The independent news organization of Duke University

Mine detection tech could identify cancer

Undersea mine detection software developed by a Duke professor can now help doctors identify cancer-related cells.

Larry Carin, William H. Younger professor of electrical and computer engineering at Duke, recently developed active learning software with the U.S. Office of Naval Research’s support. The software—which was originally intended to identify mines underseas—assists doctors in classifying tissue samples for cancerous cells. A research team at the University of Pennsylvania, led by Dr. William Lee, associate professor of medicine, hematology and oncology, is embedding the software into an automated image analysis toolkit called FARSIGHT, or Fluorescence Association Rules for Quantitative Insight.

The combination of these active learning systems and FARSIGHT has led to both improved accuracy and speed in identifying cancerous cells in humans, Lee wrote in an email Wednesday.

“As soon as we tested the new and improved software on some images, we recognized immediately that its ability to correctly identify the type of cells we were interested in studying had improved dramatically,” Lee said.

In a Pratt news release Oct. 7, Carin noted that the software could be a “game-changer” in cancer research.

“The results are spectacular,” Carin said in the release. “There is a real chance this may save lives in the future.”

Lee said the software provides a more efficient alternative to former methods of cancer cell identification.

“In doing what humans won’t or can’t do, [the software] becomes a powerful tool for studying and understanding human cancers in ways not previously possible,” Lee said.

Looking to the long-term applications of this software, Lee noted that the identification of endothelial cells—cells which form blood vessels necessary for oxygen and nutrient delivery—would allow for new treatments in which these endothelial cells, rather than tumor cells, are the target of cancer treatment. This would effectively deprive tumors of blood vessels and blood supply, Lee said.

Jason Stack, a program officer with the Office of Naval Research, wrote in an email Thursday that it is typical for offices like the Office of Naval Research to turn to university faculty for research.

“The hardest challenges demand the smartest people, and they have to have time to think deeply about the question at hand,” Stack said. “These criteria lead us to our universities. ONR has been engaging university scientists since inception in 1946—65 Nobel laureates, so far.”

Stack noted that such a re-purposing of technology is not unprecedented.

“It is common for fundamental research conducted by ONR to lead to many additional non-defense applications even though they may not directly result in new naval capability,” he said.

The Office of Naval Research supports the technology’s adaptation, Stack added.

“Solutions to the demanding problems benefit everyone,” he said. “Good ideas which overcome technical barriers are often used in other forms in other applications as is the case with FARSIGHT.”


Share and discuss “Mine detection tech could identify cancer” on social media.