New 64-slice devices drive radiology to new heights

Sixty-four slice CT makes all things radiologic better. The high spatial and temporal resolutions that work so well for cardiac imaging also improve other applications, especially vascular studies.

Whenever small vessels are involved, and whenever early detection and quantification of disease are an issue, 64-slice CT makes a difference, according to early adopters of the technology.

Motion artifact is no longer a problem for pediatric, trauma, and geriatric patients, thanks to five- and 10-second scans. Increased coverage with 64-row detectors has made stroke studies and even oncologic evaluations feasible.

There's more to these scanners' popularity, however. Radiologists don't actually have much choice about buying them. They have been replacing old CTs with new ones every few years for nearly three decades. Being on top of their profession requires that they have the latest tool, which they use to improve the quality of routine studies and boost productivity.

More patients, higher revenues, and greater patient access to healthcare constitute the trinity of diagnostic imaging. And radiologists now have an added reason to buy the latest generation of CTs: For the first time in the modality's history, radiologists specializing in CT are feeling outside pressure. Cardiologists see 64-slice scanners as the noninvasive successor to diagnostic cardiac catheterization for screening selected patients and following up those who have undergone interventions.

Hundreds of CTs from the major vendors are flowing into cardiology. Radiologists cannot hope to stem this tide, but they can maintain control of the modality if they use the latest technology to better do what they have always done best.

"We can perform much more detailed analysis of very complex anatomy with a speed and confidence that we couldn't aspire to a year ago. It has changed what we can look for and how we look at it when we find it," said Dr. Michael Vannier, a professor of radiology at the University of Chicago.

Patients benefit from the speed of 64-slice CT, which is due to the greater coverage 64-row detectors provide. The heart, brain, or both lungs can be scanned in five seconds; the whole body takes 30 seconds.

CT had fallen short of MR in stroke evaluation, largely because MR imaging could cover more area, but the advent of 64-row detectors changed that. A 16-slice scanner covers 8 to 12 mm in a single pass, a 40-slice scanner covers 20 to 32 mm, and a 64-slice device covers up to 40 mm. Greater z-coverage provides immediate benefits for perfusion measurements.

"Now CT has a much greater chance of being accepted for stroke evaluation," said Dr. Suresh Mukherji, director of neuroradiology at the University of Michigan Health System in Ann Arbor.

If 64-slice CT has a standout, it is cardiac imaging. Sixty-four-slice technology offers diagnostic-quality coronary CT angiography for identifying stenotic disease and the ability to assess risk with CT coronary calcium scoring. And there is no reason radiologists shouldn't be using it for these purposes.

At New York University, radiologists perform heart scans and coronary CTAs, examining plaques and stents and reporting their findings to cardiologists.

"We have decided to make it a joint effort," said Dr. Jill Jacobs, chief of cardiac imaging at NYU School of Medicine. "Cardiologists provide clinical input and we, as radiologists, provide the image expertise on the CT and 3D reconstruction side."

Radiologists absolutely must become involved in cardiac imaging, said Dr. Geoffrey D. Rubin, chief of cardiovascular imaging at Stanford University.

"If they don't, they risk losing other clinical activities they do with CT," he said. "If radiology is going to remain the leader in CT imaging, then radiologists need to keep up. They need to use 64-row and more advanced CT technology when it comes along and get involved in coronary CTA and other cardiac applications."

The clinical reach of radiologists allows them to examine patients not only for aortic dissection and coronary artery insufficiency but for pulmonary embolism and other causes of chest pain and shortness of breath, such as pneumonia and rib fracture. The breadth of radiologists' knowledge makes them a critical part of the diagnostic process.

Jacobs scans outpatients, but she expects that 64-slice scanning will eventually be used to evaluate patients in the ER who complain of chest pain and shortness of breath.

"With this, we can take the differential diagnosis of the more likely causes and look at each one in a diagnostic study," she said.

VASCULAR IMAGING

All vascular applications are fair game for 64-slice scanners, whose speed and high resolution are well suited to angiography. A head and neck study done on the University of Michigan's 64-slice scanner is completed in about five seconds, or three times faster than on its 16-slice scanner, Mukherji said.

