Development of a surgical navigation system for corrective osteotomy based on augmented reality

Abstract

This study develops a novel navigation system for corrective osteotomy surgery based on augmented reality. By combining camera calibration and optical tracking technologies, we can provide a superimposed view of virtual 3D models with the resection plan onto the real patient, whereby both positional and rotational discrepancies between the surgical plan and current cutting plane are displayed. To test the feasibility of the developed system, the quantitative accuracy assessments were performed by using a custommade grid phantom and a mockup model. The assessment using the custom-made grid phantom showed stable root mean square error (RMSE) between virtual points and corresponding real points regardless of the viewing angle (1.31 +/- 0.29 mm; mean +/- s.d.) and the phantom-to-camera distance (1.17 +/- 0.31 mm; mean +/- s.d.). The assessment using the mockup model with five human subjects showed small positional and rotational discrepancies between the cutting plan and real cutting plane (0.59 +/- 0.39 mm and 1.31A degrees +/- 0.38A degrees; mean +/- s.d.).