Dataset labels
Hello,
Thank you very much for sharing this dataset. I would like to ask what the ground truth labels correspond to in this dataset. My understanding is 1 -> positive, 0 -> negative, 2-> undecided. is this correct?
Hi Batu,
yes, your understanding is correct: 0 -> negative, 1 -> positive, 2-> undecided / ambigious.
Best,
Lorenz
Hi Lorenz,
Thank you for the prompt reply.
In the gt_pred_ie_consolidated.csv file, as well as all the ground_truth.csv files, I see that the labels can be 0, 1, 2, and 3. I believe these are related to negative, positive, likely positive, and likely negative labels in your manuscript. Then can I still group these labels as 0 -> negative, 1 -> positive, 2&3 -> undecided ?
Best,
Batu
Hi Batu,
What you wrote is correct.
Best,
Lorenz
Hi Lorenz,
Thank for your clarifying.
All the best,
Batu
Hi Lorenz,
Sorry for poking again. I am having some difficulties understanding the difference between the segmentation images named as "imagefeature_0" and "imagefeature_1" in vectra colon and pancreas datasets. I see that in the training script you are using only the "imagefeature_0", but just wanted to know the difference between them.
Best,
Batu
Hi Batu,
no problem. So imagefeature_0 is the whole cell segmentation, the imagefeature_1 is the nuclei segmentation.
Best,
Lorenz
Thanks for the clarification!
best,
Batu