Preventing ACL Injury: Using OpenSim to Find Effective Training Interventions

Last January, the Australian women’s hockey team and the University of Western Australia came together to develop and implement a training program to prevent knee injuries—specifically those to the anterior cruciate ligament or ACL. The regimen was primarily aimed at strengthening the women’s hip and trunk musculature because upper body mechanics seem to affect knee loading during sports activities, according to preliminary simulation research (using OpenSim) by Cyril (Jon) Donnelly, PhD, a lecturer at the University of Western Australia (UWA).


To determine whether the athletes’ training regimen in fact reduces ACL injuries, Donnelly did not just wait to see whether the women would fare better than a control group (as an epidemiologist might). Instead, he and his research group at UWA calculated surrogate measures for the risk of injury—essentially the forces through the knee joint—using the OpenSim modeling framework. To do that, they applied kinematic markers to the athletes’ limbs to mark where their bones are, and using a sophisticated three dimensional motion capture system estimated each athletes’ joint angles and external forces during running, changing direction and landing tasks.


The results should help Donnelly home in on training interventions that protect the knee and ACL.  But there is a challenge: Data-gathering and computational modeling procedures aren’t standardized across labs.  As a result, experiments are extremely difficult to repeat or reproduce. “Depending on the musculoskeletal model you use, for the same data you can get different estimates of the peak forces going through the knee joint,” Donnelly says.


In order for researchers to compare apples to apples, Donnelly along with Scott Delp, PhD, professor of bioengineering at Stanford University, and Mark Robinson, PhD, lecturer in biomechanics at Liverpool John Moores University, partnered to obtain a collaborative grant from the University of Western Australia. The grant will allow the team to implement standardized data collection and modeling procedures and then use them to collect motion-capture data from 20 athletes at each center (in Perth, Australia and Liverpool, United Kingdom).


OpenSim was the modeling framework of choice, Donnelly says, because it is open source can be used to calculate external loading, whole body kinematics, and muscle activation, just as Donnelly’s group did for the Australian hockey team. All of the data and results will be made publicly available, Donnelly says, so that other labs can try to reproduce their findings in OpenSim or even pursue research questions about upper and lower body mechanics that go beyond ACL injury prevention. “There are a lot of great researchers out there who don’t have access to these types of motion data,” Donnelly says. “We hope this open-source data can help bridge this gap.”


As for his ACL research, Donnelly says, the goal is to determine whether injury prevention training programs work and can be reproduced across labs, sports, and continents. Establishing a standard model and collecting data in a standardized way should permit him to run multi-center training interventions (like the Hockey Australia project) with confidence that any observed differences in results derive from the population or the training rather than the model or data collection procedure. 


 “We hope to show that the positive training effects we are achieving here at UWA can be repeated across centers,” Donnelly says. “This is needed if we are to effectively translate injury prevention research into injury prevention practice across Australia and the world.”

All submitted comments are reviewed, so it may be a few days before your comment appears on the site.

Post new comment

The content of this field is kept private and will not be shown publicly.
All submitted comments are reviewed, so it may be a few days before your comment appears on the site. This question is for testing whether you are a human visitor and to prevent automated spam submissions.
Enter the characters shown in the image.