BCI team awarded over £500,000 to explore AI for earlier detection of returning liver cancer

The study will focus on people with liver cirrhosis (scarring of the liver) and liver cancer who have been treated with surgery or thermal ablation (using heat to destroy the tumour tissue). The researchers will develop AI that examines CT scans alongside information from patients’ blood tests.

The researchers will use existing patient records and follow patients over time to develop and test the AI. For patients whose cancer returned, they will examine earlier scans taken before the recurrence was diagnosed. This will allow them to investigate whether the AI could identify signs of cancer that were not recognised by radiologists at the time.

The team will combine information from CT scans with changes in a blood marker called alpha-fetoprotein, which is routinely used when monitoring some patients with liver cancer. By examining changes in alpha-fetoprotein over time alongside imaging, the team hopes the AI will be able to recognise patterns linked to very early recurrence.

One of the study’s targets is to determine whether the system can improve detection of tumours smaller than 5 mm and identify recurrence at least three to six months earlier than routine radiological detection.

The project will also include a 12-month study at Barts Health involving 80 newly treated patients to assess how the AI could work in NHS care. The AI will analyse their follow-up scans alongside the usual NHS checks, but its findings will not influence their care or doctors’ decisions. Researchers will later compare the AI’s findings with routine reports from radiologists to assess how well it performs and explore whether it could be used safely and practically in NHS care.

“Our clinical colleagues identified this challenge, and their expertise and support are central to everything we are developing,” Dr Singh says. “Their involvement will help us understand how the tool could support decisions in everyday practice.”

The project builds on earlier seed funding from Barts Charity and support through the University of Edinburgh Venture Builder Incubator, where the team received a Cancer Research Horizons prize.

The new Cancer Research UK funding will allow the researchers to develop and test the approach and assess its potential for use in clinical practice. Further clinical studies would still be required before the technology could be introduced into routine NHS care, including research to establish whether finding recurrence earlier leads to better outcomes for patients.

“Our longer-term goal is to give clinicians and patients more options when liver cancer returns,” Dr Singh concludes. “By bringing together AI expertise and the experience of clinicians caring for these patients, we hope to recognise recurrence at a point when treatment can make a greater difference.”