New AI Program Identifies Lung Cancer

New artificial intelligence tool can identify tumors at a 97% rate, a year before other methods can diagnose it.

Combined with a computerized tomography (CT) scan, an artificial intelligence (AI) program can spot signs of lung cancer a year before it can be diagnosed with other existing methods, according to research presented at the European Respiratory Society International Congress.

“Screening for lung cancer would mean many more CT scans being taken and we do not have enough radiologists to review them all. That’s why we need to develop computer programs that can help,” Benoît Audelan, an investigator in the Epione project team of the Inria (France’s National Institute for Research in Digital Science and Technology) centre at Université Côte d'Azur, said in a statement. “The objective of our research is not to replace radiologists but to assist them by giving them a tool that can spot the earliest signs of lung cancer.”

The AI program was able to identify 172 of 177 malignant tumors in CT scans, making it 97% effective, the results of the study showed.

The AI program also identified 152 areas a year earlier that were later diagnosed as cancer in the same individuals.

The study results also showed that the AI program did identify areas that were not cancer, so it would need to be improved before being used in clinics, according to the statement.

CT scans have been used to spot signs of lung tumors, followed by a biopsy or surgery to determine if the tumor is malignant, but a radiologist must review approximately 300 images to look for signs of cancer, which can be difficult, the statement said.

The AI program aims to alleviate some of the process for radiologists.

Reference

Artificial intelligence could help diagnose lung cancer a year earlier. EurekAlert. News release. September 7, 2021. Accessed September 8, 2021. https://www.eurekalert.org/news-releases/927387