GE Healthcare, UW partnership sharpens medical imaging advances
When one of the nation’s top medical imaging research programs sits less than 80 miles from one of the world’s largest medical technology manufacturers, it’s no surprise the two institutions would frequently join forces.
But collaboration has taken an innovative and productive twist for GE Healthcare of Waukesha and the University of Wisconsin-Madison. Since 2001, the university and company have developed a research and intellectual property agreement that allows both company and university employees to work together as a seamless team on cutting-edge technology.
Rather than the standard project-by-project approach to university-industry agreements, UW-Madison and GE took more of an “umbrella approach” that allows for input and feedback across many lines of research and clinical settings. Patent experts with GE and the Wisconsin Alumni Research Foundation (WARF) developed predetermined intellectual property agreements meant to protect interests while also encouraging the free flow of ideas.
Jerry Shattuck, a WARF licensing manager, is one of the architects of the GE collaboration along with GE Global IP Manager Tsuri Bernstein. Shattuck says the project is without peer on campus in terms of productivity: Later this year, the project will mark its 100th patent disclosure since the start of the first agreement. The disclosures tend to be “of very high quality” since they already have industry input, he says.
Managing that much intellectual property, Shattuck says, would be impossible without the preexisting long-term relationship and good communication between the two entities.
“We meet with IP staff from GE every month to manage this huge portfolio of inventions,” Shattuck says. “This monthly meeting gives us the ability to fine-tune our partnership and make sure it’s a win-win for both sides.”
Both sides agree that the partnership, now under its second five-year contract, has been a big success story.
“I think the scope of our relationship is broader than any I’m aware of in the nation between industry and an academic group,” says Charles Mistretta, a professor of biomedical engineering, medical physics and radiology who is one of the university’s most prolific inventors, with 27 medical imaging patents.
“Both sides bring a different skill set,” says Mistretta. “They have tremendous engineering skills at GE, and here we’re putting things together with black tape and crude software. We are nowhere near being able to put something into a product.”
GE provides research grants for projects of mutual interest, and there are currently four live projects with industry backing, he says. GE also funds two post-doctoral fellows in computer tomography, where there are many joint projects, and has donated and opened research access to installed equipment.
One valuable aspect to Mistretta and others is access to “pulse sequencing,” the programming language that steers image development on magnetic resonance imaging (MRI) machines. With that access, Mistretta can evaluate how his concepts will perform on a real clinical device.
That reality check is a critical part of disseminating research into applications, he says.
“The whole idea of dissemination depends on practicality, on market demands, on clinical demands, on feedback from clinicians and so on,” Mistretta says. “GE is in a very good position to get that kind of data for us. We don’t want to be developing something that never has a chance of being disseminated and doing anyone any good. They give us a realistic perspective on where we should go.”
Mistretta notes that some of his doctoral graduates are now in high positions of leadership at GE Healthcare, and he ends up collaborating at times with GE employees who were once his students. The Wisconsin Alumni Association estimates that about 100 UW-Madison graduates work at the Waukesha GE facility.
Another novel aspect of the partnership is that GE Healthcare employs two scientists stationed at UW Hospital Clinics, working elbow to elbow with about 30 faculty and graduate students on the medical imaging team. GE Chief Scientists and Manager Jean Brittain and Senior Scientist Reed Busse have been on site for approximately 2 years.
Busse says GE’s on-site role helps “bridge the gap” between good ideas that bubble up through the research process and the engineering resources needed to develop them further. He conducts his own research and also serves a facilitative role with the UW-Madison projects.
One major assist to GE is the direct interaction with the clinical environment, Busse says. Obviously, the company builds and installs the medical machines, but is one step removed from the clinicians and patients who rely on them. The clinical feedback helps improve the functionality of existing projects, but also steers the researchers toward new challenges.
“When you talk to radiologists here, they don’t want to just push a button and get an image,” Busse says. “They want to push the envelope and to look at disease in a way they haven’t been able to before. They know the aspects of disease that they can’t see with today’s technology.”
A number of UW-Madison projects, supported both directly and indirectly by the GE partnership, are pushing that technology envelope. Mistretta, for example, is developing two new approaches to accelerating image acquisition, called HYPR and VIPR. Through a fundamental change in how images are collected, these techniques are capable of exceeding the density of current images from a factor of 10 to 1,000.
“A present scanner would take 39 hours to get what we are able to collect with HYPR in five minutes,” he says.
Similar in concept to how movies are compressed by JPEG files, HYPR and VIPR exploit the huge redundancy of information from image to image. This approach focuses on collecting only the differences in data over time and then reconstructing the redundant information, leading to the remarkable acceleration. And faster data acquisition is critical because it helps capture the most detailed images with the lowest possible x-ray exposure to patients.
Mistretta’s research has led to some of the most lucrative patents in the WARF portfolio, including an MR imaging advance in the late 1980s that is the second-highest revenue producer at WARF, behind Vitamin D. But Mistretta is convinced that HYPR and VIPR techniques have vastly more potential because they can apply to a wide range of medical imaging, including MR, computerized tomography (CT scans), positron emission tomography (PET scans), ultrasound and others.
Another promising development is from the research lab of Scott Reeder, an assistant professor of biomedical engineering, radiology and medical physics at UW-Madison. Reeder’s advance, under the acronym IDEAL, provides a way to separate water and fat out of medical images to provide a much more detailed picture of the underlying pathology. Fat signals often obscure images from MR and other technologies, especially in challenging areas of the body, such as lungs, breast tissue, and the head and neck area. The IDEAL technology is currently being applied clinically and could prove highly useful in cancer imaging and in better visualizing musculoskeletal injuries.
IDEAL will also help quantify the density of fat in the liver, one of the hallmarks of chronic liver disease. That type of test currently cannot be done without a biopsy, Reeder says.
Walter Block, an associate professor of biomedical engineering and medical physics, is also developing fast-imaging applications to look at critical structures such as cartilage, work that can lead to a clearer picture of complex knee injuries. Block’s work was funded by a grant from the Walter H. Coulter Foundation, which provides UW-Madison biomedical engineers with $1 million per year for research that shows potential for rapid transfer to the physicians and patients who need the advances through commercialization of their innovations.
Block says UW-Madison is one of only nine universities in the country to receive such Coulter funding, and the GE affiliation made the department a stronger candidate for this award by demonstrating how UW-Madison can rapidly move research into commercial use. Like Coulter, GE is trained on advances that show commercial and clinical potential.
Block says that with the GE collaboration, “perspective is key,” noting that academic and corporate researchers can have very different ways of looking at the same challenge. He offers this hypothetical example:
“In an academic paper, if you demonstrate that your process works nine times out of ten, that’s going to be a strong paper,” Block says. “But if you put an industrial product out that works nine times out of ten across an installed base of 10,000 scanners, you’re looking at a disaster. Having GE right here, as well as an hour down the road, makes us much more in tune with understanding how the business works.”
Stephen Harsy, associate dean of the UW Medical School, says the GE project “is by far the most significant relationship the medical school has with a company.” He says there is evidence that the medical imaging departments have become more competitive in getting National Institutes of Health (NIH) grants because of the strong industry link and access the latest-generation technology.
The first five-year master agreement yielded about 50 different project statements, Harsy says. The second five-year agreement started in September 2006, and Harsy says it is likely to be a period of greater commercialization of some the ideas nurtured in the first round.
Shattuck agrees that the future of the project should yield greater rewards. “MR and CT imaging have been the biggest areas to date,” he says, “but we are trying to have the same success in other imaging modalities.”