Detecting and Treating Disease with Light

Duke engineers are lighting the way to a new era in medicine—literally.

As one of the world’s leading biophotonics research teams, these  specialists focus on harnessing the unique properties of light to detect and manipulate biological materials. With colleagues from across Duke’s medical center and university, they are developing a spectrum of solutions to pressing problems in health care, from diagnosing and treating cancer, to guiding and improving surgery, to designing new tests to identify telltale biomarkers of disease.

Joseph Izatt, program director for biophotonics at Pratt’s Fitzpatrick Institute for Photonics, and collaborators are using optical coherence tomography (OCT) to  improve treatment of serious eye diseases. Early in his career Izatt helped develop the first clinically viable OCT, an optical imaging method that uses near-infrared light to  capture diagnostic images from inside the eye. Today, it’s a $1 billion-per-year industry and standard technology for diagnosing  retinal problems—but its potential is hardly exhausted.

In 2007, Izatt and his company, Bioptigen Inc., made the technology portable, designing a handheld device 40 times faster than existing OCT equipment. Suddenly, ophthalmologists could obtain previously elusive diagnostic images, such as retinal views in infants to  detect neonatal vision loss.

Recently, his team created a way for OCT to profile individual cell membranes with nanosecond spatial and millisecond temporal resolution. They used this technology to monitor the “heartbeat” of individual heart muscle cells growing in culture.

Their next goal, he said, is to extend OCT for image guidance of retinal microsurgery. While OCT doesn’t yet guide operations in real time, it is impactful, said one of Izatt’s collaborators, Duke retinal surgeon Cynthia Toth.

“OCT images give you more information about the retina before you begin surgery,” she said. “So, you operate smarter, and it translates into better patient outcomes.”

In particular, she said, OCT assists surgeons in repairing macular holes that form at the retina’s center and impair vision. The enhanced, imaging-provided detail helps surgeons select the most appropriate method to treat each hole and offers insight into why and how it formed in the first place.

The team is also working to help OCT fill a critical and growing need in cataract surgery, one of the most often-performed of all surgeries worldwide, Izatt said.

“In cataract surgery, the surgeon replaces an older patient’s cloudy lens with a plastic one, but in order to choose the new lens correctly he or she first needs to know the power of the patient’s cornea,” he said. “The current instruments for measuring corneal power were not designed for people who have had vision correction surgery such as LASIK. Adapting the three-dimensional imaging power of OCT in the front of the eye will help surgeons pick the right replacement lens in this growing population.”

Since its invention just over 20 years ago, the imaging speed and resolution capabilities of OCT have increased manifold. But each new generation of these improvements causes a sharp increase in the size of imaging datasets, said Sina Farsiu, director of Duke’s Vision and Imaging Processing (VIP) Laboratory. The hardware produces an extraordinary amount of micron-level information that cannot be analyzed or searched without computer assistance.

“Ophthalmologists get minutes to spend with patients, and they don’t have the time to analyze all the images from a scan,” he said. “They want data to be compressed into a meaningful quantitative measurement – that’s what I do.”

Farsiu’s computer-automated algorithms mine the diagnostic images Izatt’s tools capture, searching for and quantifying disease biomarkers. His latest improvements could be particularly effective for diabetes-related blindness, he said. Currently, treatment is trial-and-error. Physicians try therapies until one works. Using Izatt’s hardware and Farsiu’s algorithm,the team aspires to image patients to help physicians immediately identify the best biomarker-based treatments.

Very soon, Farsiu’s algorithms may need to process even more data. With a new multi-center grant headquartered at Duke from the National Institutes of Health, the team is hard at work on the latest hardware improvements to improve OCT imaging speed by another factor of 20. This research will produce more images, and make them clearer, by removing any motion artifacts that blur scans.

These technological developments have more than clinical impact – they’ve also wooed talented students interested in biomedical and electrical engineering research to Pratt, said Izatt, who also helps administer the Fitzpatrick Institute of Photonics fellowship programs that provide funding for selected students.

“For our projects, we’ve had to not only invent technologies to bring these capabilities to physicians, but we’ve also have to increase our understanding of how body systems such as the eye work,” he said. “Working on these endeavors gives students excellent training in biomedical science, optical engineering, and signal/image processing, because it takes a combination of those disciplines to make these advancements.”