Keras unet and then chromogen analysis for prostate histopathology images.
.
├── predict_calculate_dab # scripts for the prediction of histo images and chromogen analysis
│ ├── new_output_dir.sh # bash script to create the output directories
│ └── run_all_unet_predictions.py # runs precictions on each TA image and calculates DAB measurements
└── train_model # scripts for training the 3 unet models that are used
├── checkfiles.py # for checking the dataset for corrupt files
└── unet.py # for trianing vanilla unet model The scripts can be applied in the following way:
- Train three Unet's using unet.py
- One for acini recognition
- One for nuclei recognition
- One for stroma recognition
- Apply these Unet models to:
- A) Segment images into masked areas of interest
- B) Calculate the amount of DAB/Chromogen stain in these areas