Computer Vision News - January 2022

65 Node21: Detecting Lung Cancer With the final submission date looming, Ecem and Keelin advise everyone to check the tutorials before you submit to make sure you have not forgotten anything! There is a forum on the Grand Challenge website where people can discuss issues and contact them if necessary. Finally, could we be seeing a second edition – NODE22 , perhaps? “ I’m not sure yet! ” Keelin teases. “ There may be future developments, especially on the generation track. This is something that is quite new. There’s been a lot of work recently about generation of different kinds of images, but we haven’t seen something that’s really clinically useful yet. We’re curious to see what will happen. ” The test cases used in the challenge are especially valuable because the labeling was performed by looking at a CT scan and chest X-ray together . This is gold standard labeling , which is great for participants developing algorithms. They can see how their algorithms perform on clinically relevant data. Ecem and Keelin work in the same group at Radboudumc and tell us they share the same principles. “ We work on open science, ” Keelin says. “ With this Grand Challenge platform, we make everything reproducible and let people try things out. We always aim to do things that are clinically useful. Working in a hospital here, we really want to make clinically relevant chest X-ray applications. ”