Risk Stratification for ICU Delirium using Pervasive Ambient Sensing Information
Researchers have developed sequential neural network models to predict Intensive Care Unit (ICU) delirium using ambient sensing data, specifically light intensity and sound pressure levels. A convolutional model demonstrated strong discrimination, achieving an AUC of 0.80, with sound features proving to be the most dominant predictors. Integrating sound and light data improved short-term prediction, suggesting passive ambient sensing offers a practical method for enhancing delirium risk estimation and prevention strategies. AI
IMPACT This research demonstrates the potential of AI in healthcare for early risk stratification of conditions like delirium, using non-traditional data sources.