Abstract
Introduction
3D printed Patient Specific Guides (PSGs) can improve the accuracy of joint-replacement. Pre-operative CT bone models are used to design a PSG that fits the patient's specific bone geometry. A key design requirement is to maximize docking robustness such that the PSG can maintain a stable position in the planned location. However, current PSG designs are typically manually defined, lack a quantitative assessment of robustness, and have an unknown consistency of docking rigidity between patients. Limited research exists on the stability and robustness of surgical guides, and no software packages are available to facilitate this analysis. Our goal was to develop such a software.
Methods
In this paper, the contact between a patient's bone and the PSG is modelled using robotic grasping theory, and its docking robustness is quantified by analysis of the PSG's grasp wrench space (GWS) (i.e. the combination of contact forces and torques between the bone and PSG). To this end, a PSG design and analysis tool with a graphical user interface was developed in MATLAB. This tool allows the user to load shapes (e.g. STL bone models), select and manipulate possible contact points, and optimize the contact point locations according to the largest-minimum resisted wrench (LRW) that the grasp can resist in any direction. The LRW is a grasp quality metric equivalent to the radius of the largest (hyper)sphere contained within the convex hull of the GWS, and its value can be evaluated using frame-variant GWS calculations (i.e. centroid-dependent) or frame-invariant GWS calculations (i.e. centroid-independent).
Results
Multiple 2D and 3D shapes were loaded into the software and contact points were selected to form a ‘grasp’ and compute their wrench spaces. For a square with four contact points, the frame-variant LRW is calculated to be 0.4240 and the frame-invariant LRW is calculated to be 0.4999. These values are expected to increase after contact point optimization, with a higher value indicating higher docking robustness.
A realistic contact set for a shoulder arthroplasty PSG was created by modelling the glenoid of a patient's scapula and selecting seven contact points. For this configuration, a frame-variant LRW was calculated to be 0.0034 and a frame-invariant LRW is 0.0181.
Discussion
To date we have developed a software capable of using robotic grasping theory to analyse and validate the robustness of existing surgical guides. A set of contact points similar to those used in clinical PSGs produced a much smaller quality metric value compared to the ideal grasp for a simple shape. This quality metric value is constrained by the wrench component with the smallest value and it is clear that producing a PSG design with a robust ‘grasp’ (i.e. high docking rigidity) is non-trivial. In the future, the result of this software will be compared to experimental results to validate its predictions. Once validated, design optimization capabilities will be implemented that can significantly improve the PSG docking rigidity.
For any figures or tables, please contact authors directly.
