Feds greenlight Infervision’s deep learning tool for segmenting lung CT scans

Technology firm Infervision recently earned 510(K) clearance from the U.S. Food and Drug Administration for a tool that aids radiologists in segmenting lung CT scans.

The firm—with a North American headquarters in Philadelphia—said its InferRead Lung CT.AI product uses deep learning technology that can also help clinicians to pinpoint nodules. Currently, about 380 hospitals and imaging centers are using the algorithm across the globe, officials said Thursday.

Infervision said the system is designed to support concurrent reading and can boost physicians’ accuracy and efficiency. They’re also targeting it toward lung cancer screening programs, to aid in finding hard-to-spot smaller nodules.

"The tremendous potential for lung cancer screening to reduce mortality in the U.S. is very much unrealized due to a combination of reasons,” Eliot Siegel, MD, a professor in radiology at the University of Maryland School of Medicine, said in a statement. He added that the algorithm could potentially act as a “robust lung nodule ‘spell-checker’” that reduces rad read times.