need guidance

Question

need guidance

Opened this issue 3 months ago · 2 comments

Hello, author. You mentioned in #23 that I didn't use real training images, but I still don't understand. I noticed you referred to 12 million images. Could you please tell me where I can find them?Or where in the code can I view the import of data? I have checked the TCGADataset in tumor_til_in_text.py, as well as the DataModuleFromConfig in main.py

Answer 1 · 2024-06-25T15:46:50.000Z

See #4 . We uploaded the h5 file to Drive, which contains TCGA BRCA patches at 10x magnification.

Answer 2 · 2024-07-11T12:20:06.000Z

Hello, thank you for the work.

Could you provide the code you used to encode the images and prepare .npz and .hdf5 files? I'd like to prepare my custom dataset following the same approach you used.