Dagnino, G., Mattos, L., Caldwell, D.
Published in International Journal of Computer Assisted Radiology and Surgery, vol.10(2), pp. 217-229, February 2015
DOI 10.1007/s11548-014-1078-9

Abstract:
Purpose: Introduce and evaluate a new vision-based system for fast and accurate laser scanning control during robot-assisted phonomicrosurgery. The system aims at improving surgical quality by merging the benefits of open-loop laser control (e.g. high-speed scanning capabilities) with the robustness and accuracy of closed-loop visual servoing systems.
Methods: The required fast and accurate laser scanning characteristics are accomplished with a dual control strategy, which adds a vision-based trajectory correction phase to a fast open-loop laser controller. Evaluation of the new system is performed through real CO2 laser cutting trials on artificial targets and ex-vivo tissue.
Results: Experimental results demonstrate the proposed system is able to eliminate open-loop aiming errors caused by system calibration limitations and by the unpredictable topology of real targets. Accuracy values corresponding to pixel resolution are achieved even when smoke created by the laser-target interaction clutters the camera view. In realistic test scenarios, trajectory following RMS errors are shown to be reduced by almost 80% with respect to open-loop system performances, reaching mean error values around 30μm and maximum observed errors in the order of 60μm.
Conclusions: The observed improvements in laser aiming performance lead to the conclusion that the new vision-based laser control system can have a large positive impact on the safety and quality of laser microsurgeries.