Cornelius is an evidence-based intelligence system that uses real-time predictive modeling/AI decision-support to focus prevention and improve health outcomes.
COVID-19 Work.
We are currently building and testing patient-level predictive models of COVID-19 diagnosis and prognosis.
The first predictive model assesses the probability of positive COVID-19 test result.
The second model assesses the probability that a COVID-19-positive person would need to be hospitalized.
*****************
http://news.vanderbilt.edu/2013/10/cornelius-program-speeds-assessment-of-readmission-risk/
http://www.predictivemodelingnews.com/issues/pmnewspage1.pdf
- Project Goals
- To improve outcomes of hospitalized patients by focusing prevention resources using real-time predictive models.
- To provide rigorous scientific evaluations of the impact of the models on patient outcomes and publish the results in high-profile journals.
- To facilitate other researchers in the implementation and evaluation of new models.
- Predictive Models
- Pressure Ulcers
- Hospital Readmissions
-
-
- Sepsis
- Falls
- Hospital acquired infections
- Catheter-associated urinary tract infections
- CLABSI
- VTE
- AKI
- Research Behind Cornelius
- State-of-the-art predictive modeling techniques to ensure that the models are robust and the findings will replicate.
- Reproducible research methods.
- Modern methods of assessing validity, reliability and calibration.
- Rigorous randomized controlled trials to assess whether the models improve outcomes.
- Contact Us
- daniel.byrne@vanderbilt.edu
- henry.domenico@vumc.org
- Vanderbilt University Center for Technology Transfer & Commercialization