Autumn newsletter 2024

Published 14 Nov. 2024

🛠️ New features of Grand Challenge

Welcome to the Autumn Edition of Grand Challenge Insights!

As the crisp air sets in and the leaves turn vibrant shades of amber and gold, we are excited to share the latest updates and highlights from our community. This autumn, we're celebrating meaningful milestones, creative endeavours, and heartwarming stories from participants across the globe. Let’s embrace this season of change and reflection, as we continue to grow and collaborate together!

Control Image Registration!

Editors of reader studies or algorithms can now manage how images are registered with each other while interacting with them in the viewer. This new feature in CIRRUS interprets registration data, providing more control and precision in image alignment.

Custom Challenge Sign-Ups!

You can now gather more detailed insights from participants when they sign up for your challenge. In addition to basic information like affiliation, challenge organizers can include custom registration questions to learn more about participants' experience level or the approach they plan to take.

As we step into the rich colors and cosy atmosphere of autumn, we're excited to see new ideas and collaborations taking shape within our community. Let’s make this autumn a season of growth and creativity, as vibrant as the falling leaves!

💡 Blogposts¶

"July 2024 Cycle Report"¶

Read all about the platform improvements from our RSE team in July, including new cirrus features for image registration.

"September 2024 Cycle Report"¶

In the September cycle, the RSE team worked on enabling synchronous execution of algorithms in reader studies. Until now, algorithms on Grand Challenge typically run asynchronously, meaning they are executed in parallel when resources are available. Now, reader study editors can now provide an interactive algorithm for certain reader study questions, which readers can then execute on demand while viewing cases. Additionally, the RSE team focused on enhancing challenge sign-ups. Challenge organizers can now include custom registration questions when participants sign up for your challenge. Read more about it in this blog post!

🔦 Highlighted algorithms¶

OrbitFracNet

This algorithm automatically detects fractures of the bony orbit in CBCT scans. Accuracy is reported to be 0.925. Try out this algorithm with your data!

Automated NASH grading of liver histopathology

This model detects steatosis and inflammation in a given WSI. In addition, the model also localizes the portal regions, due to their high similarity to inflammation. The output of the models is one mask, in which the background (0 - colorless), liver tissue (1 - blue), ballooning (2 - green), inflammation (3 - yellow), and portal regions (4 - orange) have been segmented. Try out this algorithm with your data!

🚀 Upcoming and running challenges¶

Grand Challenge currently offers two types of submissions: prediction submission and algorithm container submission. The algorithm container submission type has the advantage of producing reproducible algorithms that remain accessible to the research community long after the challenge has ended. This allows for continued use and exploration of the algorithms by the community. Therefore, it should be noted that we are phasing out the prediction submission procedure in favour of the algorithm submission procedure to ensure that challenges always produce reproducible algorithms. To learn more about hosting a challenge on our platform, go to our documentation!

🥅 Goal: The goal of the PUMA challenge is to improve nuclei and tissue segmentation in melanoma histopathology. The challenge consists of 2 tracks:
1. Panoptic segmentation with three instance classes: tumor, TILs (lymphocytes and plasma cells), and other cells
2. Nuclei detection for all classes: tumor, lymphocytes, plasma cells, histiocytes, melanophages, neutrophils, stromal cells, epithelium, endothelium, and apoptotic cells
✍️ Register: Registration is now open!
⏰ Deadline: Accepting AI algorithm submissions until March 1st, 2025.
🏆 Prizes:
Up to four members of each leaderboard's top three performing teams will be invited to participate in the challenge paper as consortium authors

🥅 Goal: The goal of the MONKEY challenge is to yield top-performance deep learning solutions to automate biopsy assessment of transplant kidneys. The challenge consists of 2 parts:
1. Detection of mononuclear, inflammatory cells
2. Detection and distinguishing inflammatory cells: monocytes and lymphocytes
✍️ Register: Registration is now open!
⏰ Deadline: Accepting AI algorithm submissions until December.
🏆 Prizes for inflammation cell detection:
1. €800,-
2. €500,-
3. €300,-
🏆 Prizes for monocyte and lymphocyte detection:
1. €800,-
2. €500,-
3. €300,-

🥅 Goal: The PANORAMA Challenge focuses on detecting pancreatic cancer in contrast-enhanced CT scans.
✍️ Register: Registration is now open!
⏰ Deadline: Accepting AI algorithm submissions for Open Development Phase from April.
🏆 Prizes:
1. €1.000,-
2. €500,-
3. €250,-
4. €150,-
5. €100,-

🏆 Leaderboard of finished challenges¶

The submission phase for several MICCAI challenges has officially closed. The result were announced at the MICCAI conference.

A big thank you to all participants for your hard work and dedication in pushing the boundaries of medical image computing and AI. Best of luck to everyone, and we hope you've enjoyed the conference filled with exciting presentations, discussions, and the unveiling of challenge outcomes!