"Surgical Analysis" Feature Article, August 2004

surgical analysis

Engineering technology gets more work in the operating room.

byJean Thilmany, Associate Editor

The puzzles that engineers and doctors face have much in common. The human body, after all, can be thought of as an engineered system. Blood flowing through the body, for instance, mirrors airflow through an engine.

It's natural then, that medical researchers should appropriate for their own purposes engineering analysis technology originally developed to predict the flow of fluids through engines or of water around bridge stanchions.

Computer images form the basis of models that can guide surgeons through the unique fluid dynamics of each patient.

As these technologies find a home in hospitals and doctors' offices, they'll change the way surgeons assess patients' conditions and operate. Analysis software lets surgeons plan an operation, by mapping it specifically to the patient's body. Doctors will predict how a patient will respond to a medical implant and plot implantation accordingly, say researchers who work to configure engineering technologies for medical use.

A Cultural Coming Together

David Gosman—professor of computational fluid dynamics at Imperial College of Science, Technology, and Medicine in London—uses computational fluid dynamics to study blood flow. He says engineering simulations are finding growing acceptance in the medical realm.

"But my impression is that it's still the early days," he said. "CFD is relatively new, so it'll take a while to become known and appreciated.

"We need more research to fully understand what simulations can offer and how to use that information," he added. "And the tool needs to be put into an easy-to-use form so that it can be routinely applied to cardiovascular studies without the need for CFD specialists."

Of course, before technology used and developed by engineers can shift to the medical world, the separate cultures need to find a common point of departure.

The medical researchers, often engineers themselves, have to understand how doctors work and how they'd accept using a computer as a tool in the operating room. But in the subtle world of medical problem solving, doctors also have to accept that the cold, hard numbers that drive the computational methods behind analysis software do provide answers.

"A fascinating, but frustrating, meeting of the minds between engineers and doctors results when the two fields must by necessity work together to develop medical equipment or methods," Gosman said.

Professionals in the two fields use different terminology and approach measurement and simulation from different angles.

"We found the different jargons initially to be a real barrier, but this was gradually overcome with give and take on both sides," Gosman said.

Still, for many doctors, computers and surgery just don't mix, said Anthony Petrella, manager of the computational biomechanics group at DePuy Orthopaedics Inc. of Warsaw, Ind. The company makes almost every type of replacement joint.

Medical researchers at DePuy are working to repurpose CAE software with an eye toward visualizing and analyzing the human body. Petrella said the company wants to use analysis software to improve joint replacement surgery. Doctors would use the software to predict how a patient's new joint will likely wear and how tissue will respond to the artificial joint. Guided by those predictions, they can position the joint to the best effect and use an artificial joint of appropriate material.

Joint replacement has become almost routine. Many U.S. surgeons perform joint replacement operations. For some, that's almost all they do. Therefore they have little time to incorporate a new system into their daily routines.

"Surgeons are skeptical of computational methods," Petrella said. "Still, new doctors today are tech savvy. They played computer games as kids."

Computer-assisted surgeries, in which doctors get real-time information about their patients as they operate, are becoming more common, but computers still aren't usual operating room tools. Many surgeons don't feel comfortable using them, Petrella said. He spoke in May at a user conference sponsored by Ansys in its home town of Pittsburgh. DePuy's research uses Ansys software for analysis.

A medical tool of today could be used in the operating room to capture information about the patient's condition and feed it to a computer. According to Petrella, fluoroscopy— essentially a moving X-ray of a patient's joint—can visualize how the joint moves dynamically. That image would be the basis for the computer model to predict how the joint would wear over time and how surrounding tissue might react to the synthetic material.

Researchers can calculate when and how a patient's artificial joint will wear. But creating a model to predict how tissue will respond to the new joint is a different ball game entirely, Petrella said. Healthy tissue can respond any number of ways to the implant. It might become stiff or change shape over time. Analysis codes are hard to write for that type of intangible.

Still, Petrella and his fellow researchers say a computer program is within reach that will feed surgeons information unique to each patient.

Help For a Weak Heart

David Gosman and his team at Imperial College are adapting technology originally developed to simulate airflow through a reciprocating internal combustion engine so that it can be applied to study the human heart.

CFD is most commonly used now to plan heart-bypass surgery. Surgeons simulate blood flow through the heart via CFD then program their findings into CAD software, where they design a patient-specific arterial bypass. Gosman's research, on the other hand, aims to simulate how a weak heart muscle affects blood flow. The simulations would help researchers improve pacemakers and artificial heart valves.

"Doctors have been interested for some time in knowing about blood-flow behavior in the heart and how it differs between healthy patients and those suffering from heart illnesses," Gosman said.

Little is known about the effects of heart irregularities on the flow, Gosman said.

"And there's even less understanding of how pacemakers can be designed to control the heart motion in a way to restore the healthy pattern," he added. "What kind of motion is necessary to achieve that?"

