BMIRDS / deepslide

Code for the Nature Scientific Reports paper "Pathologist-level classification of histologic patterns on resected lung adenocarcinoma slides with deep neural networks." A sliding window framework for classification of high resolution whole-slide images, often microscopy or histopathology images.

Home Page:https://www.nature.com/articles/s41598-019-40041-7

Geek Repo:Geek Repo

Github PK Tool:Github PK Tool

question about method

wangtaoknight opened this issue · comments

First, thanks for your hard work and share the entire code on github. I was just begain to studing this process and gain mauch from your work!
I tried to run your code. I see that you don't use the labeling information of the lesion. You are using WSI_ Is every patch defined as tumor? But tumor_ Most WSI are normal tissues. Is that what you do?

@wangtaoknight
We are not using the WSI in the dataset in the paper. The pathologists provided us with annotated regions consisting only of the tissues. The WSI labelling throughout the code is not entirely accurate for the case presented in the paper. But, this code can still be used for tissues or even whole slides and achieve good performance.

Oh, I got it. thanks for your reply.