Supporting lung cancer diagnosis with AI
PenRAD successfully bid for funding from the government’s £21m Artificial Intelligence Diagnostic Fund (AIDF) to test an AI tool designed to detect lung nodules on CT scans.
The challenge
Lung nodules - small lumps that appear on X-rays or CT scans - are common findings in diagnostic imaging. Some nodules can be the earliest indicators of lung cancer. The challenge lies in distinguishing between harmless and potentially malignant nodules quickly and accurately.
Artificial intelligence shows promise as a helpful tool to assist Radiologists with increasing volumes of chest CT scans, and to track nodule progression over time. However, there is little peer-reviewed evidence whether AI can speed up diagnosis or treatment. The purpose of the AIDF funding is to evaluate its effectiveness in real-world settings.
Our approach
After a comprehensive evaluation of available technologies, PenRAD led a robust procurement process supported by Health Innovation South West and the Peninsula Cancer Alliance. Qure.ai was selected to provide its qCT software for an initial two-year period.
The qCT tool uses artificial intelligence to overlay CT scans with AI-generated insights, highlighting the size and growth of lung nodules over time.
This helps clinicians to assess whether a nodule is stable, growing, or shrinking—critical information for determining the urgency and type of follow-up or treatment required. The AI also retrieves prior CT images for comparison, streamlining the monitoring process and improving clinical decision-making.
Impact and benefits
Since implementation, approximately 1,200 chest CT scans are being processed weekly across the region using qCT. The tool has contributed to improving the efficiency and accuracy of lung nodule reporting.
qCT is assisting in 1,000 chest CT scans per week
An evaluation is underway, independently led by Health Innovation South West, which will provide real-world evidence for how and to what degree artificial intelligence is a useful clinical tool and the potential patient benefits.
Lung nodule AI assistance helps to identify and track suspicious nodules over time. The reporting radiologist can then put any such nodules into a clinical context and provide useful information for further follow up and/or treatment.
Tej Pandher, Consultant Radiologist, University Hospitals Plymouth NHS Trust
Next steps
he programme will also contribute to the national evidence base for artificial intelligence in healthcare. Current contracts are due to expire in the winter of 2026/27, and conversations are taking place with trusts to secure a working solution.
PenRAD convenes a peninsula-wide AI sub-group which leads on horizon scanning for good practice and potential AI solutions that could meet the needs of trusts in the region.