Deep learning methods for histopathology image analysis
Codes for deep learning-based pipelines for whole slide tissue image (WSI) analysis:
Segmentation of Nuclei:
Code Repository:
https://bit.ly/2IEMDp8 (https://github.com/SBU-BMI/quip_cnn_segmentation)
Trained CNN model can be downloaded from
https://bit.ly/3pBHQFE (http://vision.cs.stonybrook.edu/~lehhou/download/model_trained.tar.gz)
Related Papers:
Hou, L., Gupta, R., et al. Dataset of segmented nuclei in hematoxylin
and eosin stained histopathology images of ten cancer types. Sci Data7, 185
(2020). https://doi.org/10.1038/s41597-020-0528-1
Hou L, Agarwal A, et al. Robust histopathology image analysis: to label or
to synthesize?. Proceedings of the IEEE Conference on Computer Vision and
Pattern Recognition 2019 (pp. 8533-8542).
Datasets:
https://tinyurl.com/yyqoq2n2
Creating Tumor Infiltrating Lymphocyte (TIL) Maps:
Code Repository:
Recent codes and models developed using the VGG16 and Inception-V4 networks.
These are the recommended models for use in TIL analysis.
https://bit.ly/38JSqEN (https://github.com/SBU-BMI/quip_classification)
Trained VGG16 and Inception-V4 models can be downloaded from
https://bit.ly/3lArJWA (https://stonybrookmedicine.box.com/shared/static/bl15zu4lwb9cc7ltul15aa8kyrn7kh2d.zip)
Codes used in the Cell Reports paper.
https://bit.ly/3nnFXuq (https://github.com/SBU-BMI/u24_lymphocyte)
Related Papers:
Saltz J, Gupta R, et al. Spatial organization and molecular
correlation of tumor-infiltrating lymphocytes using deep learning
on pathology images. Cell reports. 2018 Apr 3;23(1):181-93.
Abousamra S, Hou L, et al. Learning from thresholds: fully
automated classification of tumor infiltrating lymphocytes for
multiple cancer types. arXiv preprint arXiv:1907.03960. 2019 Jul 9.
Datasets:
Dataset in the Cell Reports paper:
https://doi.org/10.7937/K9/TCIA.2018.Y75F9W1
TIL analysis results from the VGG16 and Inception-V4 models:
(The data is being uploaded, check back in a few days.)
https://bit.ly/3kJJDVT (https://stonybrookmedicine.box.com/s/qb4gi6o1dihvds0tuieaclpdbcl03qzu)
TIL/Tumor Quantification in TCGA Breast Cancer WSIs:
Code Repository:
https://bit.ly/2K8pmwh (https://github.com/SBU-BMI/quip_cancer_segmentation)
Trained CNN model can be downloaded from
https://bit.ly/3pzvqyo (https://stonybrookmedicine.box.com/shared/static/1hdfb06lgd08xfbpoly9tjp6c6i665nz.zip)
Related Papers:
Le, Han, Rajarsi Gupta, Le Hou, et al. "Utilizing automated breast
cancer detection to identify spatial distributions of tumor infiltrating
lymphocytes in invasive breast cancer." The American Journal of Pathology (2020).
Datasets:
https://bit.ly/2UtZcpO (https://app.box.com/s/1qux9ub21zcvpwao1cf81ar4milxl25x)
Segmentation of Tumor Regions in Pancreatic Cancer Cases
Code Repository:
https://bit.ly/2IJRLZp (https://github.com/SBU-BMI/quip_paad_cancer_detection.git)
Trained CNN model can be downloaded from
https://bit.ly/32Pcn9q (https://github.com/SBU-BMI/quip_prad_cancer_detection/tree/master/data/models_cnn)
Related Papers:
Le H, Samaras D, Kurc T, Gupta R, Shroyer K, Saltz J. Pancreatic cancer detection
in whole slide images using noisy label annotations. InInternational Conference on
Medical Image Computing and Computer-Assisted Intervention 2019 Oct 13 (pp. 541-549). Springer, Cham.