Autumn newsletter 2023

Published 18 Sept. 2023

🛠️ New features of Grand Challenge

Welcome to the Autumn Edition of Grand Challenge Insights!

As the leaves change colour and the air turns crisp, we're excited to bring you the latest updates on our relentless journey to transform the field of pathology viewers. This season, we're diving deep into the realm of innovation, streamlining accessibility, and enhancing collaboration with groundbreaking features.

Explore with Ease: GitHub Integration

We're thrilled to announce our game-changing integration with GitHub repositories! Now, as an admin, you can effortlessly search through over 100 repositories within your organization, making collaboration and version control smoother than ever before. Unlock the power of efficient teamwork this autumn.

But that's not all! Introducing the Combined Leaderboard

Prepare to be amazed as we unveil this revolutionary feature. Our new Combined Leaderboard allows you to observe a user's algorithm(s) performance across multiple phases or tasks. No more jumping between leaderboards to track progress; the Combined Leaderboard ranks performance via a combination of the ranks of several challenge phases, providing a comprehensive view of excellence.

Elevate Your Reader Studies with New Question Options

For our dedicated researchers and participants in Reader Studies, we're introducing an array of innovative question types, including:
- Choice-type questions for easy selection and analysis.
- Text-type answers to questions for in-depth responses.
- Slider answers, where users provide responses on a sliding bar between a minimum and maximum value, adding a new dimension to your studies.

In this season of innovation and progress, we're taking collaboration, competition, and research to a whole new level. Stay tuned for expert insights, remarkable advancements, and celebrations of our community's accomplishments as we continue our mission to redefine the future of pathology. So grab your favourite warm beverage, cosy up, and get ready to embark on an autumn journey of discovery with Grand Challenge.

👋 Special: Grand Challenge at MICCAI 2023!

This year more than 15 MICCAI '23 challenges were hosted on Grand Challenge on various imaging modalities, including MRI, CT, ultrasound, etc. with many different goals: detection, segmentation, classification, and more.

We're looking forward to all the winning algorithms!¶

Meet us at MICCAI 2023!

Chris will be waiting at the AWS booth at MICCAI 2023, eager to chat with you about Grand-Challenge.org. Whether you have fresh ideas, exciting new challenges, or valuable feedback to share, he's there to listen and engage in constructive discussions 🤝. Feel free to approach him; your input is highly appreciated! If you want to be 100% sure to catch him, you can book a 1-on-1 at the booth via this link!

💡 Blogposts¶

"July 2023 Cycle Report"¶

Read all about the platform improvements from our RSE team in July, including the expansion of reader-study question options and response systems.

"August 2023 Cycle Report"¶

In the August cycle, the RSE team worked on various tasks including the viewer for pathology images, integrating multiple leaderboards into one for challenges and improving the integration of GitHub repositories. Read more about it in this blog post!

"September 2023 Cycle Report"¶

The RSE team introduced optional hanging protocols and re-introduced the polygon-to-mask editing capabilities for the CIRRUS online viewer. Check out these features in more detail in the blog post!

🔦 Highlighted algorithms¶

Deep learning to estimate pulmonary nodule malignancy risk using a current and a prior CT image

This is a deep learning (DL) algorithm based on convolutional neural networks (CNNs) that accepts a current and a prior low-dose chest CT, the coordinates of a pulmonary nodule, and the time difference (delta) between the CT examinations as input to estimate the malignancy risk of the nodule. The output of this algorithm is a score between 0 - 1 with 3 decimal values indicating the malignancy risk in the pulmonary nodule(s). Try out this algorithm with your own data!

Spine segmentation in MR

This algorithm segments the spine in sagittal lumbar MR images and predicts whether a vertebra is completely visible within the field of view. The output of the algorithm is a segmentation of all visible vertebrae, intervertebral discs (IVDs), and the spinal canal. All structures are segmented separately and are provided with a unique label. Try out this algorithm with your own 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 Lumbar SPIDER Challenge focuses on the segmentation of three anatomical structures in lumbar spine MRI: vertebrae, intervertebral discs (IVDs), and spinal canal.
✍️ Register: This challenge will run indefinitely
⏰ Deadline: This challenge will run indefinitely and remain open to provide an available benchmark for newly developed algorithms.

🏆 Leaderboard of finished challenges¶

The task for this challenge was two-fold: to validate modern AI algorithms’ performance at csPCa detection and diagnosis. The overall winner was Noelie Debs. Congratulations! The top 5 prostate-AI teams (participating in both phases of the challenge) will be invited to join the PI-CAI consortium, and in turn, they will be listed as consortium authors on an upcoming high-impact journal paper summarizing the findings of this challenge.