THE ROLE OF PERSONALIZED PRESSURE MODELLING IN KNEE OSTEOTOMY PLANNING

Knee osteotomy is a surgical procedure that realigns the lower limb, aiming to relieve pressure on the overloaded cartilage and improve knee joint function. While this concept relies on load redistribution, direct in vivo measurement of knee joint contact forces remains challenging. To address this limitation, computational musculoskeletal modeling offers a non-invasive solution for estimating joint kinetics. This study introduces a personalized discrete element analysis (DEA) framework to calculate and compare pre- and postoperative intra-articular pressure distributions. Lower limb weight-bearing computed tomography (WBCT) were obtained from 5 patients (mean age = 39.4 years, SD = 14.36), scheduled for high tibial osteotomy. 3D osseous models were created using semi-automated statistical shape model-based segmentation. The cartilage layers were defined based on point-specific thickness maps, derived from Magnetic Resonance Imaging (MRI) of 53 healthy individuals. The menisci were modeled as triangular shapes with variable height and width, conforming to the femoral condyles. Soft tissue inclusion accuracy was validated using a leave-one-out cross validation on the MRI dataset. Personalized anatomical models were used for tibiofemoral distance analysis and discrete element analysis (DEA). The point-specific joint width was mapped as the shortest, Euclidean distance between the tibial and femoral bones. The DEA model utilized linear springs to represent cartilage-bone and meniscus-cartilage-bone interactions. A joint reaction force equal to half the body weight was applied to simulate weight-bearing conditions. Cartilage predictions showed a maximal error for patellar cartilage prediction with a median Root Mean Square Error (RMSE) of 0.75 mm. With the exception of the anterior cruciate ligament, presenting a median RMSE of 1.16 mm, all other structures exhibited a sub-millimetric accuracy. In figure 1, the pre- and postoperative configurations of a single subject were compared using a tibiofemoral distance map. We developed and validated a methodological workflow for patient-specific, morphological knee joint modeling, serving as an input for both tibiofemoral distance analysis and DEA. This methodology shows potential as an innovative tool to guide and improve high tibial osteotomy surgery. Further investigation with larger patient cohorts is warranted to validate the clinical applicability of this approach in routine surgical practice. For any figures or tables, please contact the authors directly.