Summary

A novel digital design framework for creating patient-specific high tibial osteotomy plate fixations using clinically relevant diagnostic tools such as densitometric calibration of CT scans and subject-specific biomechanics in combination with the artificial-intelligence driven Generative Design.

Abstract

Background

Osteoarthritis (OA) is the leading source of chronic pain and disability, affecting approximately 15% of the world population with a socioeconomic burden costing up to 2.5% of gross domestic product. High Tibial Osteotomy (HTO) is an effective joint preserving treatment for unicompartmental OA of the knee, offloading the mechanically overloaded compartment of the knee to relieve pain and optimise joint movement. Hinge fractures, soft-tissue irritation, loss of correction, delayed union and stress shielding are reported amongst main complications, all of which can be attributed to the ‘one-size-fits-all’ concept of HTO plate. Recently, additive manufacturing and patient-specific devices have begun to address the limitations of the ‘one-size-fits-all’ approach in HTO. Subject-specific finite element (FE) models created from densitometric calibration of CT scans have emerged to be an indispensable tool in biomechanical research. The authors have developed a novel digital design framework to produce patient-specific HTO plates from a CT scan, integrating densitometric calibration, FE analysis and the artificial-intelligence driven Generative Design (GD).

Methods

CT scan of a cadaveric left tibia was used in this study for virtual simulation of HTO, guided by an experienced knee surgeon. Following the anatomical reconstruction, a densitometric calibration was performed to apply element-specific heterogenous bone material properties, producing a subject-specific FE model of the cadaveric tibia. A biomechanical model of the HTO construct with the gold standard plate was developed and validated through FE modelling, to attain plate design requirements. Subsequently, clinical and biomechanical requirements were replicated within the GD domain to explore titanium plates for HTO. The generatively designed, patient-specific plate choices produced were comprehensively analysed under physiological load conditions, to attest their biomechanical performance and thus healing efficacy.

Results

The integrated elements of the novel design workflow from densitometric calibration, and surgical planning through to GD HTO plate exploration are demonstrated. The study also highlights the significance of bone material mapping in HTO by comparing FE models with generic material properties from literature vs subject-specific modelling, appraising the disparity in post-operative biomechanical outcomes. Three novel patient-specific HTO plates are illustrated, and their biomechanical efficacy is validated under partial weight-bearing and walking.

Discussion/Conclusion
Unlike conventional HTO plates, GD allows simultaneous integration of patient factors, surgical planning, and patient-specific biomechanics, with the aim to reduce plate stiffness and profile on soft-tissue. Subject-specific FE modelling captured the shape and the heterogenous material properties of the native tibia, enhancing the accuracy of the subject-specific HTO biomechanics. Such subject-specific FE models can be used for identifying weaker anatomical regions, allowing the optimisation of screw configuration and corresponding plate designs. This personalised approach to HTO offers improved outcomes, accounting for patient variability, especially when considering osteoporotic patients. The novel approach to HTO increases patient-specificity of plates by creating a custom fit conforming to both patient anatomy and biomechanics, minimising plate size and stiffness whilst maintaining construct stability, and thus improving mechanobiological performance and post-operative outcomes.