Session Title: New Drug Treatment and Technology I
Session Date/Time: Friday 18/09/2015 | 11:00-12:30
Paper Time: 12:12
First Author: : M.Mura NETHERLANDS
Co Author(s): : D. Iannetta N. de Jonge G. Naus M. Beelen T. Meenink M. de Smet
PURPOSE:Vitreoretinal (VR) surgery requires a high level of surgical skill. High-precision robot assistance can improve a surgeon’s skills. E.g., higher precision improves the reproducibility of existing procedures and the development of new, high-precision procedures that cannot be performed manually. The PRECEYES Surgical System was developed to assist in VR surgery. The surgeon uses a motion controller to manipulate an instrument manipulator, which controls the instrument. The purpose of this study is to evaluate positional precision and steadiness of a surgeon, comparing manual and assisted instrument positioning inside the eye.
Tests were performed by a surgeon in an eye simulator at the Medical Robotic Technologies lab of PRECEYES, Eindhoven, the Netherlands.
To measure positional precision and steadiness, a tracing test was developed. A test model was created, resembling the VR environment. The model consisted of a hollow styrofoam bottom half which is lined inside with millimeter (mm) paper. The top half, composed of transparent plastic resembling an artificial sclera was placed above. A trocar was present through which an instrument can be inserted into the model. With the tip of the instrument a square of 4 by 4 mm was traced on the mm paper, in the region corresponding to the macula. The test person was asked to freeze the instrument for 3 seconds at the corners of the square. Image analysis was used to calculate i) the deviation between the tip of the instrument and the line that was traced and ii) the deviation between the tip of the instrument and the corners of the square during the freezing step. These deviations provide measures for positional precision and steadiness, respectively. Each test was repeated 3 times in manual and robotic mode.
Manual use of the instrument resulted in an average positional deviation between the tip of the instrument and the line of 80.8 μm ± 50.6 μm. The average deviation between the position of the tip and the position of the corner was 125.7 μm ± 62.9 μm. It took 62.8 ± 7.7 seconds to complete these manual experiments. Robot-assisted use resulted in an average positional deviation between the tip of the instrument and the line of 49.6 ± 28.5 μm. The average deviation between the position of the tip and the position of the corner was 55.9 μm ± 23.5 μm. It took 95.6 ± 9.4 seconds to complete these assisted experiments. Furthermore, literature shows that manual positioning precision of a good surgeon is in the order of 125 μm. Off-line experiments with the stand-alone robotic assistant have demonstrated its intrinsic precision to be below 10 μm. Its steadiness can only be compromised by movement of the eye model, as it freezes at its exact position.
The results indicate that positional precision and steadiness in the X-Y plane, using a simulated surgical environment to be twice as high with assisted surgery as compared to manual surgery. This relative improvement is significant, whilst the difference in procedural time is relatively small. However, the absolute values of the results have to be interpreted carefully. Comparing the values from literature and off-line experiments with the measurement results demonstrates that these values are highly dependent on the imaging modality used, in this case a Zeiss OPMI stereoscopic microscope, and the interpretation of the surgeon analyzing the images in real-time and correcting the instrument positioning accordingly. Furthermore, the evaluation method will influence the results. E.g., the steadiness of the robotic system, measured whilst freezing the instrument position at the corner of the grid, was measured ± 23.5 μm. This deviation is introduced by the evaluation method and/or movement of the eye model, as the robot freezes the exact position. Summarizing, it can be concluded that assisted surgery allows for significant improvement of both positional precision and steadiness, which will benefit reproducibility of existing treatments as well as enable the development of high-precision procedures. To obtain absolute values, further experiments are required.