Summary

An individualised lower limb model is created using integration of MRI and Gait lab studies to assess PFJ reaction force

Abstract

Introduction

Patellar dislocation is characterised by a complete loss of contact between the articular surface of the femoral trochlea and that of the patella, and is a common injury in the skeletally immature population, with an estimated incidence rate of 43 per 100,000 [1] in children. To add to this problem, children and adolescents are likely to experience recurrent patellar instability following acute first-time dislocation, with a recurrence rate ranging from 38.4% to 91% [2]. It is therefore critical to determine the underlying mechanisms of recurrent patellofemoral joint (PFJ) instability in skeletally immature patients to best manage the injury. Patellar dislocation is collectively described as an abnormal interplay of lower limb alignment, bony geometry of the trochlea and the patella, passive restraints of ligaments and retinaculum and the action of the quadriceps. Nevertheless, the interplay of risk factors that predispose individuals to recurrent patellar dislocation remain poorly understood and cannot be gathered from traditional bio-statistical methods. There is a need to establish methods that can ‘unpack’ the relative contributions of bony geometry, bony alignment and soft tissue restraints for each patient to best inform clinical decision making. The broad aim of this project is to develop and validate personalised paediatric lower limb musculoskeletal models to investigate the aetiology of patellar dislocation injury in children and adolescents. The specific aim addressed in this abstract is creation and validation of the personalised models.

Methods

In an ongoing prospective study, typically developing participants and patients who present with recurrent patellofemoral instability are referred for Medical imaging (MRI) and 3D gait analysis. MRI acquisitions consist of full lower limb and knee scans at four knee flexion angles. For each participant, 3D lower limb bones, knee ligamentous structures (patellar tendon, ACL, PCL and MCL) and articular knee cartilage are reconstructed using Mimics (Materialise, Leuven). Reconstructed 3D images inform creation of a personalised knee joint mechanism for each patient, whereby passive motion of the tibiofemoral joint (TFJ) and patellofemoral joint (PFJ) are modelled as a ‘five rigid link parallel mechanism’ and a hinge joint, respectively, with surface contact conditions and ligament length constancy [3]. MRI-measured parameters are adjusted by using two optimization algorithms that avoid mechanism singularities and best match kinematic patterns from published experimental studies [4]. Validation is performed by comparing experimental TFJ and PFJ passive kinematics at the nominal and three different TFJ flexion angles. The personalised knee joint mechanism is incorporated into a personalised full lower limb OpenSim MSK model.

Results

AND DISCUSSION
Preliminary results from the rigid kinematic model were obtained for a recurrent patellar dislocator (female, age: 9yrs) and a control participant (female, age: 11yrs). The patellar dislocator exhibited higher external patellar rotation, and lateral and superior patellar displacement through early TFJ flexion range of motion (i.e. 0-30 degrees), which confirmed the presence abnormal patellar tracking of lateral patellar tilt [4]. Furthermore, the optimised model showed good agreement to the experimental imaging data at different knee flexion angles.

Conclusions

This paper presents the first passive kinematic model of the TFJ and PFJ, informed from bony geometry and soft tissue restraints, to be successfully implemented into a fully personalised lower limb musculoskeletal model of a child with patellofemoral instability. In clinical practice this method has the potential to improve surgical planning by identifying individualized risk factors that (ii) reduce patellofemoral joint stability and (ii) increase patellofemoral joint contact forces during gait.