Does Artificial Intelligence Really Speed Up Lung Cancer Diagnosis?

A large study conducted in the UK evaluated whether using artificial intelligence to prioritize the analysis of chest X-rays could help detect lung cancer more quickly. Over 93,000 X-rays were examined across five different hospitals. The results show that prioritized analysis by artificial intelligence did not reduce the time needed to obtain a scan or diagnose lung cancer.

Patients whose X-rays were analyzed with the help of artificial intelligence waited an average of 53 days before undergoing a scan, compared to 53 days for those whose X-rays were analyzed without this priority. For the 558 individuals diagnosed with lung cancer, the average time to diagnosis was 44 days with artificial intelligence and 46 days without. No significant difference was observed in other stages of the care pathway, such as the time before an urgent consultation or the start of treatment.

Artificial intelligence sometimes detected abnormalities that radiologists missed, but in nearly one-third of cases, its conclusions did not match those of human experts. Among these disagreements, about one-quarter of the exams revealed results requiring medical action. However, even when artificial intelligence flagged an abnormality, it did not speed up the overall diagnostic process.

Researchers emphasize that artificial intelligence can help identify suspicious signs, but its use in prioritizing exams had no measurable impact on the speed of diagnosis. They therefore recommend not systematically integrating this method into hospitals for the time being. According to them, it would be more effective to improve the organization of care pathways, such as enabling immediate reading of X-rays by a radiologist, which had already shown promising results in a previous study.

The study also confirms that the delay between X-ray and scan often remains too long compared to national recommendations, partly due to a lack of scanners and specialized staff. The authors stress the need to better understand how artificial intelligence could be useful without overburdening medical teams or generating unnecessary alerts. They encourage further research to identify situations where this technology could truly make a difference.

Sources and Credits

Source Study

DOI: https://doi.org/10.1038/s41591-026-04253-5

Title: AI-based chest X-ray prioritization in the lung cancer diagnostic pathway: the LungIMPACT randomized controlled trial

Journal: Nature Medicine

Publisher: Springer Science and Business Media LLC

Authors: Nick Woznitza; Lesley Smith; Janette Rawlinson; Iain Au-Yong; Bindu George; Madava G. Djearaman; Arjun Nair; Richard W. Lee; Neal Navani; Siyabonga Ndwandwe; Caroline S. Clarke; Andrew Creeden; Josh Newsome; Indrajeet Das; Sylvia Abaokporo; Richard Tucker; James Hathorn; David R. Baldwin