Computer Vision News - May 2024

Computer Vision News 44 main take-home message will be a broader understanding of the data behind medical imaging methods. “Medical imaging is a really interesting area because we’re very close to applications and have many safety and fairness-related issues we have to think about,” Dominik points out. “We often think about things that maybe aren’t the biggest topics in the main computer vision area, and that’s something we want people to think about, have new perspectives on, and maybe use in their future research.” At Intuitive, Shuoqi sees the realworld applications of medical imaging and computer vision every day. Through his work on the robotic lung biopsy tool Ion, which helps diagnose lung cancer in the peripheral lung area, he knows that precise image segmentation relies on robust solutions that cater to diverse patient profiles and healthcare settings. “Our goal is to develop robust and general solutions for lung biopsy,” he explains. “We want a solution that works in most cases – not just one dataset for one hospital. We want it to work for all hospitals.” Eagle-eyed readers will know this is not the first time we have featured Dominik in our magazine, and as he tells us, his November 2021 interview about his ICCV paper had an unintended consequence that led us to where we are sitting today. “You’re actually the reason Shuoqi and I know each other!” he exclaims. “He read your article on my paper on synthetic data and, after trying the method himself, contacted me with some questions,” he pauses. “My code’s documentation might not have been perfect!” he laughs. “We stayed in contact after that and even met at Intuitive in California in 2022. Then, when Shuoqi started planning the workshop, with synthetic data being part of the scope, he asked me if I wanted to join the team, which I was more than happy to do. So, thanks again to you – even the first interview is still creating new opportunities!” CVPR 2024 Workshop Preview