Baseline algorithm

Dear Organizers!

Can you provide a tutorial on the baseline algorithm, data preprocessing, and a guide to creating a Docker image?

Hi fy1999xu,

Thank for your interest in our challenge! Documentation on how to create a docker image for an algorithm can be found here: https://grand-challenge.org/documentation/create-your-own-algorithm/

We used Hover-NeXt (https://github.com/digitalpathologybern/hover_next_inference) and nn-UNet (https://github.com/MIC-DKFZ/nnUNet) and we create specific docker containers compatible with the GC platform. This can all be found here: https://puma.grand-challenge.org/submission-instructions/

Regarding data preprocessing and baseline algorithms, we did not modify the original baseline code from nnU-Net or Hover-NeXt. To participate, you only need to ensure that your data is formatted to be compatible with their algorithms. Additionally, your Docker container should follow the guidelines to produce the required input and output, as outlined in the submission instructions linked above.

That said, the primary goal of the challenge is to create a better algorithm, so feel free to explore other models that might outperform the baselines.

Good luck! Kind regards,

Mark