AI system detects lung nodules in CT images in 20 seconds

A research team out of Russia says it has developed a new artificial intelligence (AI) system for treating lung cancer patients that can interpret CT results in 20 seconds.

The system, named Doctor AIzimov as a tribute to author Isaac Asimov, is installed directly on a user’s computer and can provide images where findings are marked for easy visibility. Researchers from Peter the Great St. Petersburg Polytechnic University (SPbPU) in St. Petersburg, Russia, noted that their software has detected lung nodules as small as 2 mm.

“Initially, we set up an algorithm to search for nodules starting from 6 millimeters, because radiologists themselves start the treatment of tumors of this size,” Lev Utkin, head of the SPbPU Research Laboratory of Neural Network Technologies and Artificial Intelligence, said in a prepared statement. “But the system is so smart that it was able to find nodules of even smaller size.”

Doctor AIzimov uses a patented “chord method” for lung cancer classification, producing histograms that the system analyzes. It was trained with 1,000 CT images from various datasets, and the researchers say they are currently collecting data for their own dataset.

Also, while the system currently interprets CT images in 20 seconds, the team thinks it can shorten that to just two seconds over time.

Radiologists from the St. Petersburg Oncological Center also assisted with the research, according to the statement.