"The faster we go, the better the images we obtain, because there is greater likelihood we will get the contrast in the arterial phase," he said.

A single run with a 64-slice scanner produces high-resolution 3D images of the aortic arch, carotid and vertebral arteries, and the circle of Willis. Dr. Kazuhiro Katada, a professor of radiology at Fujita-Health University School of Medicine in Japan, describes 64-slice technology as the ideal tool for morphologic and functional evaluation of cerebral ischemic disease.

"Sixty-four-row CT makes it possible to cover all the important anatomy-the arch, neck, and circle of Willis-at one time," he said. "Before this, they had to be scanned separately."

The amount of contrast and radiation dose are reduced because tissue overlap that is unavoidable in separate scans can be eliminated. A single 64-slice exam also decreases costs and the time from diagnosis to treatment, Katada said. He predicts that this capability will make 64-row CT a standard tool for the evaluation of cerebral ischemic stroke.

GENERAL RADIOLOGY

Mukherji, who routinely performs 64-slice CTA studies of the brain and carotids, contends that 64-slice CT is a necessity for general radiology.

"We're able to see the smaller vessels much better and faster, as well as stenosis in the carotid," he said. "Patients who would normally need to undergo an angiogram to diagnose the extent of stenosis can now have a CT scan instead. General radiologists have to be sure they can do brain and neck CTs, because those are still the bulk of studies being done on a CT scanner."

Dr. Elliot Fishman, director of diagnostic imaging and body CT at Johns Hopkins Medical Institutions, has compiled protocols for every part of the body (http://www.ctisus.org/ mdct64/protocols/ protocols.html). His guidelines balance image quality and dose for enhancing vasculature and soft tissue. They include head and neck protocols for the carotids, abscesses, and masses; cardiac protocols for calcium scoring and coronary CTAs; chest exams of the airway and aorta; musculoskeletal studies of the abdomen, liver, kidney, colon, and pelvis; and even a whole-body screen.

Adhering to specific protocols is important because 64-slice scanners are much faster than previous generations of scanners. In oncologic studies, a 64-slice CT scan may be complete before the contrast completely opacifies the tumors. In peripheral runoffs, there is a danger that the scanners will outrun the bolus, particularly in patients with peripheral arterial disease, which can substantially delay the bolus.

How studies are done, contrast is injected, and data are acquired and interpreted must all change if 64-slice CT is to be effective. But once these changes are made, 64-slice scanners make CT easier, not harder, according to Fishman.

DOSE REDUCTIONS

The new scanners also offer the potential to reduce the contrast medium dose. Fishman optimizes the dosages by matching them to the applications.

"We do a lot of coronary arteries, and for those I use 80 cc of contrast," he said. "If I want to do a routine abscess, I can go 100 cc, no problem. But if we're looking for liver metastases, we need a lot more iodine."

X-ray dose can be managed as well. All the manufacturers offer dose-adaptive protocols: Siemens' CareDose, Philips' DoseRight, GE's Auto mA and SmartmA, and Toshiba's SureExposure. These protocols automatically determine the optimal middle ground between image quality and exposure to match patient size and body area.

But the lower dose from 64-slice scanners can feed the temptation to do more studies than necessary, Fishman said. The scanners are so fast that physicians may be inclined to perform multiphase studies when a single-phase examination would suffice.

"There are no tube heating problems with 64-slice scanners, so it is easy to do everything," he said. "You can do five phases of the kidneys, and we would if it weren't for the radiation. You have to think about whether all that information is needed and how you can minimize the acquisitions."

Dose can easily become an issue in cardiac studies, Fishman said. ECG-gated coronary angiography may expose a patient to three or four times more dose than a nongated exam because gating slows the pitch. An ECG pulsing technique might mitigate the problem, cutting dose during gating by 40% to 70%.

The idea is to turn on the x-ray rube only when it's needed; for example, at diastole. Dose then would be administered only for short bursts rather than for the entire five-second period as it is now. The idea is sound, Fishman said, but the practical aspects have yet to be worked out.

