Significant barriers to the translation of machine learning based predictive models in clinical care continue to exist. The barriers are both technical and social and successful deployments require significant effort. This panel will provide targeted case studies of machine learning based predictive models through model evaluation, deployment, and adoption. The panel will demonstrate end to end examples of machine learning based predictive models and the issues and processes leading to their successful deployment. In some case studies, model evaluation results demonstrate improvements in quality of care and reductions in cost of care.
Learning Objective: Demonstrate examples of machine learning based predictive models deployed and adopted in clinical practice.
Describe challenges in successful deployment and adoption of machine learning based predictive models.
Describe operational principles supporting the deployment of machine learning based predictive models.
Describe methods for evaluating the clinical impact of these machine learning based predictive models.
Describe interventions and measurement tools at the point of care.
Describe how organizational and engineering structure contribute to model translation.
Yindalon Aphinyanaphongs (Presenter)
NYU Langone Health
Jonathan Wilt (Presenter)
Oschner Health System
Corey Chivers (Presenter)
Mark Sendak (Presenter)
Duke Institute for Health Innovation