A Patient-Specific Knee Model Driven by In Vivo Kinematics to Better Represent an ACL-Injured Pediatric Population

Abstract

Purpose: Pediatrics and adolescents are at a higher risk to suffer an anterior cruciate ligament (ACL) injury in comparison to their adult counterparts. As well, the rate of injury is increasing. While some of this increased injury rate may be attributed to increased participation in sport, it remains unknown why only some children suffer an ACL injury. Traditionally, surface marker- based motion-capture would be used to determine this difference; however, due to the presence of soft tissue artifact, marker translation may exceed the physiological range of the knee itself. Using OpenSim, the range of motion of the knee was constrained to allow for soft tissue artifact to be reduced. Therefore, the two objectives of this thesis are 1) to create and validate a new OpenSim knee model, and 2) to use this knee model to determine whether differences in knee kinematics and ligament lengths exist between ACL injured and non-injured control pediatrics. Methods: Manuscript 1 (Chapter 5), focused on the first objective of creating and validating a new OpenSim knee model. Thirty-two healthy pediatric females performed squats, countermovement jumps, and drop-vertical jumps. OpenSim models were made and scaled to each participant and featured a 6-degree-of-freedom knee. Each knee was allowed to move in a manner dictated by published in vivo biplanar fluoroscopy studies. These resultant biplanar fluoroscopy-constrained knee kinematics were then compared to another series of in vivo constraints: bone pin-constrained kinematics. Finally, the length of the four main ligaments of the knee were tracked and compared to existing literature on healthy ligament lengths from extension to deep flexion. Manuscript 2 (Chapter 6) sought to answer the second question: Are there differences in kinematics and ligament lengths between ACL-injured and uninjured control participants? Forty ACL-injured (20 male, 20 female) and 40 uninjured control (20 male, 20 female) participants completed a countermovement jump, which drove inverse kinematics for the OpenSim model created in Manuscript 1. Knee kinematics and ligament lengths were compared between male injured to male control, and female injured were compared to female control. To isolate the effect of injury, males were not compared to females. Results: Starting with Manuscript 1, the squat task had the best agreement between biplanar fluoroscopy and bone pin kinematics, during periods of low knee flexion. At high flexion, the ACL length was shorter than literature data, for all tasks. For Manuscript 2, during the countermovement jump, female control participants obtained greater knee flexion and internal rotation when compared to female ACL-injured participants. For males, there were only small differences in countermovement jump kinematics. For ligament lengths, female controls had longer posterior cruciate ligament lengths, whereas for males, male ACL-injured participants presented with longer ligament lengths for all 4 ligaments. Conclusion: The thesis results identified how to use deep flexion biplanar fluoroscopy constraints to minimize soft tissue artifact. In comparison to previously established knee bone pin constraints, the biplanar fluoroscopy knee relied more on internal-external rotation and anterior-posterior translation to obtain the required bone positions in OpenSim. This additional anterior-posterior translation also led to the increased shortening of the ACL, as the linear distance between ligament insertion points decreased. Next, this thesis identified small differences in kinematics between ACL-injured and control pediatric females during a countermovement jump, where control participants obtained greater knee flexion in the preparatory and landing phase of the task. Few differences in kinematics were found in males. Differences in ligament lengths between injured and control groups were based on differences in knee kinematics, or by participant anthropometrics; taller participants had longer ligaments.