models predicted intact ligaments with 93% accuracy.
- These models identified sectioned dorsal or volar ligaments with 84% accuracy.
- Optimal prediction performance was observed with the wrist in 10-30 degrees of flexion.
- High prediction rates, sensitivity, specificity, and kappa values demonstrated model viability.
Conclusions:
- A noninvasive, CT-based predictive model for wrist ligament injuries is viable.
- Neural network analysis of kinematic data shows promise for diagnosing scapholunate instability.
- The developed tool can accurately predict the presence and location of ligamentous injuries in the wrist.