Radiologists at Case Western Reserve University expected change when they acquired a multislice CT scanner two years ago. They anticipated that it would save time, but they also knew they’d face archiving issues for the huge quantity of images generated.

What they didn’t foresee was that physicians outside the radiology department would become enamoured of the machine, according to Dr. Robert C. Gilkeson. He was surprised to find his referring physicians demanding MSCT.

“Surgeons and referring physicians aren’t really cued into imaging,” said Gilkeson, physician-in-charge of chest radiology. “This is the first time they’ve really gotten into the technology. They get in a huff if it isn’t done on this machine.”

As imaging centers take the plunge into multislice CT imaging, they find that the new devices save time, as promised, but also offer challenges along with opportunities for changes and refinements to practice. Diagnostic Imaging asked several early adopters of MSCT to discuss their experiences with the technology: what they’re doing with it, and where it might take radiology in the future.

Case Western’s MSCT is the machine of choice for almost every cardiovascular case, according to Gilkeson, whether it’s virtual endoscopy of the aorta prior to surgical reconstruction or an evaluation of pulmonary embolism. The images are better and more consistent than those from single-slice CT.

“We use it for almost all of our cardiac evaluations, whether it’s coronary artery disease or aortic disease,” he said. “When our surgeons do complex aortic reconstruction, they demand that the scans be done on this machine.”

Case Western has six CT scanners. Only one is multislice, but Gilkeson estimates that 20% to 30% of all scans are done on that machine. He predicts that share will rise to half fairly quickly, due in large part to pressures from referring physicians. The technology is useful for mapping congenital heart defects in the pediatric population, for example.

“Traditionally, MR has been used, and one of the difficulties is the issue of sedation in the pediatric population,” Gilkeson said. “An MRI can take half an hour. With MSCT, the scan takes eight seconds. You can often not even sedate the patients.”

The flood of images MSCT produces poses reading and archiving issues that he and his colleagues are only beginning to address.

“We read everything off monitors,” he said. “You have to-there’s no way you could read all that film. It puts a certain stress on departments where there are older radiologists who just want to look at film.

Older single-slice machines still have a place in practice, according to Gilkeson. Many indications can be imaged adequately without multislice. Radiologists just need to agree on what those indications are. In the meantime, the growing popularity of Case Western’s MSCT with nonradiologists is taking on a life of its own.

“The clinical pressures have been such that we’ve had to move the machine upstairs to the ER, where physicians are demanding its use in trauma cases,” he said.

Coping With Volume

Multislice CT provides new ways to look at the volume of data, accelerating the routine use of three-dimensional imaging in diagnosing, said Dr. Alec Megibow, vice chair of radiology at New York University Medical Center.

“The output of multislice is so huge, there’s no way to look at it the way we used to look at CT,” he said. “Very soon we will no longer be sending referring clinicians stacks and stacks of film.”

NYU collects 1-mm slices for most applications, but instead of looking at slices, radiologists look at the entire volume of data at once. They sandwich slices together to look at the sum and can then define what slice thickness is desired.

“It’s the volume we look at,” Megibow said. “Maybe the whole clinical question is answered on one slice. On multislice, our approach would be to combine all of the slices and create one or two key images. We can do that in any phase and then can begin to take advantage of 3-D imaging.”

The radiologist is able to interact with the surgeon in new ways, according to Megibow. It becomes possible to look at laparoscopic planning for minimally invasive surgery, for example. He considers these interactions to be the major value of MSCT.

NYU has had MSCT for almost two years, and it has increased the volume of cases to be read, which in turn increases demands on personnel. MSCT vendors are working with radiologists to improve the user interface on machines to present the data in a more user-friendly manner.

MSCT allows new applications such as calcium scoring and angiography and improves on others such as virtual colonoscopy, Megibow said. He predicts an explosion in use as more locations add a multislice machine, replacing applications such as cervical spine film and coronary CT in the emergency room. The possibilities are enough to make single-slice machines obsolete, even though patients in some situations are still well served by the older technology.

