/adverse-events

Bayesian optimization-guided topic modeling for automatic detection of sepsis-related events from free text

Primary LanguagePythonMIT LicenseMIT

adverse-events

license

This repository contains the source code related to the manuscript accepted to the IEEE International Conference on Bioinformatics and Biomedicine (IEEE BIBM 2021):

@ARTICLE{,
AUTHOR = {Yan, Melissa Y. and Høvik, Lise Husby and Pedersen, André and Gustad, Lise Tuset and Nytrø, Øystein},
TITLE = {Preliminary Processing and Analysis of an Adverse Event Dataset for Detecting Sepsis-Related Events},
YEAR = {2021}}

See here to download the exact version of the source code used in the publication (v1.0).

Usage:

  1. Clone repo:
git clone https://github.com/andreped/adverse-events.git
  1. Create virtual environment, activate it, and install dependencies:
cd adverse-events/python
virtualenv -ppython3 venv
source venv/bin/activate
pip install -r /path/to/requirements.txt
  1. Create the project structure as defined below:

  2. Run scripts for training and evaluating different classifier models:

python3 main.py misc/default-params.ini

Different parameters relevant for the analysis, building of models, evaluation, plotting results, and similar, may be modified in the INI-file.

Project structure

├── adverse-events
    ├── python
    │   ├── multi-class
    │   ├── topic-analysis
    │   ├── utils
    │   └── ...
    ├── data
    │   ├── EQS_files
    │   ├── file-with-all-notes.csv
    │   └── file_with_annotated_notes.csv
    └── output
        ├── history
        ├── models
        └── figures

Made with ❤️ and Python