Po-Han Lin
MIT EECS | Angle Undergraduate Research and Innovation Scholar
Machine-Learning Based Model to Predict Intensive Care Unit Length of Stay
2019–2020
EECS
- Artificial Intelligence & Machine Learning
Amar Gupta
Digital healthcare applies advances in technologies to conventional healthcare systems in order to increase effectiveness and quality. Recent developments in artificial intelligence allow analysis and inference on large datasets. The project aims to design and develop machine learning models that allow for betting prediction of length of stay (LOS) for patients in intensive care units (ICU). Most current ICU LOS models use linear-regression-based methods and are inaccurate, while machine-learning-based models have demonstrated the ability to improve mortality prediction with less complexity. The outcomes of the project can better inform the stakeholders and are crucial for resource planning, cost forecast, and setting expectations.
“I am joining SuperUROP because it will provide me in-depth research training and the opportunity to participate in the advances of a field. Coming from a background in CS and Bio-engineering, I am interested in digital healthcare and personalized medicine. Through this project of building a platform for medical data collection and analysis, I wish to sharpen my full-stack development/machine learning skills and learn more about healthcare digitization.”