This repository contains the code and resources for the paper "Machine-Learning based Prediction of Blood Stream Infection in Pediatric Febrile Neutropenia". The goal of this project is to develop a machine learning model that can predict the risk of bloodstream infections (BSI) in pediatric patients with febrile neutropenia (FN).
Febrile neutropenia is a serious condition that can lead to life-threatening infections in children undergoing chemotherapy. Early and accurate prediction of bloodstream infections can help in timely intervention and better management of these patients.
In this project, we explore various machine learning algorithms to predict the likelihood of bloodstream infections in pediatric patients with febrile neutropenia. The features used for prediction include clinical and laboratory data collected at the time of hospital admission.
The main objectives are:
- To identify key risk factors associated with BSI in febrile neutropenic children.
- To build and evaluate machine learning models for predicting BSI.
- To provide insights that can aid in clinical decision-making.
To run the code in this repository, you need to have the following software and packages installed:
- Python 3.8.8
- NumPy 1.21.2
- Pandas 1.2.4
- Scikit-learn 1.1.2
- Matplotlib 3.3.4
- Seaborn 0.11.1
- Shap 0.40.0
- Scikitplot 0.3.7
- Statsmodels 0.12.2
- Jupyter Notebook 6.3.0 (optional, for running notebooks)
You can install the required packages using the following command:
pip install -r requirements.txt