On the corner of Vanderbilt’s Medical Center Drive and 21st Avenue is a research institute that houses what is likely the single largest, most comprehensive imaging center in the country. The Vanderbilt University Institute of Imaging Science puts the most advanced imaging techniques literally at the doorstep of Vanderbilt University Medical Center physicians who want to find new ways of advancing medicine, education specialists at Peabody College looking to improve learning methods, and engineers researching new technologies.

John Gore

One of the most highly regarded research groups in the world in biomedical imaging, the imaging institute explores virtually every aspect of imaging science, from the underlying physics of imaging to applications of imaging techniques to detect, diagnose and treat disease.

“We have the tools that provide the right information and we don’t have the constraints of being spread around,” Gore says. “We were fortunate that the administration had a vision for how everything should be knit together.”

That vision was implemented by Gore, the renowned imaging expert who came to Vanderbilt from Yale in 2002 to set up the institute. Gore assembled a team of the best imaging scientists to develop advanced research using the latest, most powerful imaging equipment available.

In summer 2010, the center received a $3.45 million federal stimulus grant to purchase a new magnetic resonance imaging (MRI) scanner to study small animals. The 15 Tesla scanner (one Tesla is roughly 20,000 times the strength of the magnetic field of the Earth) can produce detailed images of the brain and body, as well as measure minute levels of key compounds. It will be used in noninvasive studies of genetically engineered mice and other small animal models, creating opportunities for breakthrough research in basic understandings of cancer, diabetes and brain disorders, along with other possibilities.

The institute already houses seven powerful magnets including a 7 Tesla human scanner, one of only 13 in the world being used in human studies. Built in 2006, the building includes four floors of research, classroom and office space. Great care and planning went into every facet of the $19.7 million facility, including creating a mock MRI scanner so human research subjects can adjust to the feeling of lying prone in the cylindrical tube. The center also houses other advanced imaging devices, including a 3-T whole body scanner, X-ray, ultrasound, PET, optical and CT scanners.

The state-of-the-art center provides ready access to the latest equipment while facilitating constant communication among potential collaborators.

“We’re not a lab living in a bubble,” says Mark Does, associate professor of biomedical engineering. “(Interaction is) how we identify the relevant questions. We were built from day one for greater collaboration.”
Currently, 24 faculty members and 50 graduate students, mostly engineering majors, are associated with the institute. In addition to addressing questions brought by “visionaries,” institute faculty conduct their own research, much of it designed to improve imaging technologies and methods and to use them in new applications.

“We’ve got the hammer and we’re often looking around for the right nail,” says Adam Anderson, associate professor of biomedical engineering. “We rely on the rest of the university to bring us novel questions. We develop methods, and use them to provide new knowledge, but there are many physicians or educational researchers who study problems with which imaging can help. We rely on them to describe what they need.”
Among the research projects ongoing at VUIIS are studies furthering understanding of the human brain and how it functions, including asking why gray matter is gray and white matter is white, and building tools for understanding how the brain is organized. Other work involves better understanding blood flow in tumors, how the brain changes when a psychiatric disorder is present, and the impact of certain medicines on brain function.

“A lot of projects here draw heavily on engineering and the applied sciences to make them work,” Anderson says. “There are a lot of algorithms being developed to help the interpretation of information from different modalities.”

“I love being involved in fast-paced, high-risk, high-reward startups,” says Limp, BS’88, a successful entrepreneur and chief operating officer of BrightKite, a social networking Web site. Limp, who earned his degree in computer science from the School of Engineering, specializes in ventures in the high-tech arena.

Good entrepreneurial ideas abound, he says. The tricky part is crossing the space between an idea and getting to market with a product no one else has, one that customers need. “To be a successful engineer/entrepreneur you have to have experience in all parts of business,” the Silicon Valley veteran advises, as well as a willingness to stake your name and reputation on a product.

No Shortcuts

Limp has made a career of such experiences. Just out of VUSE, he spent eight years at Apple Computing. Then he was at Liberate Technologies, serving as vice president of marketing, later chief strategy officer. Next he was vice president of business development for PalmSource and then a venture partner with Azure Capital Partners, which backs entrepreneurs, including BrightKite.

“Having a holistic view of what it takes to run a business is fundamental. That’s where my Vanderbilt engineering education is an asset. It’s provided a balance between theoretical and practical,” Limp says. “The practicality of the curriculum has paid dividends countless times as I apply core problem-solving skills to startups that are changing dramatically daily.”

