PHG Foundation provides AI expertise for cancer project
2 July 2020
Project DELTA aims to diagnose up to 50% of cases of oesophageal cancer earlier, leading to improvements in survival, quality of life and economic benefits for the NHS.
The project is a collaboration between the Universities of Cambridge, Oxford, Kings College London, the PHG Foundation and Cyted. The team will be using the innovative Cytosponge™-TFF3 test - the ‘sponge on a string’ - to diagnose Barratt’s oesophagus in patients with heartburn. Barrett's oesophagus, a condition that can turn into cancer of the oesophagus, is more common in patients who suffer with heartburn. AI algorithms will be developed to identify at-risk individuals and to assist pathologists with rapid diagnoses. People diagnosed with Barrett’s oesophagus can then be monitored regularly for early signs of cancer. Oesophageal cancer takes the lives of around 8,000 in the UK each year, often because the cancer is diagnosed too late.
The PHG Foundation will be exploring the ethical, legal and regulatory challenges of embedding artificial intelligence in a personalised prevention pathway for oesophageal cancer. At one end of the pathway, this will include analysing the impact of using machine learning to identify at-risk individuals and for recruitment to screening - at the other, for the safe and efficient analysis and reporting of biopsy samples.
Sharing a £16m grant from UKRI’s industrial strategy challenge fund, Project Delta is one of six health projects from across the UK to use AI to bring together and better interpret data from multiple sources.
Science Minister Amanda Solloway said:
'The University of Cambridge project we are backing today will help ensure more lives are saved and improved, as it aims to diagnose up to 50% of oesophageal cancer cases earlier.'
Alison Hall, the PHG Foundation's Head of Humanities who is leading our contribution to the work said:
'I'm delighted that we are contributing to this exciting project to improve personalised prevention and accelerated diagnosis of cancer; machine learning has great potential to improve healthcare, but to make it work in practice it is essential to address relevant ethical and regulatory issues.