Lung cancer is the leading cause of cancer death in Canada. In fact, an estimated 1 in 14 Canadians are expected to be diagnosed with lung cancer in their lifetime. That’s why detecting the disease early, when treatment is more likely to be successful, is so important.
However, there are barriers to early detection. Existing guidelines recommend lung cancer screening only for current or former smokers, which excludes the significant number of non-smokers who are diagnosed with the disease. And CT scans, which effectively detect the presence of incidental pulmonary nodules (IPNs) in the lungs – indicating an increased risk of lung cancer – are often not followed up on appropriately.
With donor support, Canadian Cancer Society (CCS)-funded researchers Drs Renelle Myers and Rayjean Hung are developing an artificial intelligence (AI) powered test that identifies biomarkers in breath and blood that indicate lung cancer. At the same time, they will also create an AI method to determine if IPNs detected by CT scans are cancerous or non-cancerous.
If successful, these tests will improve the early detection and diagnosis of lung cancer for more people, saving more lives.
“AI provides promising potential to improve lung cancer early detection,” say Drs Myers and Hung. “We can use machine-learning analytics to detect tumour signals in data and use deep learning to predict imminent tumour occurrence based on CT images. This will reduce health care resource utilization, not only improving patient care but optimizing healthcare systems across Canada and possibly the world.”