Computer Vision News - April 2020

3 Summary Stamatia (Matina) Giann ou 11 With the ultrasound, we have less detail, but we can see how deep the tumor is going.” One challenge is these imaging probes can be very thin, have a small Field of View, and you need to keep a constant distance between the tissue and the probe. With deformations and hand tremor, scanning with your hands can be very difficult. If the scan is compromised, the quality of the information it provides for diagnosis support will be ineffective. This is why Matina is proposing to use robotic tools : “We’re using the cameras of the surgical microscope to get the 3D structure of the surgical environment, and then we are using robotic tools which are holding the imaging probes to scan the tissue surface and capture characteristics. We can then analyse this to give the surgeon diagnosis support and to help guide resection during the operation.” The team have proposed a technique which will be presented at the 2020 International Conference on Robotics and Automation – ICRA (which will probably be restructured to a virtual event). It uses the da Vinci robot to hold anultrasoundprobe anddo autonomous tissue scanning. “The advantage of this work is that we don’t make any assumption about the motion of the tissue,” Matina explains. “Previous work has relied on either a static environment or periodic motion, where the system has to learn the motion first and then move the robot according to the tissue motion. In our framework, we’re recovering the motion of the tissue during the scan and then moving the probe accordingly in order to achieve motion compensation.” At the Hamlyn Centre they have several da Vinci systems, used both for operations and research, and a da Vinci Research Kit (dVRK), which means they have access to the APIs. The da Vinci system has many applications; however, it isn’t currently used for neurosurgery, so the team are building small handheld probes and robots that can be used for microsurgery. S. Giannarou, M. Ye, G. Gras, K. Leibrandt, H. Marcus, G.-Z. Yang, “Vision-based deformation recovery for intraoperative force estimation of tool–tissue interaction for neurosurgery”, IJCARS, Jun 2016