HIBRID

HIBRID is built on two pipelines: STAMP for image preprocessing and MARUGOTO for model training. Please follow the instructions below to run the code.

Installation

For setting up a local environment, note that the following steps are for Ubuntu Linux systems. First, install OpenSlide using either the command below or the official installation instructions:

apt update && apt install -y openslide-tools libgl1-mesa-glx # libgl1-mesa-glx is needed for OpenCV

Second, install conda on your local computer, create an environment with Python 3.10, and activate it:

conda create -n stamp python=3.10
conda activate stamp
conda install -c conda-forge libstdcxx-ng=12

Then, install the STAMP package via pip:

pip install git+https://github.com/KatherLab/STAMP

Next, initialize STAMP and obtain the required configuration file, config.yaml, in your current working directory, by running the following command:

stamp init

To download required resources such as the weights of the feature extractor, run the following command:

stamp setup

This will trigger a prompt asking for your Hugging Face access key for the UNI model weights.

Preprocessing

Ensure the Config file is filled with the correct PATH (Details can be viewed in the form of comments in the Config file)

stamp --config /preprocess/stamp/config.yaml preprocess  

Example Commands

Train

python model/marugoto/train.py \
-ct /path/to/clinical_table.csv \
-st /path/to/slide_table.csv \
-o /path/to/output_location \
-f /path/to/feature_directory \
-t OS OS_E DFS DFS_E

Deploy

python model/marugoto/eval.py \
-ct /path/to/clinical_table.csv \
-st /path/to/clinical_slide_table.xlsx \
-o /path/to/eval_results \
-f /path/to/feature_directory \
-m /path/to/model_output \
-c cohort_name \
-t OS OS_E DFS DFS_E

Additional Information

ct = clini table, using format:|PATIENT|FILENAME|OS|OS_E|DFS|DFS_E|
st = slide table, using format:|PATIENT| (required but redundant as slide info read from ct)
o = output location
f = feature directory
t = stats: OS overall survival, OS_E os event (i.e. dead/alive), DFS disease free status, DFS_E DFS event
m = model path (location of .pth output from train.py script)
c = cohort (additional name for output of eval.py)