Views on a new surgical assistance method for implanting the glenoid component during total shoulder arthroplasty. Part 2: From three-dimensional reconstruction to augmented reality: Feasibility study - 05/12/18

Doi : 10.1016/j.otsr.2018.08.021

Julien Berhouet ^a,^b,^c,^⁎ , Mohamed Slimane ^b, Maxime Facomprez ^b, Min Jiang ^b, Luc Favard ^a,^c
^a Service d’orthopédie traumatologie 1C, faculté de médecine de Tours, université François-Rabelais de Tours, CHRU Trousseau, avenue de la République, 37170 Chambray-lès-Tours, France
^b Équipe reconnaissance de forme et analyse de l’image, laboratoire d’informatique fondamentale et appliquée de Tours EA6300, école d’ingénieurs polytechnique universitaire de Tours, université François-Rabelais de Tours, 64, avenue Portalis, 37200 Tours, France
^c Société d’orthopédie de L’Ouest (SOO)/HUGORTHO, 18, rue de Bellinière, Trélazé, France

^⁎Corresponding author. Service d’orthopédie traumatologie 1C, université François-Rabelais de Tours, CHRU Trousseau, avenue de la République, 37170 Chambray-lès-Tours, France.Service d’orthopédie traumatologie 1C, université François-Rabelais de Tours, CHRU Trousseauavenue de la RépubliqueChambray-lès-Tours37170France

Sous presse. Épreuves corrigées par l'auteur. Disponible en ligne depuis le Wednesday 05 December 2018
Cet article a été publié dans un numéro de la revue, cliquez ici pour y accéder

Abstract

Introduction

The main goal of this study was to propose a new method of surgical assistance for the implantation of a total shoulder prosthesis, with the use of augmented reality (AR). The advantage of this approach is that it supplements information, on the one hand pre-existing or disappeared due to a pathological process, such as the premorbid glenoid, and on the other hand already existing but not accessible to the surgeon during the procedure, such as the so-called “hidden” face of the scapula.

Material and methods

Several information preparation steps were needed. The first consisted in the three-dimensional (3D) rendering of the pathological glenoid, from a point cloud corresponding to the premorbid glenoid based on previously developed regression equations. A library of “healthy” generic glenoids was then created by hierarchical bottom-up analysis. From this database, a so-called adequate normal generic glenoid was fused and matched to the pathological glenoid reconstructed using a morphing technique. An experimental AR application was constructed. Smart glasses were used to display the prepared 3D information.

Results

A pathological 3D glenoid was reconstructed and used for the AR application. A complete display of the scene, reconstructed glenoid and full scapula was obtained. However, an offset from reality was observed. The main limitations were technical, related to the connected tool itself and the operating software.

Discussion/Conclusion

This was a feasibility study of the different steps required to apply AR, from information preparation to its visualization. A new parameter crossing experiment is needed to optimize each step of this process.