Taking Risks

Currently, Limp is also collaborating on a Web site called Education.com, which provides a consolidated reputable resource for parents about educating children in the same way WebMD consolidates medical information. “It’s averaging 2 million hits a month,” he says. “Advertisers love it because they can get to customers they want—moms and dads interested in saving time.”

But Limp says not every hit is a home run. He helped develop a palm-sized computer that fell flat. “It was a brilliant piece of electrical engineering. It failed because consumers went to lower price points with bigger keyboards,” he says philosophically. “In the end, we sold the assets. Yet I wouldn’t trade time there for anything.”

The hallmarks of the engineer/entrepreneur are a combination of courage and unrelenting focus, he notes. “Don’t chase the next shiny penny,” he advises. “Sure, the next idea is out there, but its success is predicated on doing everything you can to make the current one a success.”

Things break. Forbes magazine says breakage costs American manufacturers $30 billion a year in warranty payments. If manufacturers could predict breakage and adjust warranties, they could save more than double that amount—not to mention the other benefits they’d reap from improved reliability, performance and quality.

Robert Tryon, PhD’96, holds scaled-down aircraft turbine disks before and after testing by VEXTEC.

Anticipating breakage on everything from jet engines and health care devices to energy technology is the entrepreneurial niche that gave birth to VEXTEC. Founded by Bob Tryon, PhD’96, chief technology officer, and Animesh Dey, MS’94, PhD’96, chief product development officer, along with CEO Loren Nasser, VEXTEC is a front-runner in using a computational framework for simulating and predicting the breakdown of manufactured products—and the potential impact on a company’s finances.

“Like the DNA of any living cell, every material has a microstructure that determines its behavior. This is the key to predicting how and when failure will happen,” says Tryon, a material science engineer who previously worked at General Motors and Ford. At the core of VEXTEC’s success is a patented system that simulates the life expectancy of a part, an engine or an entire product fleet. The system then combines that information with data from other components in a product to predict the product’s overall reliability. Clients range from manufacturers to oil and gas, aerospace and electronics companies.

In 2009, Forbes lauded VEXTEC as America’s most promising company, predicting its Virtual Life Management (VLM) product for forecasting failure will hasten the pace of innovation in its market niche.

Entrepreneurs in Reliability

“It’s hard to be an entrepreneur at GM,” shrugs Tryon, who worked first in the gas turbine division at GM. Concerned about product reliability, GM sent him to Vanderbilt for a doctorate in engineering and to study under Thomas A. Cruse, then H. Fort Flowers Professor of Mechanical Engineering and the guru of probabilistic structural analysis methods. At VUSE, Tryon met fellow doctoral candidate and computational reliability modeling expert Dey, studying under Professor Sankaran Mahadevan. Under Cruse and Mahadevan’s tutelage, Tryon and Dey saw a future ripe with possibilities.

When Tryon and Dey graduated in 1996, the sale of the GM division in which Tryon had worked made him a free agent. Initially, he and Dey honed their chops as consultants. Their first customer? Chrysler.

Their work took two paths: material science experts focused on developing a virtual material simulation tool, while computational experts focused on correlating the simulation from material behavior to predicting fleet performance and business impacts. In 2000, the two joined with Loren Nasser to found VEXTEC. Nasser husbanded the management and infrastructure of the self-funded firm while Tryon and Dey focused on refining the VLM product. Today, the three still own 100 percent of the company, which has 28 employees and posted $3 million in sales in 2008.

Animesh Dey, MS’94, PhD’06, shows a mini-aircraft engine used to test for structural failure of the disks.

Custom-tailored Sales and Service

In providing the only accurate and efficient computational framework for simulating and then predicting product behavior, VEXTEC’s VLM has the potential of eliminating the need for trial-and-error product testing. It also provides insight about products and how to make them better.

Dey says listening to individual customers is critical in helping those customers improve their products. “Don’t tell customers what you know,” he says. “Tailor your message to what they need and want to know.” Dey customizes VLM to meet the exact needs of each client and industry. The challenge is to focus, identify need and create a solution in ways that positively affect clients’ operations and bottom line.

Although their pioneering technology, management team and huge market potential led to the Forbes honor, Dey and Tryon say that the personalized approach is key to the company’s success. It’s not enough to have a great product—VEXTEC must meet customers’ needs to grow.

“Fundamentally, to be successful, your product has to make someone’s day-to-day life easier,” Tryon says.

Cummings

Peter Cummings may know the roads between the Vanderbilt campus and Oak Ridge National Laboratory better than he knows his own neighborhood. Cummings divides his work and time between the two institutions 170 miles apart, focusing on fundamental research in two areas with enormous potential: energy and cancer.

