How did you train your network on TCGA-NSCLC dataset for three manifications?

BearCleverProud / HAG-MIL

Repository for Hierarchical Attention-Guided Multiple Instance Learning

14 stars 2 forks source link

How did you train your network on TCGA-NSCLC dataset for three manifications? #2

Open Lewislou opened 3 months ago

Lewislou commented 3 months ago

I have noticed some WSI in NSCLC dataset do not have level 2, so how did you extract patches for these images for your framework? For example: TCGA-05-4384-01Z-00-DX1.CA68BF29-BBE3-4C8E-B48B-554431A9EE13.svs, TCGA-63-A5MH-01Z-00-DX1.596077FF-9CE1-4EA7-9BB6-222C69872CA4.svs and many others.

Lewislou commented 3 months ago

BearCleverProud commented 3 months ago

I have noticed some WSI in NSCLC dataset do not have level 2, so how did you extract patches for these images for your framework? For example: TCGA-05-4384-01Z-00-DX1.CA68BF29-BBE3-4C8E-B48B-554431A9EE13.svs, TCGA-63-A5MH-01Z-00-DX1.596077FF-9CE1-4EA7-9BB6-222C69872CA4.svs and many others.

We actually operates on level 1 and level 0 only for TCGA datasets, NO level 2. Sorry for the confusion.

bravePinocchio commented 2 weeks ago

May I ask how this problem can be solved?