Applied Radiology

The promise of 16-slice computed tomography (CT) scanners became a clinical reality in recent weeks as four major manufacturers rolled out the latest additions to their CT lines. Siemens Medical Solutions (Iselin, NJ) began shipping the Somatom Sensation 16, Philips Medical Systems (Bothell, WA) began commercial shipment of their Mx8000, and GE Medical Systems (Waukesha, WA) began full production of the LightSpeed ¹⁶ . In addition, Toshiba America Medical Systems (Tustin, CA) received marketing clearance for their version of the 16-slice CT, the Aquilion 16, in mid-July and is expected to begin commercial shipment momentarily.

Detector technology

All of the new 16-slice scanners offer sub-millimeter slice thickness and half-second or less rotation time. The Philips Mx8000 has a 24-mm wide detector and can operate in two modes: 16 * 0.75 mm or 15 * 1.5 mm. With the addition of Infinite Detector Technology (IDT) and a 0.42-second rotation time, it can collect 16 slices of data simultaneously, and 38 slices of data per second, thereby covering more than 4 cm of patient anatomy in 1 second with sub-millimeter isotropic accuracy. The system also transfers data from the detector array at a rate of 1 Gb per second.

The Siemens Somatom Sensation 16, with a 0.4-second rotation time, has a voxel size of 0.5 * 0.5 * 0.6 mm. The system can acquire 72 mm of coverage per second with 1.5-mm slice thickness.

The GE LightSpeed ¹⁶ provides 16-slice scanning in all modes with sub-millimeter voxels and a 0.5-second rotation time.

The Aquilion 16 from Toshiba utilizes a 40-row quantum detector capable of generating 16 simultaneous 0.5-mm, 1-mm, or 2-mm slices with each 0.5-second gantry rotation. According to the company, the system can complete a CT angiogram that scans from the abdomen to the feet in 2-mm slice thickness in 15 seconds.

Technology evolution

"If you look at CT over the past 4 years, I think the most amazing thing has been the vast evolution in the applications and the hardware changes that enabled that," said Peter Arduini, general manager, global CT at GE. "Just from 1998 until today we've evolved from 4 to 8 to 16 slices."

The challenge to get there was threefold, noted Markus B. Lusser, segment manager, radiology and computed tomography at Siemens. "The main challenge in designing such a scanner was not only to design a machine that is able to acquire more slices," he said, "but to acquire thinner slices and higher resolution and, in the end, to handle the massive data sets very quickly."

"A number of things evolved over time," said Bill Kulp, manager, product marketing at Philips. "First, the computers had to be fast enough to process all the data." Next, he credits two major technological advances with helping to make the 16-slice CT possible. One was the development of cone beam reconstruction algorithms. "When you get the wider detector and the thinner slices, if you don't have cone beam reconstruction, you start getting artifacts that become clinically significant when you get to 16 slices. These algorithms have been known for many years," he noted. "The challenge in implementing them was the fact that it takes approximately an order of magnitude, or 10 times, more data to process and reconstruct an image in the cone beam. Also, there is a whole new set of artifacts and other things that have to be fine tuned out to make the images look good. The other major challenge," he continued, "was just getting the massive amount of data off of the detector. When you are going from 4 to 16 slices, there is obviously four times more data, and if you are going to a faster rotation speed, there is even more. This scanner is going to four images per second and we are going to be going faster in the near future," he noted.

"I think the cone beam reconstruction is going to be a key to image quality," agreed Bryan Westerman, manager, clinical sciences at Toshiba, "because with this advanced technology, we have to be very careful about image quality."

Clinical evolution

As the technology has improved, the clinical role for CT has evolved as well. "When you go above 4-slice, all your applications and protocol work change pretty radically," said Arduini. With the introduction of multislice CT, the modality moved from one that was used primarily for core radiology--central nervous system and basic body imaging--into a variety of clinical practices, including imaging the heart, lungs, and colon.

"The 4-slice CT scanner opened the door to new applications," noted Sholom Ackelsberg, general manager, global CT advance products at GE. "It showed you could do CT angiography. The 8-slice made 1.25-mm imaging routine for short coverage and 2.5-mm routine for everything else. It expanded the ability to do CT angiography on a routine basis and it showed the way for being able to look at CT colonography and for some of the more advanced applications, such as accurately sizing nodules in the lung and polyps in the colon. So, the 4-slice showed what was possible and the 8-slice made it real. The 16-slice just makes thin slices, 1 mm or smaller, routine everywhere in the body."

Clinical applications for 16-slice CT

"Just about every area of imaging is going to be made better by 16-slice scanning," said Kulp. "The reason is that you can use thinner slices over bigger volumes. Thinner slices are the Holy Grail of imaging because it gives you more detail, more image quality, and better resolution."

Most experts agree, however, that cardiac imaging is the field expected to benefit the most from the new technology. "Without question, the biggest advantage I see would be in cardiac imaging," said Westerman.