At the same time, as the principal scientist at the Center for Nanophase Materials Sciences at Oak Ridge, Cummings leads planning of the research agenda for a team of 95 researchers and support staff. The center is a U.S. Department of Energy/Office of Science Nanoscale Science Research Center.

And there’s more. In April 2009 the White House announced the establishment of a new multimillion-dollar Energy Frontier Research Center at Oak Ridge. The Fluid Interface Reactions, Structures and Transport (FIRST) Center is one of two planned for the facility. Cummings, the John R. Hall Professor of Chemical Engineering at Vanderbilt, now serves as a member of the FIRST leadership team and as a co-principal investigator. “The center represents an important investment in the basic research that will underpin new energy sources, energy storage methods and energy production techniques,” Cummings says.

The technology projects share a core intent. “The work we’re doing to understand what happens at the interfaces of different materials is crucial to a huge range of energy problems,” Cummings says. “Our work is not very applied—we’re working at a level that focuses on understanding very fundamental things that can lead to completely new energy technologies.”

Part of Cummings’ research focuses on developing theories that will become design tools for molecular electronic devices, which have the potential to replace silicon in future computer chips. Current chip manufacturing methods, based on lithographic etching of silicon, have enabled computer speeds to double every 18 months by carving ever smaller computing elements into the silicon. These methods will reach their limit in the next decade or two, when the etched structures will be too small to be stable.

“With molecular electronics, instead of trying to etch features into silicon, you make devices using a bottom-up technique called self assembly, with computing elements consisting of single molecules,” Cummings says. “Our goal is to understand how bottom-up self assembly works and can be controlled.”

Understanding from the bottom up is also at the core of Cumming’s cancer research. With colleagues in the Vanderbilt University Medical Center’s cancer biology department, Cummings upends the traditional model of understanding cancer at the tumor level and instead predicts tumor behavior by understanding and modeling how individual cells within the tumor move and interact with each other and their environment.

Using computer modeling, the researchers focus on the point at which cancer begins to move around the body. “We’re trying to understand how the properties of cells and nature of the environment impact whether a tumor will become invasive or not,” he says.

“Philosophically at the root of everything I do is the idea of understanding large, complex entities by understanding how the component entities move and interact with each other and their environment.”

Schmidt

With an enemy missile hurtling toward their aircraft, fighter pilots shouldn’t have to wonder whether their defense systems will work in time. Testing how such systems perform before they’re used in a hostile environment is just one of the many projects that Professor Doug Schmidt directs using building-block middleware computer software he, his students and staff developed.

“Our work relates to making it easier to develop and test these large-scale information systems to make sure they perform the right functionality at the right time,” Schmidt says. “These are very large, complex software applications and we use middleware platforms—reusable software that coordinates the application and infrastructure components of an IT system—and tools to validate and enhance confidence in mission-critical systems.”

Schmidt, professor of computer science and computer engineering and associate chair of computer science, oversees a team that currently performs such computer testing for a wide range of sponsors, including the U.S. Navy, Lockheed Martin, Raytheon, BBN Technologies, the U.S. Air Force, the Australian navy and Northrop Grumman.

Through an Air Force Research Laboratory grant, for example, Schmidt works with several companies to help develop a system to link defense fighters seamlessly to the Defense Department’s Global Information Grid (GIG). The goal of the GIG is to enable military personnel in the field to securely and reliably connect to needed information whether via Internet, cell phone, e-mail, GPS or technology yet to come.

Applications for his team’s work are not just defense-related: The European Space Agency is using Schmidt’s middleware technology as a building block for the Galileo global satellite navigation system, the continent’s own global positioning service. Europe’s new CoFlight air traffic management system, which will modernize and unify European air traffic, also uses his middleware technology.

Health care applications exist as well, particularly for medical imaging and picture archiving. Siemens and GE have been shipping medical imaging products based on Schmidt’s middleware for over a decade.

“Our ACE and TAO middleware are distributed using an open-source licensing model, similar to the widely used Linux operating system,” Schmidt says. Users can also customize ACE and TAO or consult with Schmidt’s R&D team for help in converting the baseline open-source middleware into specific applications.

“Smart” is the operative word for another area of research interest. As part of a new course, Schmidt is working with students to build applications that connect smart phones (such as Google Android and the Apple iPhone) to smart communication systems (such as cloud computing, where computing services are provided via the Internet). The project uses hardware and software provided by Google and Apple.

