Skip to content

Template repo for Data Science for Biomedical Informatics (BMIN503/EPID600) final project

Notifications You must be signed in to change notification settings

mcarr04/BMIN503_Final_Project

 
 

Repository files navigation

BMIN503/EPID600 Final Project

BMIN5030 Final Project Project Title: Effectiveness of AI-Based Models in Predicting Patient Mortality Outcomes

Description: This project explores the effectiveness of AI-based models in predicting patient mortality outcomes compared to traditional clinical methods. It analyzes patient data from the MIMIC-IV database and evaluates the accuracy and potential benefits of AI in healthcare.

Objectives: To assess the predictive accuracy of AI-based models for patient outcomes such as ICU readmissions or mortality. To compare AI-based predictions with traditional clinical prediction methods. To identify key factors influencing prediction accuracy.

Dataset: Source - MIMIC-IV database

Description: Contains de-identified health-related data associated with critical care unit patients.

Tools and Technologies: Programming Language - R IDE - RStudio Libraries - tidyverse for data manipulation and visualization, caret for model training, randomForest for AI-based predictions, survival for survival analysis

Repository Structure: data/: Contains preprocessed datasets used for modeling and analysis, scripts/: R scripts for data processing, modeling, and analysis, results/: Outputs, including model results and visualizations, docs/: Project documentation and presentation files, README.md: Project overview (this file).

Results: The analysis demonstrates [briefly describe key findings]. Insights into the potential benefits of AI-based predictions in clinical settings are highlighted.

Limitations: The project is limited to the scope of the MIMIC-IV dataset. Computational constraints may affect model training and evaluation.

Future Work: Extend the analysis to other patient outcomes. Integrate additional datasets to improve model generalizability. Compare other AI techniques for prediction accuracy.

Author: Madison Carrigan, Email: [email protected]

Acknowledgments: Special thanks to Blanca Himes, Anastasia Lucas, Jakob Woerner, Dr. Lama Al-Aswad, and peers of BMIN5030 for their guidance and support throughout this project.

License: This project is licensed under the MIT License.

About

Template repo for Data Science for Biomedical Informatics (BMIN503/EPID600) final project

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published

Languages

  • HTML 100.0%