"You are looking for a 50% stenosis on a 2-mm vessel," said Ackelsberg. "You need to be able see <1 mm in your spatial resolution and you need to be able to see it in three dimensions because the vessels don't necessarily align themselves along one axis of the scan.'

"We, and I think others, have shown pretty dramatically that when you are dealing with coronary arteries that may be 1 or 2 mm in diameter, you really need sub-millimeter slices to get good accurate information," agreed Westerman. "If you are using the sub-millimeter slices, you are back in the old situation of coverage versus time," he explained. "With the 4-slice scanners, we cannot use 0.5-mm slices and cover the whole heart in a reasonable single breath hold. Therefore, most people are using the 1-mm slices. When we have the 16-slice available, that gives us the option of using the sub-millimeter slices. The 0.5-mm slices will give you the best spatial resolution and you'll be able to cover the whole heart in a very reasonable breath hold of 18 to 20 seconds."

"I think the biggest effect has been that it really takes the vascular studies to a whole new level of reliability and resolution," noted John R. Haaga, MD, chairman of radiology at the University Hospitals of Cleveland. Dr. Haaga participated in the beta testing for the Philips Mx8000 and has been using the system clinically for approximately 8 months. "I think we've all been pleased with the ability to start looking at the blood vessels with the multislice scanners. That the slices are thinner, more reliable, and faster just gives us a purer quality. It really does almost rival what can be obtained from a direct study. People have talked about using CT in this fashion for many years, but I really think that until we have these new devices it really hasn't been as consistent and reliable as we would have hoped, but these really are proving to be true."

"Cardiac imaging is the obvious one," said Kulp, "but beyond that, areas like trauma are very important as well. In a trauma situation, you want to be able to get as much diagnostic information as fast as possible. With this scanner, you come right in, you do the 16-slice CT, you are able to have the coverage with thin-slice imaging capabilities, and you can get all the diagnostic information you need right off the bat. And time is life in these situations."

Changing protocols

With scanning times reduced to <1 minute with the 16-slice scanners, most agree that clinical protocols will need to be adjusted. 'As the customer goes from 4-slice to a 16-slice scanner, they have to change the protocols they use,' said Kulp. 'If they don't do that, they are not taking advantage of the performance of the new scanner; all they are doing is having a very expensive new tool that they are using like the old tool.'

In particular, contrast protocols may require modification. "First of all, the acquisition is over after an average of 10 or 11 seconds in high-resolution mode or in sub-millimeter mode probably in 5 or 6 seconds through a 300- or 400-mm volume," said Lusser, "so contrast media protocols have to be adapted to utilize the most power of the machine."

"Shorter scan times require a little more care in how you try to track the contrast material," said Westerman. "Certainly, I think there is going to be another bit of impetuousness in increasing injection rates. People are starting to talk about injection rates of 5 or 6 mL per second. This is a little higher than most places have tended to work with. It's going to be interesting to see how this works out. It also may be possible to reduce contrast volume somewhat."

Haaga said that at his institution they are using "somewhat less" contrast media with the 16-slice scanner depending on the study. "Of course, timing is a little more critical, so it is important that we use special programs to track the bolus before we begin the scan," he added.

Ackelsberg noted that it has been proposed that a 50% reduction in contrast media may be possible with the use of a 16-slice scanner and tighter boluses with accurate injection timing. "But," he cautioned, "this still has to be proven clinically."

Looking ahead

While Toshiba currently has a 256-slice scanner under development, most manufacturers are focusing more on clinical applications than on increasing the number of slices. "We don't see a big trend toward higher number of slices," said Lusser. "Of course, there will be a trend toward thinner slices."

"I think that until now the world has been very much focused on the hardware revolution," said Arduini. "I think we are at a point now where we have a lot of speed throughout the body and can perform a lot of new applications. What we are focusing on at GE is evolving the overall application package; everything from colonoscopy packages, lung capability, and tools, such as computer-aided detection applications, to enhance not only the overall productivity of the clinician but to really help, over time, to make better decisions about disease management, particularly in oncology. Down the road, we are going to evaluate at which level of slice and coverage real value can be brought above 16. Our hope is to understand where perfusion studies of the heart and full-organ imaging in a rotation can be gained and what new applications they can bring as well."

"For conventional CT applications, 16 is pretty darn good and it gets you very close to where you want to be theoretically," agreed Kulp. "But as you go faster, as you get wider detectors and thinner slices, new possibilities open up. I can see CT being used for things that it was never used for before and that is really where you want to go; for example, whole organ perfusion. For that, you are going to need a detector that, instead of being 2 or 2.2 cm, is 6 or 8 or even 10 cm. And you're going to need thin slices over that kind of a volume."

"CT is coming more and more to the front of the hospital," concluded Arduini, "and I think the 16-slice scanner is going to be more and more of a force in that evolution." AR