We want you to be able to become familiar with the technology the challenge will use and how to properly handle and address the training data. Before starting the beta test on the AWS platform we offer you to locally install the Eisen.ai framework and load sample labeled training data which has the same quality and structure as the real data you will later work with during the challenge. Of course, sharing patient data would not be wise and we will never do this. So we needed another solution.
Therefore, we created a toy dataset from real Covid-19 CT data that is synthetic (i.e. none of them is a real patient anymore) but quite realistic and pretty cool. You can read more about it on Ahmad’s Medium.com blog post.
The method he came up with to synthesize the data is pretty new and although we think it is safe we do not want to take any unnecessary risk. That’s why we feel this data should not be made publicly available without any rules. Therefore, we have a short data usage agreement that we want you to comply with.
So, in order to download the toy data, you should register yourself on this website (here is a quick link) and check the box that you will follow the simple rules. It will then lead you directly to the download page.
Have fun and give us feedback!