MICCAI 2019

MICCAI 2019 DAILY 11 Zongwei Zhou )周纵苇( Zongwei explains the practical consequences of this work: “Our work is for disease detection. We don’t focus like doctors do on helping a patient to be healthy again. You have a scan and we tell you that you have a disease, but you still need to go to a hospital. Disease detection is usually done by radiologists. We want a machine, image by image, to really quickly and really accurately detect a disease.” This will certainly be faster, but will it also be more accurate? He says yes, for some diseases. Lung nodules , for example. There are some cases where machine learning is significantly better than radiology at detecting disease, some where the two are similar, and others where radiology is better. It’s a sliding scale. Thinking about next steps, Zongwei says they need some collected effort from other industries. This work is self-supervised learning without the need for human effort. Everything can be downloaded from the internet and the model can learn it. However, their lab is just a research lab, and they have limited GPU and human mind power. They have already published their work so far and invite people to contribute. If you want to learn more about this work, come along to Zongwei’s oral today at 11:26 and poster [T-5-B-013] tomorrow at 13:00. The paper is nominated for MICCAI 2019 Young Scientist Award.