Computer Vision News - August 2022

46 Innovation Prize at Hamlyn 2022 “ We are a big, multifaceted, interdisciplinary team working on this project , ” Lukas tells us. “ We have all this expertise and close collaboration with the UCL Ear Institute, which helps us immensely in the design process. Your readers know Sophia Bano and Dan Stoyanov , so I think they’ll know how good and well-established they are in the field of vision. Jeref Merlin is a research assistant with a biomedical engineering background, and he’s working with me on the robotic side, the hardware development and integration, and the fabrication techniques we’re using to build the robot. We also have our clinical collaborators, Nishchay Mehta and Joseph Manjaly , who work at the UCL Ear Institute and are very experienced in this field. ” From a clinical translation perspective, this work has a relatively linear validation process because there is no need to puncture the tympanic membrane initially – the robot can be deployed in the ear canal and perform some movements. The team can acquire imaging data and see how the The team started out using microscope images , primarily. Now, they are looking into domain adaptation because they have an endoscopic view of the tympanic membrane, which is not easily transferable. Also, working with custom-made medical phantoms , they need to find ways to make the model more generalizable from the microscope views they get from publicly available data sets. They have found a data set that is more applicable to their work and are investigating how that improves the algorithm’s performance.