“We’re always asking ourselves how to keep students’ interest in computing. This is it,” Schmidt says. “We give them something technical they can hold in the palm of their hands. We teach them the fundamentals and then we allow them to build on these by developing interesting applications. We want to unleash the creative power of smart students.”

Miga

Image-guided surgery enables skilled physicians to perform difficult operations. But the images used for guidance are generally taken before surgery begins. How do surgeons account for changes that take place in tissue while the surgery is ongoing—changes brought on by the pressure of an instrument, a shift due to an incision or other factors?

That is the primary work of Michael Miga, director of Vanderbilt’s Biomedical Modeling Laboratory and associate professor of biomedical engineering. Miga and his colleagues produce computational modeling techniques that mimic these effects and are then used to compensate for tissue changes during surgery.

Miga’s computer modeling techniques so far have chiefly focused on brain, liver and kidney surgery. His lab is developing a novel computational framework that interacts with current operating room systems. It uses software and computer and measurement equipment to make calculations that modify the images for deformations during surgery.

This is important in brain operations, for example, as certain drugs shrink the brain during surgery, or when tissue is retracted during removal of a tumor. The computer modeling would cost-effectively augment the image-guided surgery techniques already in place and account for the tumor’s new location.

Miga says that unlike the brain, which is largely held in position by the skull, other organs are more flexible and can move during surgery.

Each organ has unique characteristics and different surgical approaches, requiring that the engineers apply separate research methods, new approaches to computation and different algorithmic design. Research done by Miga using laser-range scanning to capture the liver shape for image-guided liver surgery has already been incorporated into a new product that has received FDA approval. It is available in the marketplace and is being tested clinically (see Engineering Vanderbilt, spring 2009). Miga is currently focused on using the changes measured by the laser-range scanner to incorporate deformation correction into the product. It is expected to be the first of its kind available commercially.

Computer modeling also affects medical imaging. With breast cancer detection, current imaging modalities cannot document a tissue’s stiffness, an important biomarker of disease. Miga and his team are researching ways to use new noninvasive imaging methods to detect changes in tissue stiffness.

Using similar methods, but in a very different context, Miga also focuses on bone fractures. In this work, models and algorithms attempt to determine how well a fracture is healing by looking at the stiffness of the tissue at the fracture site.

“The common thread is that these are all mathematical model-based analysis approaches with a characterization, or interventional, aspect,” Miga says. “Strewn throughout each research project is the integration of computer models, soft tissue mechanics and analysis, with a central focus at translating the information to direct therapy or characterize tissue changes in an active way.”

Sztipanovits

As director of the Institute for Software Integrated Systems (ISIS), Janos Sztipanovits oversees more than $10 million in systems and information science and engineering projects involving more than 100 researchers, staff and graduate students.

These projects engage ISIS, and Sztipanovits, the E. Bronson Ingram Distinguished Professor of Engineering, in information systems, health care, and defense and national security.

Currently the development of projects on high-confidence system design with defense applications are ongoing, particularly for avionics. High confidence systems (those that developers and users have a high degree of assurance that they will not fail or misbehave) need to be secure and durable. “We are also interested in investigating the resilience of large information systems against cyberattacks,” says Sztipanovits, who chaired a study on the operational readiness of the U.S. Air Force against cyberattacks last year.

The largest effort in the defense area at ISIS, Sztipanovits says, is the development of a battle command system for the U.S. military’s Future Combat System program. This initiative uses ISIS-pioneered, model-integrated computing design principles that combine various technologies into a consistent and reliable system.

“This whole area of model-integrated design of systems and software is really rapidly moving into the mainstream and that creates a quite new approach to engineer, integrate and operate large networked systems,” he says. “It is also the foundation for new system design methods and tools that go beyond the conventional programming languages.”

Other projects range from the development of online training materials for homeland security purposes to creating a countersniper system. This cost-effective sniper location system detects when weapons are fired and the direction of the bullets. The countersniper technology would allow soldiers or police officers to react rapidly and minimize injuries.

ISIS’s expertise in understanding and incorporating complex privacy, security and systems integration issues dovetails into its work in health care. Currently ISIS researchers work with Vanderbilt University Medical Center toward creating trustworthy health information systems. Vanderbilt is joined in this National Science Foundation partnership by scientists from Berkeley, Stanford, Cornell and Carnegie Mellon universities.

A patient management system for managing sepsis is another program developed with medical center colleagues, Sztipanovits says. In May initial capabilities for the system were deployed in the medical center for testing, where it has already demonstrated viability. “We also are discussing how to extend that system toward other areas, such as home-based management of congestive heart failure patients,” he says.