The MATLAB script evaluate_sepsis_score.m and Python script evaluate_sepsis_score.py evaluate predictions from your algorithm using a utility-based evaluation metric that we designed for the PhysioNet/CinC Challenge 2019. These scripts produce the same results. For PhysioNet/CinC 2019, we use the utility score, which is the last (fifth) score in the output of these scripts.
You can run the MATLAB evaluation code by running
evaluate_sepsis_score(labels, predictions, 'scores.psv')
in MATLAB, where labels is a directory containing files with labels, such as the training database on the PhysioNet webpage; predictions is a directory containing files with predictions produced by your algorithm; and scores.psv (optional) is a collection of scores for the predictions (described on the PhysioNet website).
You can run the Python evaluation code by installing the NumPy Python package and running
python evaluate_sepsis_score.py labels predictions scores.psv
where labels is a directory containing files with labels, such as the training database on the PhysioNet webpage; predictions is a directory containing files with predictions produced by your algorithm; and scores.psv (optional) is a collection of scores for the predictions (described on the PhysioNet website).
This repository contains evaluation code for the PhysioNet/CinC Challenge 2019. Looking for Julia, MATLAB, Python, or R example prediction code? See the repositories in https://github.com/physionetchallenges.