New AI-based method for early detection of lung cancer

A simple cheek swab could detect lung cancer years before symptoms appear, using an AI system that reads subtle biochemical changes in cells. Lung cancer is often caught late because early symptoms are vague or absent, and current screening methods like CT scans are expensive, require specialist equipment, and are not widely accessible. This project addresses that gap by developing a test that could be performed in a GP surgery, pharmacy, or even at home, using only a few hundred cells from the inside of a person’s cheek. The test, called SM-S1, combines infrared spectroscopy with machine learning to spot spectral differences between healthy and cancerous cells. In the latest clinical trial, it achieved 93% sensitivity. If successful, this could transform lung cancer screening from a hospital-based, resource-intensive process into a routine, low-cost check that catches the disease early, when treatment is far more effective. The same technology platform is also being developed to detect oesophageal and colon cancers, and could eventually analyse biopsy samples as well.