UT, ORNL researchers pursuing technology to help those with diabetic retinopathy

Two scientists from two different research institutions at opposite ends of Tennessee are doing something many think is impossible – combining their respective intellectual property to start a company.

Ed Chaum is the Plough Foundation Professor of Retinal Diseases and Director of the Retina Service at the Hamilton Eye Institute at the University of Tennessee Health Science Center (UTHSC) in Memphis. Ken Tobin is Director of the Measurement Science and Systems Engineering Division at Oak Ridge National Laboratory (ORNL) and an ORNL Corporate Research Fellow.

The duo had not met until a 12-person UTHSC delegation came to ORNL in 2004 to explore possible collaborative opportunities. Eight years later, the two are well on their way to making a difference for those facing diabetic retinopathy, the leading cause of blindness among working age adults. Chaum is the company’s President and Chief Medical Officer, while Tobin serves as Chief Technical Officer.

In a recent interview with teknovation.biz, Chaum vividly recalled the 2004 trip that was designed to answer the question, “What are the unmet needs and opportunities open to technological innovation that would allow UT to leverage ORNL now?”

For Chaum, the biggest challenge he faced was solving the “grossly inefficient” way in which patients are screened for diabetic eye disease.

“There are 25 million diabetic patients in the U.S. today, and we are unable to screen even half of them,” he said.  That number will grow to 125 million in 30 years,” The task of screening is mammoth. Worldwide even today, we need to be looking at a billion eyes a year.”

Chaum admits that he was somewhat skeptical that the ORNL visit would help him address his needs. In his view, the solution to addressing the volume challenge involved use of the Internet, a robust database and specially designed algorithms.

“They split us into three groups,” he said. By chance, the ORNL presenters in the group included Tobin, a nationally recognized scientist for his work in computer vision and machine learning…

As Chaum listened to Tobin’s presentation on content-based image retrieval, he had an “Aha! moment”. He found significant alignment between “what Ken was doing as an engineer and what I was doing subconsciously as a clinician.”

“I thought this was exactly the way we should be screening for diabetic eye disease, using content-based methods,” Chaum said.

“It seemed like a perfect ‘mashup’ of technology and need,” Tobin said. “Content-based methods bring informatics to large image databases so that you can find – and predict – things based on how images look. Whether it is diagnosing semiconductor device manufacturing errors or disease in the human retina, the technology is a perfect fit.”

The partnership that was formed in 2004 quickly took-off as ORNL provided funding to help advance the development of the database of images through the Laboratory-directed Research and Development (LDRD) fund.

“We built a library of 400 images and began testing and validating the algorithms,” Chaum said. He admits that timing has a good deal to do with success.

“Nine months into our LDRD grant, the National Eye Institute (NEI) issued an RFP for exactly what we were working on,” Chaum said. The team secured one of three NEI grants in 2005 and leveraged it into a second grant to analyze how to use metadata to strengthen diagnosis capabilities, e.g., data such as a patient’s age or how well they are controlling their sugar.

To further advance their work, the duo needed to deploy cameras in the field and deliver patient care. The Plough Foundation, which endowed Chaum’s chair at UTHSC, agreed to purchase and deploy five cameras. The Delta Health Alliance in Northwest Mississippi secured a federal Health Resources and Services Administration grant to deploy an additional six cameras in the region.

These grants launched TRIAD, the acronym for the Telemedical Retinal Image Analysis and Diagnosis network. More than 5,000 patients from the underserved, health disparity communities in three states have been screened through TRIAD thus far.

Along the way, Chaum and Tobin decided to found the company that is now known as Hubble Telemedical. For two years, they self-funded the start-up as they better understood the business model. In December they secured their initial outside investment from Memphis-based MB Venture Partners and landed Chuck Witkowski, a well-known Knoxville entrepreneur, as Hubble’s Interim Chief Executive Officer (CEO).

“Telemedicine is currently an infrastructure highway with little sustainable traffic,” Chaum says in explaining how Hubble Telemedical could transform the underutilized “lanes” with its revolutionary approach for screening diabetic patients for retinal diseases.

The company’s goal is “real-time diagnostics in 90 seconds or less.” The system works by installing commercially available cameras in the offices of primary care physicians. Pictures are made of the retinas of diabetic patients, and their medical information and photos are immediately sent over the Internet to Hubble where the photos are currently read by Chaum and his colleagues.

“We want to leverage the technology into a more fully automated approach,” Chaum says, explaining this goal requires U.S. Food and Drug Administration approval. Retinal images will be processed “to generate content-based image retrieval search indices from the patient archive used to estimate the statistical probability of a specific diagnosis.” The computer-generated diagnosis will be reviewed by an ophthalmologist, confirmed or modified, and returned to the primary care physician with a patient management plan.

Chaum and Tobin are justifiably proud of what they have accomplished. “We’ve gone literally from an idea to a venture-backed company in about seven years,” Chaum says, adding that they also secured $7 million in grants and several national technology innovation awards in the process.

“Ed and Ken have done all this work,” Witkowski said. “We are now in the thick of transitioning from a research to a commercial enterprise.”

There’s still much work to do, but the founders and their Interim CEO are optimistic that some exciting news will be announced before the end of 2012. And for those who want to see more collaboration among the generators of intellectual property, it will validate their beliefs.

Stay connected with us on Twitter and LinkedIn. Article ideas and other suggestions should be sent to tballard@pyapc.com. Include the name and contact information (phone and email) for follow-up.