New technology overcomes false-positives in CT for lung cancer

Volume 7, Issue 3, 2018


Researchers used radiomics to test variables to distinguish a benign nodule from a cancerous nodule.

Tobias Peikert, M.D.

Tobias Peikert, M.D.

A team of researchers that includes investigators from Mayo Clinic has identified a technology to address the problem of false-positives in computerized tomography (CT)-based lung cancer screening. The team's findings were published in PLOS One.

"As physicians, one of the most challenging problems in screening patients for lung cancer is that the vast majority of the detected pulmonary nodules are not cancer," said Tobias Peikert, M.D., a pulmonologist and critical care specialist at Mayo Clinic in Rochester, Minnesota. "Even in individuals who are at high risk of lung cancer, up to 96 percent of nodules are not cancer."

False-positive test results cause significant patient anxiety and often lead to unnecessary additional testing, including surgery, Dr. Peikert said. "False-positive lung cancer screening results also increase health care costs and may lead to unintentional physician-caused injury and mortality," he said.

To address the problem of false-positives in lung cancer screening, Dr. Peikert and Fabien Maldonado, M.D., of Vanderbilt University, along with their collaborators, used a radiomics approach to analyze the CT images of all lung cancers diagnosed as part of the National Lung Screening Trial. Radiomics is a field of medicine that involves extracting large amounts of quantitative data from medical images and using computer programs to identify disease characteristics that can't be seen by the naked eye.

Researchers tested a set of 57 variables for volume, nodule density, shape, nodule surface characteristics and texture of the surrounding lung tissue. They identified eight variables that enabled them to distinguish a benign nodule from a cancerous nodule. None of the eight variables were directly linked to nodule size, and the researchers didn't include any demographic variables, such as age, smoking status and prior cancer history, as part of their testing.

Dr. Peikert said that while the technology looks promising and has the potential to change the way physicians evaluate incidentally detected lung nodules, it still requires additional validation.