Study on Laser Steering for Robotic Surgery

Abstract

Robot-assisted minimally invasive surgery (RMIS) allows access to internal anatomy through natural orifices or small external incisions. Lasers can be used to perform surgical procedures in RMIS, such as ablating, cutting, and excising tissue without or with small incisions. The laser is an effective treatment modality, which is projected on the target in a non-contact manner for surgical operations, minimizing the risk of contamination and infection. However, the requirement for a straight working path of the rigid optical components challenges the application of laser surgery within a confined environment. Traditional micromirror-based reflective laser manipulators require additional components to maintain the optical axis, resulting in a bulky size and an obstructed field of view. Moreover, in a 3D surgical scene, targets distributed at different depths cause laser defocusing, leading to the failure of laser ablation. Axial scanning by mechanically stretching the laser manipulator or stage suffers from mechanical jitter and slow response time. This study aims to address these challenges in three parts by developing laser steering techniques for robotic surgery.

First, a liquid-driven laser scanner installed on the end effector of a continuum endoscope is proposed to perform fast and reliable laser steering in a confined environment. The designed liquid-driven laser scanner is 7 mm in diameter and 7 mm in length. It uses a transparent liquid prism based on electrowetting on dielectric to achieve laser scanning. The liquid prism has no mechanical moving component, eliminating mechanical fatigue and providing a compact structure. An economical and reliable method for fabricating the liquid prism based on the liquid-infused membrane is developed, facilitating trapping charge migration and hindering dielectric breakdown. The performance of the liquid prism has no evident degradation in 1000 cycles, demonstrating the prism’s excellent stability. The maximum driving voltage of the liquid prism is reduced to 15 V, ensuring safety for biomedical applications. A theoretical model is established to predict the laser spot position quantitatively. Moreover, the laser steering ability is tested experimentally. Simulations and experiments verify the controllability and stability of the laser scanner. The designed liquid-driven laser scanner has the advantages of miniature size, low driving voltage, transparent structure, transmission mode, and high stability, which are favorable for laser-assisted surgery in confined spaces.

Second, a deep learning–assisted 3D laser steering strategy for robotic surgery capable of eliminating laser defocusing, increasing working distance, and extending scanning range is presented. A compound optofluidic laser scanner is designed to perform 3D laser steering and address the challenge of laser defocusing when scanning targets at different depths in a 3D surgical scene. The optofluidic laser scanner uses a liquid lens for laser focusing and a liquid prism for 2D laser scanning. A deep learning-based monocular depth estimation method with strong generalization provides real-time target depth estimation. Thus, the focal length of the laser scanner can be adjusted for laser focusing. In the axial direction, the laser can be focused from 40 mm to 180 mm, where 180 mm is the maximum depth for the endoscopic datasets SCARED, Hamlyn, and SERV-CT. Simulations and experiments indicate that the compound optofluidic laser scanner has the merits of miniature size, fast response time, and high repeatability, which can significantly increase the working distance and maintain laser focusing while performing 2D laser scanning.

Third, a focused laser ablation scheme based on the liquid lens is studied by using optical coherence tomography (OCT) imaging to analyze the laser ablation performance at different focal lengths quantitatively. An image-based method that can identify the ablation crater in real time and measure the crater diameter accurately is presented. In the axial direction, OCT imaging is used to measure the ablation crater depth. The relationship between the crater depth and the laser radiation time under different laser powers and focal lengths is also illustrated. OCT imaging has the advantages of being label-free, real-time, non-invasive, non-radiation, non-contact, and high resolution, which allow for the real-time continuous tracking of the impact of laser energy on the target. Experiments indicate the liquid device’s excellent power handling and laser focusing abilities through OCT imaging, providing a meaningful reference for future biomedical applications.

In summary, this study proposes feasible laser steering schemes for robotic surgery by achieving stable and reliable laser scanning and laser focusing. It provides meaningful insights for laser energy delivery in robotic surgery.

Date of Award	29 Aug 2024
Original language	English
Awarding Institution	City University of Hong Kong
Supervisor	Gang Gary FENG (Supervisor) & Dong SUN (Supervisor)