Computer Vision News - January 2022

63 Node21: Detecting Lung Cancer analysis . Most of the researchers put their efforts into solving these problems using publicly available datasets, with labels generally extracted from radiology reports by text mining. Despite the importance of this nodule detection task, the only papers covering itwere fromcompanieswithenough resources to collect large, private datasets, which aren’t available to the community. ” NODE21 is a MICCAI-recognized challenge , which shows how important this issue is for the MICCAI leadership. Ecem sees it as more of a collaboration than a competition, presenting the opportunity as a research community together to develop a good systemfordetectingnodules and tocompare the performance of different algorithms in a fair way. It is a two-track challenge. The detection track is for developing automated algorithms Lung cancers , whichare visible in the lungs as nodules, are often fatal. Themost important resource available for fighting them is time. Once someone is showing symptoms, it is often already too late. Hence there is a need for easy and affordable imaging to identify these nodules at the earliest opportunity. Routine screening is one way to improve a patient’s chances and chest X-ray is by far the most cost-effective option. Chest X-rays are used in many other health-related scenarios, so checking them for nodules as a standard task would be one way to improve outcomes for lung cancer patients. “ Early detection of lung cancer is vital and is one of the most important applications for chest X-ray, ” Ecem tells us. “ Recently, we have published a review paper where we covered around 200 papers that were using deep learning and chest X-ray