1. Digital Twin Model Predicts Sepsis Mortality.
- Author
-
Kling, Jim
- Subjects
DIGITAL twins ,INTENSIVE care units ,COVID-19 pandemic ,RESEARCH personnel ,PULMONOLOGY - Abstract
A digital twin model successfully predicted adverse outcomes in ICU patients treated for sepsis by updating time-sensitive values every 15 minutes. The model, presented at the CHEST Annual Meeting, can potentially reduce the need for risky interventions by predicting patient outcomes. Researchers used a dynamic one-dimensional convolutional neural network to predict outcomes within the first 24 hours of admission, incorporating time-sensitive data like vitals and interventions. The model was tested on data from over 28,000 patients and showed improved accuracy over time. [Extracted from the article]
- Published
- 2024