Stanford symposium finds high level of interest in multislice CT

August 4, 1999

Most early adopters opt for resolution over speedIn computed tomography, thinner—not faster—is definitely better.At a symposium held in San Francisco from June 26 to 27, top-flight radiologists extolled the virtues of multislice CT

Most early adopters opt for resolution over speed

In computed tomography, thinner—not faster—is definitely better.

At a symposium held in San Francisco from June 26 to 27, top-flight radiologists extolled the virtues of multislice CT scanners, which make four slices for every one delivered by conventional scanners. These new scanners provide the option of completing scans very quickly—or making slices much thinner. Early adopters, describing their experience with multislice scanning, resoundingly chose thinner slices because of their increased ability to uncover disease and to assist in the differential diagnosis of difficult cases.

The clinical results support the expanded use of these new scanners, just as the level of interest demonstrated at the symposium indicate that sales of multislice scanners could boom. An informal poll of the more than 300 attendees at the Stanford-organized symposium indicated that only about a dozen currently have multislice scanners. At least a third of the audience, however, raised hands when asked if they would consider the purchase of such a scanner in the next year.

GE Medical Systems of Milwaukee is the only company currently delivering multislice scanners, having shipped its 100th system just prior to the symposium, according to a source at GE. But Toshiba America Medical Systems, as well as Siemens Medical Systems and Picker International, who are partnering on the development of components for two different multislice scanners, are expected to join GE in delivering their own proprietary scanners before the end of this year or early next. Philips Medical Systems could round out the group of top CT vendors with multislice scanners, perhaps later in 2000.

Comparing images from conventional and multislice scans documented the advantages of this new genre of systems in scans of the central nervous system, thorax, and abdomen. Thin slices can resolve densities in the lymph nodes indicating necrosis or demonstrate disease distribution in liver, said Dr. Gary Glazer, professor and chairman of the Stanford Department of Radiology.

“Going with thin sections allows better detection and characterization and provides optimal use of contrast (media),” Glazer said. “There’s no way seven millimeter lesions would be seen without thin sections.”

Speed beats out thin slices, however, when assessing trauma patients. The lightning speed of multislice scanners not only freezes motion but can provide a head-to-toe assessment of the patient. Radiologists at the University of Michigan Medical Center in Ann Arbor have proven the potential of such scans, programming their multislice CT with fields of view and algorithms optimized for different parts of the body.

“The result is a total spine, a total body, and a head evaluation,” said Ella Kazerooni, UM associate professor of radiology. “This can only result in better patient decisions and reduced morbidity and mortality.”

Other investigators noted that multislice CT might be applied to angiography. In several instances, the new CT rivaled the gold standard based on x-ray fluoroscopy. Still beyond the grasp, but possibly within the future reach of multislice CT in the years ahead, is cardiac scanning.

The ability to do coronary calcium scoring, already possible with conventional spiral scanners, was well documented for multislice scanners by the symposium presenters. Additionally, scans with the new technology showed more and better cardiac information. Stanford radiologists announced they could make determinations not possible on conventional CT. In one case, they noticed a defect in one of the coronary arteries, later confirmed on cardiac catheterization to be a 90% stenosis.

“Our patients are coming in with these problems and we need to look for them,” said Dr. Geoffrey Rubin, an assistant professor of radiology at Stanford. “We can diagnose them on chest CT more frequently than previously. And this is not even a dedicated cardiac package.”

Still unresolved is how best to view the data generated during multislice scans, particularly those that take advantage of the thinner slice capability. Whereas a conventional CT scanner might produce dozens of images, a typical multislice study kicks out hundreds. The severity of the problem depends on the application.

When used to evaluate some trauma victims, radiologists at the University of Michigan Medical Center produce about 1000 individual images. Kazerooni noted that the physicians do not have enough time to look at every image, leading some symposium attendants to worry about legal liability in the event that an injury or pathology is missed. Making matters worse, the opportunity to go back for a second look is fleeting.

“We have to dump the raw data because we don’t have enough archive space,” she said.

Multislice scanners are problematic even in routine work-ups, as they typically generate 500 or more images per study, regardless of the application. Many symposium presenters relied on 3-D reconstructions to display volumes of data quickly, extracting slices to demonstrate pathology in familiar ways. Others showed how virtual fly-throughs of the airways, colon, and even blood vessels might be used in diagnosis. The inefficiencies of point and click viewing, however, remained an obstacle to assessing the data.

The key to the future, according to Rubin, is a better human-machine interface. Engineers have to give up the computer mouse for a more intuitive connection, he said. One possibility is the prototype Rubin and colleagues have built at Stanford. The “pixel wall”—a 6 x 2-foot electronic light box—might display digital images in response to a wave of the hand, a spoken word, or a flash of a laser light pen. Imaging protocols and equipment need to be integrated, if the potential of multislice CT is to be achieved.

“We have to be able to navigate and interpret efficiently,” Rubin said. “To do this we must have good control of the visualization.”