An Artificial Intelligence technology shows promise in predicting the danger of pulmonary nodule cancer | Technology
|An Artificial Intelligence technology shows promise in predicting the danger of pulmonary nodule cancer|
Artificial Intelligence tool shows promise in identifying lung nodule cancer risk
The new artificial intelligence program is another example of how new computer technologies can benefit doctors in treating patients.
As artificial intelligence and machine learning technologies continue to develop, they can become powerful tools in many fields, including medicine.
AI, which complements human experience and judgment, has already shown promise as a forecasting tool. Recent research using an artificial intelligence program to help identify lung cancer risk from the results of chest scans is an example of the technique in action.
Lung cancer is the second most common form of cancer worldwide, according to the World Cancer Research Fund. In Australia, it is the leading cause of cancer deaths and Cancer Australia estimates that lung cancer accounted for 17.7% of all cancer deaths in 2021.
Computed tomography (CT) is a type of three-dimensional image that is often used in the detection and diagnosis of cancer. Small abnormal areas, called pulmonary nodules, are sometimes found on chest CT scans. Most lung nodules seen on scans are not cancer, but some may be at risk of becoming cancerous growths. A key part of cancer detection is determining from chest scans the likelihood of cancer developing from lung nodules.
Study lead author Anil Vachani, director of clinical research in the section of Interventional Pulmonology and Thoracic Oncology at the University of Pennsylvania Perelman School of Medicine in Philadelphia, USA, says CT scans are more favored than x-rays for chest imaging.
“A nodule would show up somewhere between 5% and 8% of chest X-rays. The chest CT scan is such a sensitive test that you will see a small nodule in more than one-third to one-half of cases. We've gone from a problem that was relatively rare to one that affects 1.6 million people in the US each year,” says Vachani.
Vachani's team used the AI diagnostic tool developed by Oxford, UK-based Optellum Ltd., to help doctors assess lung nodules found on CT scans. AI, they argue, can go beyond the basics of a nodule, such as size and edge characteristics.
“AI can analyze very large data sets to generate unique patterns that cannot be seen with the naked eye and that end up predicting malignancy,” says Vachani.
The researchers provided six pulmonologists and six radiologists with CT imaging data from 300 scans of indeterminate lung nodules (those between 5 and 30 millimeters in diameter) to estimate the risk of malignancy. The experts were also asked to make management recommendations, such as CT checks or additional diagnostic procedures. They were asked to do this with and without the help of the AI tool.
“Readers judge malignant or benign with a reasonable level of accuracy based on the images themselves, but when you combine their clinical interpretation with the AI algorithm, the level of accuracy improves significantly,” says Vachani.
"The level of improvement suggests that this tool has the potential to change the way we judge cancer versus benign and hopefully improve the way we manage patients."
“We took the first step here and showed that decision making is better if the AI tool is incorporated into the radiology or pulmonology practice,” she says. “The next step is to take the tool and run some prospective trials where clinicians use the AI tool in a real-world setting. We are in the process of designing those trials.”