Regarding the number of pixel classes corresponding to the Task1 Labels images

Hello, thank you for providing the baseline code. I encountered an issue during the data preprocessing of the baseline nnunet: You have configured the dataset.json file in nnunet_results as follows: "labels": { "background": 0, "tumor": 1 }, However, when I was configuring the baseline nnunet preprocessing, I encountered the following error: Error: Unexpected labels found in file F:\DeepLearning_ICG\Hjl_Lx\Projects\nnUNetFrame\dataset\nnUNet_raw\Dataset090_PantherTask1\labelsTr\10061_0001.mha. Expected: [0, 1] Found: [0. 1. 2.] It seems that there is a third pixel class in the data, but only two pixel classes are defined in the dataset.json file.

Hello, the provided labels include two classes: tumor(1) and pancreas(2). The baseline only uses the tumor mask, so nnU‑Net expects labels 0 and 1. You can re‑save the labels with class 2 removed (or mapped to 0) before preprocessing. The pancreas masks are just optional additional information for participants and aren’t used in the baseline model.