Research Topic: Use of AI to improve the early detection of upper GI cancer using routine blood tests
Supervisors: Mr Owen Johnson, Dr Lesley Smith, Prof Gordon Cook
About Victoria: I graduated from the University of Sheffield with an MEng in Bioengineering in 2020. During this degree, I specialised in Medical Devices and Systems
Project description: Cancers of the upper gastrointestinal (GI) tract are commonly diagnosed at a late stage; for example, 58% of pancreatic and oesophageal cancers are diagnosed at stages 3 and 4. This is due to non-specific symptoms and diverse presentations in patients. As late diagnosis is associated with poorer outcomes in cancer patients, there is motivation to develop methods for earlier detection. This project aims to use machine learning methods to determine whether markers and trends in routine blood tests can detect patients at high risk of upper GI cancers before they may otherwise have presented to primary care. The project will be a retrospective cohort study using electronic healthcare records to develop algorithms capable of utilising sparse and irregular data to exploit the full longitudinal records.