Multidisciplinary CTA research could improve CAD screening
 | |
| Streamline visualization of a patient's left coronary artery with color indicating velocity magnitude (red is high and blue is low). Image courtesy of Harvard School of Engineering and Applied Sciences. |
X-ray angiography serves as the gold standard for diagnosing CAD, but the imaging method is both invasive and limited in its detection of stenoses. “Tim Russert is our poster child, we want to prevent more of those cases,” said Charles L. Feldman, ScD, a lecturer on medicine at Harvard University School of Medicine and director of the Vascular Profiling Lab at Brigham and Women’s Hospital in Boston.
Russert, the host of NBC’s Meet the Press, died in June of 2008 of an MI after a coronary plaque ruptured. His death occurred despite a negative stress test not long before his MI, which indicated that the lesion was not causing major flow limitations.
“A problem with coronary angiography is that some plaques can appear minor that aren’t necessarily so, and others can be larger but never amount to anything,” Feldman continued. “Sometimes you don’t know which ones to go after and which ones not to.”
Feldman is part of the Multiscale Hemodynamics Project at Harvard, which brings together a group of cardiologists, radiologists, physicists and computer scientists. “The ultimate goal of our project is to anticipate which lesions will become problematic in the near future and to intervene early on,” Feldman explained.
The researchers have been using CT angiography (CTA) to reconstruct the geometry of the coronary tree in an effort to more accurately assess the risks of patients’ lesions. They are trying to minimize some of the guesswork involved in assessing coronary plaques on conventional angiography by using CT to also evaluate endothelial shear stress.
Instead of relying on visual assessment of plaques, the group is making use of supercomputers, as well as clinical studies, to determine through simulation the frictional force exerted by blood flow on the vessel walls. Although a particular lesion might appear relatively benign, they’re finding what matters just as much is the stress exerted by the blood on the arterial walls.
“What makes our simulation unique is our use of the lattice Boltzmann method,” explains Michelle Borkin, part of the hemodynamics research group and a PhD candidate in applied physics at Harvard’s school of engineering and applied sciences. “Think of a whole bunch of particles bouncing around. We are able to simulate how these blood particles move through the complicated geometries of the coronary tree in a very elegant method.”
Borkin and colleagues use the geometries acquired from CTA to map patients’ coronary trees, and then run additional CTA measures through supercomputers to simulate the force of blood flow on the lesions that line vessel walls in patients with CAD. They are attempting to show that these measures of endothelial shear stress, which angiograms like Russert’s fail to depict, offer improved predictions of whether arteries are likely to rupture.
“The grand vision in the end is that this work will lead to a new diagnostic tool, one that’s noninvasive, very accurate and can show a doctor even more information about the plaque deposits than conventional angiography,” Borkin said.
Feldman added that Borkin’s interactive HemoVis provides cardiologists and radiologists with an important visual tool that allows them to accurately and quickly absorb the scores of data fed into the systems.
For now, x-ray angiography still provides better spatial resolution, Feldman said, though he indicated that this would likely change within five years. Also slowing research progress is computation speed—even with the use of supercomputers, simulating the flow of blood particles, instead of performing the calculations for a uniform liquid, uses up considerably more computer processing units. Nonetheless, the group is looking forward to the release of a 500-patient trial which Feldman predicted will have a “considerable impact,” but declined to discuss until publication.
Feldman and Borkin envision CTA as providing a noninvasive and in many ways more accurate alternative to conventional angiography, providing plaque visualizations as well as endothelial shear stress measures for the entire coronary tree.
To view the interactive visualization program click here.