The lung cancer is the most common cause of cancer death, with some 1.8 million lives lost worldwide every year. Therefore, a new study has focused on creating an artificial intelligence (AI) program capable of detecting signs of this disease a year before they can be diagnosed with existing methods.
According to research presented at the International Congress of the European Respiratory Society, this new technology uses CT scans to detect these symptoms of lung cancer , a disease that is usually diagnosed at a late stage, when treatment has less chance of success.
The researchers hope that the use of AI to support lung cancer screening can streamline and streamline the process and ultimately help diagnose more patients at an early stage.
The AI study process and its results
Computed tomography or CT scan is already used to detect signs of lung tumors, followed by a biopsy or surgery to confirm whether the tumor is malignant. But in each scan, an expert radiologist looks at about 300 images and looks for signs of cancer that can be very small.
Computed tomography trials to detect people at high risk for lung cancer have shown promise, but screening is hampered by the practical difficulty of a radiologist going through each image , one at a time, to decide who needs further testing.
The new study has been presented by Benoît Audelan, a researcher from the Epione project team at the Inria center (French National Research Institute for Digital Sciences and Technologies) of the University of the Côte d’Azur, who worked with colleagues from the University of the Côte Azul, Therapixel (software company specialized in artificial intelligence for medical images) and the University Hospital of Nice.
The researchers trained their AI program on a set of CT scans of 888 patients who had already been examined by radiologists to identify suspicious growths.
They then tested it in another set of 1,179 patients who were part of a lung screening trial with a three-year follow-up , using CT scans that were done in the last two years of the trial. Among them were 177 patients who were diagnosed with lung cancer by biopsy after their last scan in the trial.
The program identified 172 of the 177 malignancies on those scans , meaning it was 97% effective at detecting cancers. The five tumors the program did not detect were near the center of the chest, where the tumors are more difficult to distinguish from healthy parts of the body.
The researchers also tested the program in scans conducted a year before tumors were diagnosed in the same 1,179 patients, and it was able to identify 152 suspicious areas that were later diagnosed as cancer.
However, the researchers say that the program also identifies too many suspicious areas that are not cancer (false positives) and that it would have to be improved long before the program could be used in the clinic, since investigating all of them would lead to unnecessary biopsies.
According to Audelan, “Lung cancer screening would involve doing a lot more scans and we don’t have enough radiologists to review them all. So we need to develop computer programs that can help. Our study shows that this program can find possible signs of lung cancer. up to a year before, “he adds.
“The goal of our research is not to replace radiologists, but to help them by providing them with a tool that can detect the first signs of lung cancer,” he explains.
The researchers plan to work on a new system to better differentiate malignant from non-malignant tissue to help radiologists decide which patients should be further investigated.
“Diagnosing lung cancer earlier is vital to improving survival rates.”
Professor Joanna Chorostowska-Wynimko, secretary general of the European Respiratory Society and consultant for respiratory medicine at the National Institute of Tuberculosis and Lung Diseases in Warsaw (Poland), who was not involved in the research, points out that “earlier diagnose lung cancer it is vital to improve survival rates and screening would be an important step towards that goal. Research shows that CT screening could reduce lung cancer deaths, “he stresses.
“This work is promising because it shows that AI could help us review many scans quickly and even detect signs of cancer at an earlier stage. However, before this program can be used, researchers will have to improve their ability to distinguish between lung tissue that is abnormal but benign and that which is probably cancerous, “he adds.