| title | emoji | colorFrom | colorTo | sdk | sdk_version | app_file | pinned |
|---|---|---|---|---|---|---|---|
H&E-to-IHC Stain Translation |
🪄🧬🌈 |
red |
blue |
gradio |
5.7.1 |
app.py |
false |
Gradio App based on Adaptive Supervised PatchNCE Loss for Learning H&E-to-IHC Stain Translation with Inconsistent Groundtruth Image Pairs (MICCAI 2023)
Online demo:
Original folder: lifangda01/AdaptiveSupervisedPatchNCE
Original paper:
This repository features a Gradio-based application built on the methods introduced in the MICCAI 2023 paper, "Adaptive Supervised PatchNCE Loss for Learning H&E-to-IHC Stain Translation with Inconsistent Groundtruth Image Pairs." The application facilitates automatic virtual staining, transforming H&E (Hematoxylin and Eosin) images into corresponding IHC (ImmunoHistoChemistry) images.
Users can generate virtual IHC stains for four key biomarkers critical to breast cancer diagnostics:
- HER2: Human Epidermal Growth Factor Receptor 2
- ER: Estrogen Receptor
- Ki67: Antigen KI-67 (cell proliferation marker)
- PR: Progesterone Receptor
This tool simplifies and accelerates the analysis of histopathological samples, making advanced diagnostic insights more accessible through virtual staining technology.
| Input H&E Image | Output IHC Image (Ki67) |
|---|---|
 |
 |