DATA MANAGEMENT

Handling the data acquired is also a concern. A single peripheral runoff on a 64-slice system produces 3500 images. The users of 64-slice CT need PACS and IT capabilities to match.

Postprocessing tools promise to keep radiologists on top of the hundreds or even thousands of images 64-slice scanners generate. Volumetric imaging has become routine at University of Chicago Hospitals.

"Referring doctors insist on 3D reconstructions. They won't tolerate a collection of slices," Vannier said.

But the need for 3D visualization varies with the type of examination. Stanford radiologists routinely evaluate a study with 500 or fewer images using standard PACS software that allows them to flip through one transverse image after the other.

"In the end, it is less about having fancy software tools to reformat the data into beautiful images than it is having robust software tools and hardware that allow you to move through the images efficiently and intuitively," said Rubin, who is codirector of the 3D Laboratory at Stanford University.

If the computer constantly has to swap with the hard drive, and getting the data through the networks is so difficult that rate-limiting bottlenecks are forming, the result is like having a fast car stuck in second gear, he said.

Vendors have developed workstations designed to handle the flood of data. They can process up to 20 images per second, moving among the images and reconstructing them almost instantly. Storage of the data, however, can overwhelm a PACS.

The University of Chicago Hospitals perform about 4000 CT scans every month, each including 1000 to 2000 discrete slices. And those numbers are increasing.

"We have to rethink how we can compress and store so much data yet retain easy access to the images when we need them," Vannier said.

Increasing reliance on computers seems inevitable. Computer-aided detection promises to pinpoint specific regions in the thousands of slices where pathology is most likely to be found. The prospect of applying CAD is especially attractive for lung cancer. But capitalizing on this opportunity is proving difficult.

Even the lung cancer experts cannot agree on the differentiation of benign and malignant lesions, and until they do, the rules for distinguishing one from the other cannot be written into computer code.

"The biggest limitation to lung CAD is trying to figure out what is important to detect," Rubin said.

Further from the limelight are productivity tools that optimize throughput and access to services. They may be running in the background onboard CT scanners and workstations, or they may walk into the site on two legs. GE, for instance, offers Realize, a technology implementation program that mixes computerized tools and staff training to reduce variability and bottlenecks in the exam process.

FUTURE ADVANCES

The need for such programs will increase as technology continues to advance. The speed and coverage of the latest generation of scanners allow studies of dynamic processes previously not possible.

Several years ago, Mukherji proposed to work with University of Michigan oncologists in using a 16-slice scanner to evaluate therapy for cancer of the tongue. Perfusion changes from pre- to post-treatment indicated the effect of treatment. But oncologists turned down the proposal because the 16-slice scanner was too limited; scans covered just 2 cm. The 64-slice system covers twice the area, and the Ann Arbor physicians are now working together to study CT perfusion as a predictive measure for the efficacy of cancer therapy. Radiologists and oncologists at the University of Chicago Hospitals are involved in similar research.

"Instead of just monitoring changes in tumor size, we can watch the perfusion of a contrast agent as it moves toward, around, and through a tumor," Vannier said. "This can provide an early view of how a patient is responding to therapy. It helps us predict, rather than simply describe, responses to treatment."

Technological achievements in the future will increase the flexibility of the acquisition itself. Multiple x-ray sources generating higher peak x-ray power will be matched to multiple detectors-or a segmented detector-capable of producing higher spatial resolution as well as fluoroscopic and radiographic imaging. Operators will be able to vary the isotropic spatial resolution of the images to home in on regions of interest.

The continued evolution of automatic exposure control will advance CT as a low-dose modality. The journey to this point has already begun with algorithms that balance dose and image quality.

As CT becomes more powerful, covering more area in less time and providing more diverse clinical opportunities, getting the most from this equipment will present challenges. Vendor engineers will continue their efforts to keep users ahead of the curve, building smart technologies that translate clinically-based commands into those that tell the machine what to do. But the greatest challenge for imagers will be recognizing the machines' capabilities and discovering how to capitalize on them, especially in applications that traditionally have been outside the reach of radiologists.

Mr. Freiherr is business editor of Diagnostic Imaging.