Three-dimensional preoperative planning in the weight-bearing state: validation and clinical evaluation

Hooman Esfandiari,Tabitha Roth,Fabio Carrillo,Philipp Fürnstahl,Sandro F Fucentese,Reto Sutter,Giulia Ceschi,Lazaros Vlachopoulos,Wolfgang Wein,Matthias Wieczorek

doi:10.1186/s13244-021-00994-8

Abstract

Objectives3D preoperative planning of lower limb osteotomies has become increasingly important in light of modern surgical technologies. However, 3D models are usually reconstructed from Computed Tomography data acquired in a non-weight-bearing posture and thus neglecting the positional variations introduced by weight-bearing. We developed a registration and planning pipeline that allows for 3D preoperative planning and subsequent 3D assessment of anatomical deformities in weight-bearing conditions.MethodsAn intensity-based algorithm was used to register CT scans with long-leg standing radiographs and subsequently transform patient-specific 3D models into a weight-bearing state. 3D measurement methods for the mechanical axis as well as the joint line convergence angle were developed. The pipeline was validated using a leg phantom. Furthermore, we evaluated our methods clinically by applying it to the radiological data from 59 patients.ResultsThe registration accuracy was evaluated in 3D and showed a maximum translational and rotational error of 1.1 mm (mediolateral direction) and 1.2° (superior-inferior axis). Clinical evaluation proved feasibility on real patient data and resulted in significant differences for 3D measurements when the effects of weight-bearing were considered. Mean differences were 2.1 ± 1.7° and 2.0 ± 1.6° for the mechanical axis and the joint line convergence angle, respectively. 37.3 and 40.7% of the patients had differences of 2° or more in the mechanical axis or joint line convergence angle between weight-bearing and non-weight-bearing states.ConclusionsOur presented approach provides a clinically feasible approach to preoperatively fuse 2D weight-bearing and 3D non-weight-bearing data in order to optimize the surgical correction.

Highlights

Malalignment of the lower limb often causes a shift of the load-bearing axis away from the center of the knee, leading to an imbalance in load distribution within the knee joint
While progressed degeneration requires surgical treatment through unicompartmental or total knee replacement (TKR) [5], a joint-preserving realignment surgery achieved through a corrective osteotomy is the benchmark in younger patients with unilateral OA [6]
We have recently shown that the relevant radiological metrics in the context of preoperative planning, such as the mechanical axis (MA) and the joint line convergence angle (JLCA), are significantly different between 2D weight-bearing and 2D non-weight-bearing as well as between 2D non-weight-bearing and 3D non-weightbearing conditions [12]

Summary

Introduction

Malalignment of the lower limb often causes a shift of the load-bearing axis away from the center of the knee, leading to an imbalance in load distribution within the knee joint. While progressed degeneration requires surgical treatment through unicompartmental or total knee replacement (TKR) [5], a joint-preserving realignment surgery achieved through a corrective osteotomy is the benchmark in younger patients with unilateral OA [6]. In a corrective osteotomy intervention, the pathologically deformed bones are cut, realigned and subsequently fixed with an orthopedic implant, transferring the load from the pathological to the healthy compartment of the knee. Precise postoperative alignment to a neutral weight-bearing axis or even mild over-corrections have been found to result in improved functional scores [7]. Accurate preoperative planning and radiographic measurements are crucial

Objectives

Methods

Results

Discussion

Conclusion