Vitreoretinal surgery remains one of the most challenging types of surgery due to the delicate nature of the eye, the minimally invasive nature of the procedure, the small operating space and lack of depth perception. One of the procedures that is performed during vitreoretinal surgery is Epiretinal Membrane (ERM) Peeling , during which the surgeon has to peel a membrane of the retina without damaging the retina. Robotic membrane peeling can greatly improve clinical outcome because of increased accuracy, tremor reduction and little fatigue effects. However, depth perception remains a problem, because the surgeon is looking at the procedure from directly above through a microscope, therefore creating a 2-D vision. The topic of this study is a tele-operated robotic system that is specialized in vitreoretinal surgery. The PRECEYES surgical system is equipped with an OCT -sensor that can estimate how far away the tip is from the retina, but not accurately enough for traditional 'forbidden region' assistance. This study aims to design and evaluate a proof-of-concept for haptic support during robot-assisted vitreoretinal surgery, by means of an artificial damping-field on the master device. The main contributions of this study were 1) to create a simulation environment for robot-assisted vitreoretinal surgery; 2) to design and implement a damping field on this simulator; and 3) to perform a human-in-the-loop experiment to compare unassisted control behaviour to assisted control behaviour
To study the effect of the damping field, a simulation of the procedure was created. A physical haptic master device controls a virtual slave in a simulated environment. The participants (n=16) were asked to move the instrument towards the retina, and then gently move it through a ring on the surface of the retina. This task forced the participant to make a peeling motion that is typical for ERM peeling. The experiment had 4 conditions: a small or a large target, with the damping field switched on or off. Each of the 16 participants had to perform 10 successful runs per condition. A retina puncture resulted in a failure and is not counted as a successful run. Results showed that for the small target, the average number of punctures per participant decreased from 2.4 to 0.9 when the damping field was enabled. The average completion time was comparable for both small target conditions, and likewise for the large target conditions. It was observed that participants decreased their velocity at the same distance of the retina for multiple runs when the damping field was active, which indicates that the participants were able to estimate the distance to the retina using the damping field. Furthermore, the damping field enabled the participants to employ a larger safety margin (distance between the instrument and the retina) between the instrument and the retina. For the experimental conditions studied, the damping field decreased the amount of punctures without affecting the completion time. Improvements to the damping field, such as an exponential increase of the damping coefficient and moving the starting point closer to the retina could increase the performance.