“We lived fine on single-slice, but I don’t think anyone would buy or replace a single-slice machine now,” Megibow said. “What MSCT is going to do is establish CT as the general radiologic procedure.”

Sold on Speed

There are two steps a radiology group must take when acquiring a new multislice CT scanner, according to Dr. Michael Vannier: Throw out all the old protocols, and develop new protocols for everything.

“There’s little justification for continuing old practices,” said Vannier, a professor of radiology at the University of Iowa. “Some groups realize this sooner than others, but eventually everyone changes their protocols, or they don’t realize the benefits of MSCT.”

In Vannier’s practice, single-detector CT now accounts for only 20% to 25% of the routine load. It suffices for routine follow-up or repeat scans on an inpatient to check for catastrophic changes-an intracranial hemorrhage or new hydrocephalus. It’s used for many routine functions when a positive result is not anticipated. If one occurs, additional imaging procedures will be done. Those using single-slice in the future will most likely do so because the time has not yet come to replace the scanner, and the compromise in image quality and speed is acceptable.

“I can’t imagine anyone expressing a preference for single-slice CT,” Vannier said.

Most of the time spent in MSCT involves getting the patients set up as they are placed on the table, administering contrast, and taking patients to the waiting room. The actual scan time is negligible in most cases, he said.

“The higher heat capacity of the x-ray tube and greater efficiency of the detector make it unnecessary to wait for tube cooling, and x-ray tubes seem to last forever since they are not as heavily loaded,” he said. “So the reliability of the system is better and up-time is higher.”

Faster scanning doesn’t translate into faster reading of images, however. Every slice still must be reviewed.

“Some radiologists who are reluctant to change insist on filming every third or fifth slice, but their resistance is limited and ultimately they must change,” he said. “It’s impressive how much easier it is to read scans once you’re familiar with the stack mode. I’m convinced that we miss important findings when we don’t use it, especially if they’re subtle.”

Without a doubt, MSCT has shifted the threshold for routine practice, Vannier said.

“We can now see smaller structures more easily,” he said. “Isotropic imaging with CT—once thought to be impossible except with MRI—is now nearly routine.”

Evolution, Not Revolution

Multislice CT is an evolution, not a revolution, according to Dr. Brooke Jeffrey, chief of abdominal imaging at Stanford University. MSCT doesn’t open up new worlds for the radiologist-it improves on existing capabilities.

“There is no advantage at all to having anything other than multidetector CT,” he said. “This makes everything we do better. It has increased speed, it allows thinner collimation.”

Where a routine scan might have been done at 3 mm using a single-detector machine, it is now done at 1.25 mm. And there are a few procedures-such as CT angiography-that can be done only using MSCT.

“You can capture the entire chest,” he said. “This could replace catheter angiography. Whether CT angiography or catheter angiography will win out in the chest, we don’t know.”

Jeffrey still looks at every image collected, but he “track-balls” through them on the monitor. User interface technology has lagged behind the advances on the imaging front, he said. MSCT could necessitate CAD technology.

“If you’re looking at virtual colonoscopy, for example, you might generate 1000 images looking for a 1-cm polyp,” Jeffrey said. “You could get carpal tunnel syndrome looking through all those images. We need software to tell us, ‘Instead of looking at these 1000 images, look at these 10 images.’“

Stanford has invested in a PACS to archive and store images, which Jeffrey said is just about essential with MSCT. Almost all centers with MSCT also have PACS, which is considered a necessity for dealing with stacks of filmless data. When its MSCT arrived, Stanford had to change all its protocols, scanning with thinner collimation and delaying acquisition longer. Adjustments in contrast were also necessary. But changing the protocols was not difficult, even though the MSCT machine posed a bit of a learning curve. Stanford is installing its third multidetector CT and is phasing out its last single-slice machine.

“As far as we’re concerned, single-slice is obsolete,” Jeffrey said.