Internship project: Accuracy of machine learning for detecting deteriorating patients

Opportunity

The project aims to compare the accuracy of clinical alerts detected via a machine learning algorithm with those recorded in a clinical information system.

Project Objective

• Describe the characteristics of machine learning alerts being triggered by Electronic Medical Record data.
• Characterise what triggered the machine learning alert using Electronic Medical Record data.
• Evaluate the accuracy of a machine learning alert for recognising deteriorating patients.
• Develop recommendations of how machine learning can be used to support clinical practice related to deteriorating patients.