AI Better at Diagnosing Lung Cancer Than Doctors—US Study

AI

Artificial Intelligence (AI) is better at diagnosing lung cancer compared to specialist doctors, a collaborative study suggests / Photo by: Seksan Mongkhonkhamsao via 123RF

 

Artificial Intelligence (AI) is better at diagnosing lung cancer compared to specialist doctors, a collaborative study suggests, in which researchers hope AI could enhance the effectiveness of cancer screening.

Researchers at Northwestern University in Illinois and Google Inc. believe that AI would play a significant role in the future of medicine, even though present software is not yet ready for clinical application.

They opted to focus on lung cancer, which kills more people (1.8 million) compared to any other type of cancer. The sheer rate alone pushes the United States government to recommend screening for people who are at high risk due to years of heavy smoking, BBC News reports.

However, these screenings could result in invasive biopsies for those who turn out to be clear of the illness and could also miss some tumors. The researchers used AI to see if it can improve the analysis of these scans.

The team trained the software with 42,290 CT lung scans but did not program it to look for anything specific, and just taught the technology which patients developed cancer and which did not. BBC News says the researchers found that AI improved diagnosis of cancer by five percent while also reducing false-positives (people falsely diagnosed with the ailment) by 11 percent.

"The next step is to use it on patients in the form of a clinical trial," Mozziyar Etemadi, from Northwestern University, told the news site. Etemadi said the AI uses a "little bit of a black box" in identifying cancer.

"Sometimes it highlights a lung nodule (a growth) that for all intents and purposes looks benign but the model thinks it isn't. It's usually correct and one area of scientific inquiry is figuring out why," he added.

Etemadi also said that collaborative work between AI and doctors would bring about more enhancements in diagnosing illnesses and that AI plays a "huge" role in medicine.