Comparison of ACL strain estimated via a data-driven model with in vitro measurements

Abstract

Computer modeling and simulation techniques have been increasingly used to investigate anterior cruciate ligament (ACL) loading during dynamic activities in an attempt to improve our understanding of injury mechanisms and development of injury prevention programs. However, the accuracy of many of these models remains unknown and thus the purpose of this study was to compare estimates of ACL strain from a previously developed three-dimensional, data-driven model with those obtained via in vitro measurements. ACL strain was measured as the knee was cycled from approximately 10° to 120° of flexion at 20 deg s⁻¹ with static loads of 100, 50, and 50 N applied to the quadriceps, biceps femoris and medial hamstrings (semimembranosus and semitendinosus) tendons, respectively. A two segment, five-degree-of-freedom musculoskeletal knee model was then scaled to match the cadaver’s anthropometry and in silico ACL strains were then determined based on the knee joint kinematics and moments of force. Maximum and minimum ACL strains estimated in silico were within 0.2 and 0.42% of that measured in vitro, respectively. Additionally, the model estimated ACL strain with a bias (mean difference) of −0.03% and dynamic accuracy (rms error) of 0.36% across the flexion-extension cycle. These preliminary results suggest that the proposed model was capable of estimating ACL strains during a simple flexion-extension cycle. Future studies should validate the model under more dynamic conditions with variable muscle loading. This model could then be used to estimate ACL strains during dynamic sporting activities where ACL injuries are more common.

Keywords:

• Knee
• ACL
• injury
• computer simulation

Acknowledgement

The authors wish to acknowledge the funding provided by the Multidisciplinary Seed Grant from the Office of Research at Old Dominion University, Norfolk, VA. The authors also thank Julie Choisne, Bobbie Irmischer, Megan Houston and Michael Samaan for their assistance with data collection and analysis.