Dec 22 2015
The imaging skills that undergraduates in Duquesne University’s biomedical engineering program used to develop a pothole-detecting drone now are being applied to spotting cancer cells.
Students tackled digital imaging and engineering issues as they designed and built a quadcopter drone that nabbed the attention of NASA and an MIT spin-off at the Maker Faire Pittsburgh in October. Constructed with 3-D printers, from purchased parts like microcontrollers and a camera, the drone features unusual energy-saving construction that propels it forward as it captures GPS coordinates corresponding with camera images of potholes.
One of the imaging challenges was to develop a way to distinguish a pothole from a manhole, a pothole from a speed bump, and a pothole from new asphalt—software and skills that can be applied to medical research, said Dr. John Viator, director of the biomedical engineering program.
“You use the same skill set—shape analysis—to identify tumors in an MRI or to identify pathological cells in blood samples,” Viator explained. “It’s an active area in research.”
Image processing is at the intersection of engineering technology and the medical arena. The students demonstrate how cells in a blood sample look the same in a computer-generated image, whether they are cancer cells or typical cells. With their software, the students can assign gradient colors based on nonvisible pigmentation differences in specialized areas of the cell and its edges. The program produces a rainbow-colored version of the cell images—showing which cells are cancerous.
Students Marc Hazur, Colin Moore, Elizabeth Petrell, Cecilia Lee-Hauser and Andrea Sajewski came up with the idea of a drone pothole patrol after learning about drones being used to inspect bridges.