Lung Cancer AI Study Suggests Its True Benefit Might Be Medical

Just because someone is a nonsmoker doesn’t mean they’re free from the risk of developing lung cancer. According to the American Cancer Society, between 10% and 20% of lung cancer patients are lifelong nonsmokers. The reasons why they develop cancer are varied, but as the organization notes, there are plenty of environmental hazards that can increase the risk of cancer that have nothing to do with cigarettes.

This also poses a challenge to health care professionals in determining who should be screened for lung cancer. Screening people with a long history of smoking makes sense; trying to figure out which nonsmokers should be screened is trickier. All of which makes a recent study involving machine learning that much more promising as far as its implications are concerned.

“[L]ung cancer is increasingly common in never-smokers and often presents at an advanced stage,” said researcher Anika S. Walia, the lead author of a new study that used AI to analyze chest x-rays of thousands of people classified as “never-smokers.” The study utilized deep learning and ultimately designated 28% of the patients as high risk; 2.8% of them went on to be diagnosed with lung cancer. That’s more than double the six-year risk threshold for which the National Comprehensive Cancer Network recommends lung cancer screenings.

“This AI tool opens the door for opportunistic screening for never-smokers at high risk of lung cancer, using existing chest X-rays in the electronic medical record,” said Dr. Michael T. Lu, another of the study’s authors, in a statement. Dr. Lu also pointed out that with fewer people smoking cigarettes, this technology would likely become even more valuable in the years to come.

Could AI and Machine Learning Expand What We Know About Alzheimer’s?

It’s part of a growing medical trend

As valuable as this study is on its own, what it represents is even more promising. There’s been a lot of discussion of how good (or not) AI might be at creating text or images. Here, though, it’s looking for patterns that humans might not pick up at all — and potentially improving human health as as result.