Grand Challenges in Biomedical Image Analysis
Every year, thousands of papers are published that describe new algorithms to be applied to medical and biomedical images, and various new products appear on the market based on such algorithms. But few papers and products provide a fair and direct comparison of the newly proposed solution with the state-of-the-art. We believe that such comparisons can help the research community and industry to develop better algorithms. We support the organization of these comparative studies and the dissemination of their results.
Organizing and participating in challenges is not the only way to facilitate better comparisons between new and existing solutions. If it were easy to publish and share your data, and the code you used to evaluate your algorithm's performance on that data, and possibly the algorithm itself, others could directly compare their approach to yours, using the same test data and the same evaluation metrics. With this site we provide tools to make it as easy as possible for you to publish your data and your evaluation for any paper you've written.
Why Challenges? describes the rationale for organizing grand challenges, provides advice for those who want to organize such events, and discusses where we hope the field will move to next.
All Challenges provides an overview of all previous, ongoing and upcoming challenges in biomedical image analysis that we are aware of. Drop us a note if you want your event listed on this overview.
Create your own project explains how you can set up your own challenge site in a matter of minutes, based upon the COMIC platform, open source and hosted on github, that we are developing within an international consortium. The idea is that you can easily reuse all the tools we have developed to set up challenge sites, and instead of a full-blown challenge, you can also create sites for sharing data, evaluation code, and algorithms. We also link to other platforms that offer similar solutions and invite everybody to help us build better platforms.
Grand challenges in medical image analysis would not be possible without the support and enthusiasm from all contributors.