School of Computing Research Colloquia

Process Mining in the NHS: opportunities and a new frontier

Owen Johnson, Applied Computing in Biology, Medicine and Health Research Theme

Abstract: Medical practice is increasingly protocol driven, clinicians are guided by process patterns know as care pathways which prescribe sequences of events regarded as best practice for their patients. The e-health patient record systems used by the UK NHS contain traces of these events and new data mining techniques such as process mining can extract and analyse these events to examine actual patterns, check for conformance and investigate variety. In particular it may lead to clinical insight into identifying the most appropriate clinical protocols to achieve the best health outcomes for patients. In this talk I will give a worked example of process mining of e-health data to examine patterns of care and describe the research opportunities through conventional data mining and opportunities for linking to other forms of health data and research in natural language processing, machine learning and AI. Through the Leeds Institute of Data Analytics we have access to some of the richest and largest e-health record resources in the UK, and the clinicians who use them. I hope to stimulate discussion on broader opportunities for Computing researchers looking for future challenges in medicine and health.