Clinical Note Representations That Predict ICU Mortality/Re-Admission
Our focus is on improving health-related outcomes by creating machine learning models that improve human risks.
Predicting health-related outcomes in the current fragmented healthcare system is a difficult endeavor. We leverage data and create models in healthcare settings to advance how we predict health-related outcomes, specifically creating statistical models. In this work, we examined readmission rates of psychiatric patients, and the overall severity of patients’ general illnesses. Readmission rates in hospitals contribute to rising healthcare costs, are complicated to foresee, and can oftentimes be the only option for patients. The ICU is also a particularly challenging environment, because each patient’s case is constantly evolving.
Our team aims to identify predictive factors that may not be readily available in electronic health records and create statistically relevant models. In doing so, we can get closer to creating models that guide interventions plans, and address risks before it’s too late.
