Development of a patient-specific FE model of the knee to study cartilage degeneration following ACL reconstruction
A sensitivity analysis to evaluate the effects of cartilage and meniscal Young's moduli on cartilage degeneration
More Info
expand_more
Abstract
Osteoarthritis (OA) is a progressive joint degeneration disease resulting in joint pain, stiffness, and loss of mobility. Among patients with anterior cruciate ligament reconstruction (ACLR), OA incidence is increased. Articular cartilage (AC) degeneration is irreversible, therefore prevention and early detection are essential. Specifically, the possibility to predict the risk of AC degeneration would provide options for patient-specific early interventions.
This thesis presents a Finite Element (FE) workflow for a patient-specific FE knee model of a patient with ACLR. A degeneration algorithm is implemented to predict AC degeneration. Uncertainties in the parameters of a FE workflow can affect the model's outcome. The sensitivity of the workflow to the AC and meniscal Young's moduli should be determined, as these influence the stress distribution in the joint but in-vivo measurement is not possible for patient-specific values. Therefore, the developed FE knee model is used to answer the main research question: how sensitive is the predicted AC degeneration of a patient-specific Finite Element ACL reconstructed knee model to changes in the AC and meniscal Young's moduli?
The FE knee model was created based on patient-specific MRI and gait analysis data. For the sensitivity analysis, 24 models were simulated with AC and meniscal Young's moduli varying between 5 MPa and 35 MPa, and 59 MPa and 80 MPa, respectively. A degeneration algorithm was implemented for the calculation of AC degeneration, based on max principal stresses.
Minimal AC degeneration was calculated for both the tibial AC and the femoral AC, ranging from no degeneration to a degeneration level of 0.9812, and 0.9694, respectively. The 5-year follow-up MRI showed no AC degeneration either. Statistical analysis was performed with the volume of degenerated AC. A multiple regression analysis showed an exponential relationship between the degenerated volume and the AC Young's modulus for the tibial AC (R2 = 0.995, p<0.05), and the femoral AC (R2 = 0.989, p<0.05). The meniscal Young's modulus did not affect the degenerated volume. The sensitivity analysis demonstrated that increasing AC Young's modulus resulted in an exponentially larger volume of degeneration.
In conclusion, the established FE workflow showed promise for the calculation of AC degeneration following ACLR. A foundation was laid for future development of the model. The sensitivity of the workflow to the AC Young's modulus was determined, highlighting the need for patient-specific estimation of the AC Young's modulus for reliable patient-specific AC degeneration results.