Clinithink and AstraZeneca launch AI project for lung cancer detection

Clinithink and AstraZeneca Oncology UK have teamed up for a new project that uses artificial intelligence to identify patients with early stage lung cancer, so they can receive early treatment for better patient outcomes.

The project aims to flag patients who could benefit from lung disease screening, in the hope that the condition is identified early so it can be treated.

As well as potentially improving the outcome for patients, it will also reduce costs for the NHS. According to Cancer Research UK treating patients at a later stage incurs higher costs.

Dr Satoshi Hori, oncology medical affairs head at AstraZeneca Oncology UK, said: “Addressing healthcare ecosystem challenges to enable earlier detection and diagnosis of cancer is one of our UK oncology missions.

“Focusing specifically on early detection of lung cancer, our partnership with Clinithink is a great example within AstraZeneca Oncology of an external partnership with the common goal of improving UK cancer outcomes.”

The first stage of the project involves a retrospective analysis of patients’ unstructured electronic medical records to determine if it’s possible and cost-effective to use the AI technology. If it is deemed cost-effective and scalable the team will develop a prospective model in mid-2023 to prove its validity in the real world.

The technology will use both machine learning and natural language processing. These will be used to identify both symptomatic and asymptomatic patients who are in the early stages of lung cancer. Using the outputs, the team will then develop predictive models that can flag high-risk individuals.

Chris Tackaberry, co-founder and CEO at Clinithink, said: “Harnessing and understanding unstructured medical data creates enormous opportunities to transform the treatment of disease, reduce NHS costs and improve population health.

“Our technology, CLiX unlock, can process millions of detailed medical records in hours – a process which would take years if completed manually – to deliver valuable clinical insights.

“We hope to use the insights uncovered in this real-world evidence study to develop predictive models that will flag high-risk individuals at a much earlier, more treatable, stage of disease – when neither they nor their GP know they have lung cancer, or even suspect it.”

Last summer Portsmouth Hospitals University NHS Trust announced it was trialling AI technology from Sectra to help detect lung cancer in chest x-rays.