amarcu5 / gbm-resectability-prediction

Reference code for "Improved Prediction of Surgical Resectability in Patients with Glioblastoma using an Artificial Neural Network"

GBM resectability prediction

This repository contains the source code for the publication "Improved Prediction of Surgical Resectability in Patients with Glioblastoma using an Artificial Neural Network" by Adam P Marcus, Hani J Marcus, Sophie J Camp, Dipankar Nandi, Neil Kitchen, and Lewis Thorne.

Usage

To ensure reproducibility the entire project runs within a Docker container. If needed, please install Docker before proceeding. Building, testing, and then running the project can be done as follows:

docker build -t gbm-resectability-prediction .
docker run -it --rm -v $(pwd):/home/ gbm-resectability-prediction build test run

Please note that FANN formatted training data files are required and should be placed under ./data/raw/. Ethics and privacy concerns prevent sharing of the original data set.

Contributing

Contributions are welcomed! The project's structure is based on Cookiecutter Data Science. All C++ code should adhere to the Google Style Guide with two allowed exceptions: frequent use of unsigned integers (to facilitate integration with the FANN library), and lack of namespaces (to shorten identifiers as small project and clashes are unlikely). Comments should be compliant with Doxygen.