-
Regarding the “preprocessed feature vector” you provided, were the coordinate points extracted at level 1? Additionally, could you please specify the segmentation parameters used when preprocessing th…
-
非常感谢你们的工作，我想请教一下，1.TCGA-LUNG数据预处理过程，提取的patch size是多大，在那个倍率下提取的。
2.提供的LUNG的feature vector 是基于ResNet18的还是ResNet50，3.特征提取器是否经过自监督对比训练。
-
Hello Fabian,my questions is :
My data is from [camelyon16,](https://camelyon16.grand-challenge.org/) It's a **big** **.tif** 2D data,I want to use nnunet for this ,but I can not convert it to nii…
-
When extracting features for camelyon16, when extracting tumou78, I encountered this error:
These are the following parts of the error：
File "/home/ubuntu/anaconda3/envs/clam/lib/python3.7/site-pa…
-
Thanks for the inspiring paper.
I was trying to replicate the result for Camelyon 16 on TransMIL, while I was reading the config, I was confused by the split path (.npz), as I think CLAM datasplit do…
-
Hi, your work is very fascinating！
I would like to replicate your work on TCGA-Lung datasets. I noticed that there are many empty .md files in the repo. Are there any pending contents to be written?…
-
Could you please let me know how to patch the camelyon16 dataset with the same settings as yours?
Could you also let me know how to get 130 slides for testing? When I tried downloading the camelyon…
-
Dear bin,
Thank you for your great work!
- When I reproduce the results on c-16 and TCGA, I follow the provided readme: 1) Using pre-computed features from `Download feature vectors for MIL netwo…
-
Thx for your great work!
I followed the instructions written in the paper and its appendix to reproduce the results on camelyon16.
However, I observed that its training was hard to converge (the l…
-
Hello,
You said, "3.2 million patches at 20× magnification" you get in the paper.
But I found some of it is at magnification 20x and the other is 40x(as in fig 1) after I checked the Camelyon16 da…