Hi! I noticed, in section 3.6, submitting models, there is some information about the data description. However, I can not understand it completely. Do we have the imageview information of each training image in the sub challenge 1? And it seems we have the information in the testing as the image_metadata.txt shown in the testing section. For sub challenge 1, is the question, given an image, and we need to classify it as a norm or cancer? Thanks!

Created by Wentao Zhu wentao
Hi, > Do we have the imageview information of each training image in the sub challenge 1? The imageview is available for every images in the challenge. > For sub challenge 1, is the question, given an image, and we need to classify it as a norm or cancer? There is a confusion in the usage of "mammograms" (exam vs image). Thanks for bringing this to us. This will be clarified. To answer your question, we are asking in Sub-challenge 1 to determine if a patient will develop a cancer within one year (1) or not (0) given a series of mammograms (images) taken during an exam. I expect most of the participants to develop a method that generate a prediction at the image level, however the output requested is at the exam level.

view information is avaible or not in the training and testing page is loading…