According to a recent investigation, artificial intelligence (AI) has the capability to evaluate the likelihood of lung cancer in non-smokers by analyzing daily chest X-ray images through the “CXR-Lung-Risk” AI software, which identifies relevant patterns linked to lung cancer. The research results are scheduled to be disclosed at the annual conference of the Radiological Society of North America in Chicago on Wednesday. These findings reveal that individuals identified by the AI system from stomach X-rays are confronted with a twofold increase in the risk of developing lung cancer.
Dr. Michael Lu, a senior researcher at Massachusetts General Hospital in Boston, underscored the importance of this AI innovation in pinpointing never-smokers at an elevated risk of lung cancer. The objective is to utilize existing neck X-rays in electronic health records for the early detection of lung cancer, a critical initiative as smoking rates diminish.
Despite advancements in cancer treatment, lung cancer continues to be a primary cause of cancer-related mortality, with an estimated 127,000 deaths and nearly 240,000 new cases forecasted in the US this year. Strikingly, 10% to 20% of lung cancer instances arise in individuals who have never smoked, underscoring the necessity for advanced screening techniques that go beyond conventional guidelines tailored to smokers.
Although the US Preventive Services Task Force advocates low-dose CT scans for specific age groups with smoking histories, it does not endorse screening for never-smokers. However, given the escalating incidence of lung cancer in non-smokers, the creation of the CXR-Lung-Risk model utilizing AI technology offers a promising avenue for early detection.
Anika Walia, a medical student at Boston University School of Medicine, highlighted the growing occurrence of lung cancer in non-smokers and the complexities in predicting their susceptibility. Through the analysis of an extensive dataset of neck X-rays, the research team successfully identified a substantial portion of never-smokers at a heightened risk of lung cancer using AI algorithms.
The initial outcomes of the study indicate a notable number of never-smokers identified with a high risk of lung cancer by the AI model, surpassing the recommended screening thresholds. The straightforwardness and availability of additional chest X-ray imaging suggest that this approach could be beneficial in augmenting early detection endeavors.
It is essential to acknowledge that these results are preliminary and await publication in a peer-reviewed journal, emphasizing the continuous research and progress in harnessing AI to advance lung cancer screening and detection methodologies.
