Anterior cruciate ligament (ACL) injury is quite common among young athletes, with the number of injury cases exceeding 120,000 annually in the United States alone. Over 70% of which account for non-contact injuries. Forces and moments acting on the knee joint play essential roles in these injuries. Motions of the other body segments are effective in increasing or decreasing these loads. In this study, the effect of whole-body (WB) kinematics on the knee joint biomechanics was investigated using in vivo and in silico methods.
Motion analysis experiments were done on 14 able-bodied young participants wearing a full-body marker set and performing two variations of single-leg landings, using their left and right limbs (4 tasks). Marker trajectories and force plate data were recorded from the in vivo experiments of these participants.
The in silico investigations consisted of two separate parts. First, musculoskeletal simulations were done to obtain whole-body kinematics, kinetics, and muscle forces, using inverse kinematics, inverse dynamics, and static optimization techniques. The next part was non-linear dynamic finite element (FE) analyses. A FE dynamic/explicit knee model was developed from medical images of a healthy young female and validated against in vitro experiments for the knee joints kinematics and ligaments strains. Ligaments’ material properties for the knee cruciate and collateral ligaments were obtained through optimization to the experimental tensile test data in the literature. Then, the participants’ data from musculoskeletal simulations were used as the input to the FE analyses. The FE outputs included ACL strain, knee joint contact forces, contact pressures, and soft tissue stresses.
In order to find the relationship between WB kinematics and knee joint biomechanics, correlation analysis was used. Using Spearman correlation coefficients and P-values, the correlation between WB modifiable parameters and knee biomechanics along with their statistical significance were determined. Receiver operating characteristic analysis (ROC) and logistic regression were also used to predict high ACL strain cases for each of the four tasks. All the input kinematics, kinetics, and muscle forces were entered to the ROC, and logistic regression was able to predict high ACL strain trails from a few input parameters.
Parameters of the WB kinematics that were both detrimental to ACL strain and knee contact forces and stresses included increased stance knee abduction, increased stance knee internal tibial rotation, increased lumbar extension, decreased pelvis forward tilt, increased stance hip abduction, increased stance hip internal/external rotation, reduced stance ankle plantarflexion. Therefore, preventive training programs should focus on modifying these motions. Also, kinetic parameters that should be reduced because of their destructive effects on the ACL loading and knee contact forces and stresses were stance knee anterior shear force, stance knee superior force, and ground reaction forces.
Muscle forces that were indicators of high ACL strains included higher rectus femoris force, higher vastus medialis force, higher vastus intermedius force, and higher vastus lateralis force. Thus, muscles acting in the opposite direction like hamstrings should be strengthened. While muscle forces that protected the ACL were higher medial gastrocnemius force, higher lateral gastrocnemius force, higher semimembranosus force, higher semitendinosus force, and higher left and right internal and external abdominal oblique muscle forces. Medial and lateral gastrocnemius muscle forces were also inversely correlated with peak medial tibial cartilage (MTC) contact force, and peak lateral tibial cartilage (LTC) contact force and von Mises stress. Therefore, intervention exercises should aim at increasing the strength of these muscles during single-leg landings.
Finally, the most influential parameters for each task were amplified incrementally in four cases (one female and one male for each of the two single-leg landing tasks) to investigate their effects on ACL strain and knee contact forces by inducing motions that were not safe for the participants to perform in vivo. Increases of up to 2.41%, 2811N, and 5843N were seen in ACL strain and contact force on the lateral and medial tibial cartilages, respectively. Exaggerating the input kinematics led to the injury in three of the four analyzed cases due to the extreme forces generated in the medial component of the knee, which caused the contact forces on the medial tibial cartilage to exceed the failure values reported in the literature. Kinematic values that produced these failure cases included peak lumbar rotation of 8.2о toward the stance limb, peak hip extension of 36.3о, and peak hip internal rotation of 37.8о. This work is one of the few to investigate high-risk motions during dynamic single-leg landings for simulating the real injury scenarios that cannot be tested using live subjects.