Artificial intelligence can effectively recognize high cancer risk in non-smokers from lung X-rays. The research supporting this will be presented at the annual meeting of the Radiological Society of North America (RSNA).
Although smoking is considered a major risk factor for lung cancer, approximately 10-20% of lung cancer cases occur in never smokers, that is, people who have never smoked or have smoked fewer than 100 cigarettes in their lifetime.
In a new study, researchers from Harvard Medical School developed the “CXR-Lung-Risk” deep learning algorithm using 147,497 chest X-rays obtained from more than 40 thousand people.
The researchers tested the model on 17,407 patients with an average age of 63. AI classified 28% of them as people at high risk of lung cancer. Even after adjustments for sex, age and previous lung infections, the risk of lung cancer was 2.1 times higher than in the low-risk group. 2.9% of these patients were later diagnosed with lung cancer. This incidence rate means that these patients should be offered cancer screening using CT, a more informative test.
This study proves that AI can be used to identify non-smoking patients who need more comprehensive lung cancer screening.
Previous scientists to create A way to detect stomach cancer with 90% accuracy.