NIH Reporter #5K23HD113816-03

Predicting Clinical Deterioration in Mechanically Ventilated Children using High-Frequency Physiologic Data

This award funds the development and validation of machine learning models to predict clinical deterioration in mechanically ventilated children using high-frequency physiological data, aiming to improve early identification and intervention in critical care settings.

Description

Early identification of children at high risk of deterioration is a challenge, but a data-driven approach that incorporates high-frequency physiological data from bedside monitors will improve our ability to identify this at-risk population. The proposed work will derive, validate, and design for the implementation of a machine learning-based prediction model that identifies mechanically ventilated children at risk of new or worsening cardiorespiratory dysfunction. This is a critical step towards implementing acceptable and effective real-time, longitudinal prediction models at the bedside of the critically ill child.

View original record