Abstract
Purpose :
The objectives of this study are to map the scleral forces present between the sclera and the tool-shaft using force sensing retinal instruments. The impact of robot assistance as compared to freehand techniques on tool to scleral forces is quantitatively assessed.
Methods :
This study was performed in three phases: 1) in a dry phantom, 3D printed orbit and globe; 2) in 20 ex vivo porcine eyes; 3) in vivo, using 12 New Zealand Rabbit eyes. A cooperative robotic platform was used (Steady Hand Eye Robot 2.1, Johns Hopkins University), with a force sensing instrument, based on Fiber Bragg Grating. The task was to move the instrument, adjacent to the retinal surface, in radial directions, from the center to the periphery and back. The tool was also moved circumferentially around the macula, and along the equatorial line. A control method to adjust instrument velocity in the direction of increased force was implemented by the robot and tested.
Results :
In the 3D printed model, 160 trials demonstrated the presence of larger forces for circular movements (207 ±91mN) than for axial movements (150 ±76 mN), p=0.001. Overall, the regions with greatest force in the axial directions were inferior temporally (173 ±113mN), and nasally (162 ±54mN), while lower forces were found temporally (130 ±49 mN) and inferior nasally (141 ±80 mN), figure 1. The pattern of force distribution varied with port position, with increased forces on circular movements when separated by 120o (244 ±97mN), as compared to 180o (171 ±71mN), with p=0.001. In ex vivo eyes, the transmitted forces were similar between robotic and manual assistance, respectively 146 ±88 mN and 133 ±65 mN. In in vivo eyes however, the forces were 102 ±85mN using robotic-assistance, 80 ±56mN freehand, and 61 ±34mN using robotic assistance with a control algorithm.
Conclusions :
Scleral forces during robotic surgery are a potential safety concern and crucial to coordinated movements and surgical awareness. This study quantitatively maps scleral forces for various instrument positions with and without robotic control. Differences are attributed to translational movements in circular motions, instrument stress to reach specific regions, as well as anatomic orbital constraints. Despite differences between models, the use of dry phantom to study this question of emerging relevance is valuable and may limit the need for animal use in future studies.
This is a 2020 ARVO Annual Meeting abstract.