The tool can detect tiny specks of lung cancer-infected tissue on a CT scan with a 95 percent accuracy rate, which exceeds the 65 percent accuracy rate of the human eye, according to the scientists behind the research.
Engineers at the school’s Computer Vision Research Center announced the development Wednesday.
Facial recognition and brain development in humans guided the work, said Rodney LaLonde, who was on the team.
“You know how connections between neurons in the brain strengthen during development and learn? We used that blueprint, if you will, to help our system understand how to look for patterns in the CT scans and teach itself how to find these tiny tumors,” LaLonde said in a press release.
More than 1,000 CT scans were fed into a scanner to analyze patterns and help the computer learn how to find tumors.
Non-cancerous masses and tissue, along with nerves, were ignored by the computer system.
“I believe this will have a very big impact,” UCF assistant professor Ulas Bagci said in the release. “Lung cancer is the number one cancer killer in the United States and if detected in late stages, the survival rate is only 17 percent. By finding ways to help identify earlier, I think we can help increase survival rates.”
The research can be seen here: https://arxiv.org/pdf/1805.02279.pdf.