CFD simulations of interior organs may be used to study the effects of circulatory flow on the walls of the aortic arch.

Though CFD has already been used extensively to study artificial valves, little research has been done to evaluate how they affect blood flow when actually placed in a patient, Gosman said. His simulations will look at that.

Measurement techniques like magnetic resonance imaging give doctors information about patient blood flow and offer them a glimpse of how it might be affected by a pacemaker or artificial heart valve, but MRIs don't return the level of detail that CFD simulations depict, Gosman said. His approach combines MRIs with CFD simulation.

The information obtained by an MRI is in the form of a thin, two-dimensional image slice. Gosman and his team take MRIs at varied parts of a ventricle as it fills with and empties of blood. The images combine to show how the chamber looks during the entire cycle.

That image is then modeled with CFD. For modeling, the team uses Star-CD software from CD-adapco of New York. Gosman is also director and vice president of technology of CD-adapco.

With the model, doctors can predict how blood flow would change if a pacemaker or artificial valve were placed in the heart. It also lets doctors look at what-if scenarios that they can't run on the real subjects.

At first, the simulations would be used to help diagnose patients, and Gosman hopes that doctors eventually will study a particular patient's CFD simulations to plan surgery.

Help Flow Blood Freely

Researchers at the Georgia Institute of Technology in Atlanta are also studying blood flow. They have an eye toward using the information to design better artificial heart valves.

Numerical modeling techniques originally developed by the institute to simulate how water flows around hydraulic structures like bridge foundations is finding a second use in helping researchers better understand blood flow patterns through artificial mechanical heart valves.

The research could yield the most accurate depiction yet of the turbulent environment that blood cells and platelets encounter as they pass through the mechanical heart valve, said Fotis Sotiropoulos, an associate professor in the schools of civil and environmental engineering and mechanical engineering. He's researching the system along with Ajit Yoganathan, who directs Georgia Tech's cardiovascular fluid mechanics laboratory.

Doctors replace poorly functioning natural heart valves with prosthetic valves. But present-day designs are far from ideal. They can destroy blood platelets or permit a particle that has broken away from a blood clot to block a blood vessel. These complications likely come about because the blood is exposed to excessive stresses from the turbulent flow in the vicinity of the mechanical prosthesis, Sotiropoulos said.

Doctors also think the complex blood flow patterns around the valve may trigger a cellular response that can lead to the onset of heart diseases.

But before the artificial valves can be redesigned, researchers need an in-depth understanding of the flow fields that the valves induce.

To date, the fluid mechanics of heart valves has been largely studied by way of experiments, Sotiropoulos said. His research is an attempt to apply both CFD and experimentation to the problem.

Cardiovascular flows generally pose unique challenges to even the most advanced CFD tools available today. The interaction between blood motion and compliant vascular walls leads to a very complex fluid-structure problem, he said.

Yet Sotiropoulos believes that CFD is developed well enough today to tackle the problem. "Computational resources and especially the advent of massively parallel clusters in the past decade have made it feasible to attempt such complex computations in very demanding simulations," he said. (Massively parallel clusters assemble many conventional CPUs into a network to perform large-scale computer simulations or solve a large problem.)

CFD simulations of interior organs also may be used to observe in detail a ventricle in the process of pumping blood.

Still, his team relies on the results of physical experiments to prove the accuracy of their CFD simulations. An in vitro experiment can use an anatomically realistic model of the human heart, through which researchers can pump blood. They can compare results against those of a CFD model. In vivo information is also available in the MRIs of patients.

"Experiments have to accompany CFD modeling to provide it with the credibility medical practitioners need to be convinced they can rely on such modeling tools to make decisions that could ultimately affect the span and quality of the life of the patient," Sotiropoulos said.

The Georgia Tech team ultimately wants their tools to be used in what they call virtual surgery. Doctors would model a patient's heart and blood flow to plan the best way to operate—before any incision is made.

Manufacturers of heart valves can also use the CFD method to optimize their designs to minimize hazards to blood elements.

Along with the exciting opportunities, however, comes the challenge of ensuring that CFD doesn't become colorful science fiction, but remains grounded to the physical reality it's intended to simulate, Sotiropoulos added. Advances in other medical technologies make that likely.

A major difficulty in applying CFD to cardiovascular flows stemmed from medical researchers' lack of understanding about the exact geometry of the various blood vessels that serve as the basis for a CFD model. Nowadays, major advances in MRI techniques have improved the accuracy of models of blood vessels.

"The availability of anatomically realistic geometries has opened new horizons for CFD and paved the way for CFD to make a major mark in the field of biomedical engineering," Sotiropoulos said.

"The term virtual surgery no longer sounds like catchy phrases from a science-fiction novel; it's going to become reality in the coming decades," he said.