Figure 1. Visualization of a vascular islet cell tumor produced on a 16-slice CT scanner. Courtesy of Elliot K. Fishman, MD, Johns Hopkins hospital, Baltimore, MD.

The rapid development of multidetector-row CT (MDCT) scanning is having a profound impact on practice patterns, and in fact has resulted in what amounts to a revolution in imaging capabilities. Compared to other imaging techniques, MDCT has the ability to scan a large area quickly, alleviating the amount of time a patient has to hold their breath. The decreased length of patient breathhold eliminates artifact due to patient respiratory motion and narrower collimation.

The modality also gives radiologists the ability to retrospectively select scan width and provide 1-mm collimation. Reconstruction of individual images is done at a rate of nearly 2.5 images per second as opposed to single-detector CT (SDCT), in which reconstruction of data takes 6-10 seconds per slice. Overall, images returned with MDCT are proving to be of higher resolution and increased detail and quality.1 While MDCT lends itself well to nearly all applications previously taken care of with single-slice scanning, the multidetector-row CT is already changing practice protocols in several key areas: lung cancer screening, general oncology, trauma, and angiography.

Protocol No. 1: Lung Cancer Screening

The ability of the latest MDCT scanners to detect nodules as small as a few millimeters has sparked renewed interest in lung cancer screening,2 which is the focus of the New York Early Lung Cancer Action Project (ELCAP) and the International ELCAP (I-ELCAP). The purpose of both studies is to evaluate the cost-effectiveness of a single baseline low-dose CT scan for lung cancer screening in high-risk individuals-defined as those aged 60 years with at least 10 pack-years of cigarette smoking and no other malignancies.3 “We have been using an eight-slice MDCT for a little over a year in various areas of the facility, and one of those areas is lung cancer screening,” says David Yankelevitz, MD, professor of radiology at Weill Medical College at Cornell University in New York? which is the central site for ELCAP. “This is exciting work, and it’s all driven by the multi-row scanners, which allow us to take images of the entire lung at a very high resolution in a single breathhold.”

The guidelines for scanning currently are quite broad, Yankelevitz explains, though the basic idea is to take the thinnest sections.

“We usually use 1.25-mm collimation, but I suspect we’ll switch to even thinner,” Yankelevitz says. “The thinner the better, really, because it allows us to better estimate the volume, and we want to identify nodules and characterize them in terms of both size and volume. Depending on the scanner, we operate as low as 40 mA and as high as 80 mA.”

Patients in whom an abnormality is identified are scanned again 3 to 6 months later to determine the rate of growth of nodules. So far, the results of repeating the screenings annually found that false positives were uncommon, and 83% of lung cancers discovered were in the earliest, most curable stage. A positive result is defined as newly detected, noncalcified pulmonary nodules with interim growth, including nodules as small as 2 mm.4

“For small nodules, that is probably going to be the best method we’ll have to assess what to do with them,” Yankelevitz says. “For those nodules in the 5-mm-size range, PET scanning is not good at this time, and biopsy is challenging, so we are left with having to follow up on the growth. We want to develop the best protocols and measurements to see how they are growing.”

In fact, as technology evolves, Weill is evolving its protocols as well, and that has so far included developing pathology protocols for how to evaluate the lung cancers that are dissected. Yankelevitz says the team is also at work on establishing what the guidelines will be for screening for lung cancer, and defining the appropriate age and prior risk factors to recommend people entering such programs.

“Cancer screening has the potential to save more lives than virtually any other thing I can think of in medicine in the United States today,” Yankelevitz says. “Lung cancer is the leading cause of cancer death and is a wonderful area to develop the appropriate approach to screening; without question, this is being enabled by advances in technology. “

The imaging community really needs to understand that technology is moving very rapidly and will allow for tremendous improvements in what we’re doing,” Yankelevitz says. “We have to keep up with that technology and evaluate it quickly and efficiently to continue offering the highest level of care. Understanding how to incorporate screening and this imaging aspect of screening efficiently is one of the critical issues in medicine of our time.”

Protocol No. 2: Oncology

Apart from the specialized use of MDCT in screening for lung cancer, the modality also offers advantages in terms of staging known cancers, according to Elliot K. Fishman, MD, professor of radiology and oncology and director of diagnostic radiology and body CT at Johns Hopkins Hospital in Baltimore. Johns Hopkins has been using MDCT for about 5 years, recently shifting from a four-slice to a 16-slice scanner.

MDCT “changes everything,” Fishman says, pointing out that when the vessels become critical, such as in staging pancreatic and renal cancer, the advanced modality allows radiologists to create CT angiograms that are significantly better than those done with the four-slice scanner. For the pancreas, liver, and kidneys, collimation is 0.75, with reconstruction sections done every 0.5 mm, and the scans are done at 120 kV and 160 mA.

“The simple way of looking at it across all applications is that you can do everything better with 16-slice MDCT,” Fishman says. “For instance, we used to be able to reconstruct one image every 8 seconds, but the new machines let us reconstruct six images per second. That certainly does increase the throughput.

“Oncology patients are often more difficult, and they are sicker,” Fishman adds. “A very practical point is that we do not want those patients staying around, and this is easier on the patient. Just think about breathhold in pediatric patients. With MDCT, 99% of our pediatric patients no longer need to be sedated because of the speed of the scan. That is better for us and for the patients.” Although cost is the same from a patient perspective, the MDCT generates around images with much better resolution.

“Because we can scan faster without motion artifact, that increases our level of accuracy,” Fishman says. “Liver lesions can be picked up early and we are better able to stage the degree of vessel invasion. MDCT overall is more convenient and easier.”

Protocol No. 3: Trauma

Figure 2: Sagittal MPR from MDCT of a patient sustaining blunt force injury clearly demonstrates tear in dome of diaphragm (arrowhead) with herniation of fat into the thorax. Courtesy of Stuart Mirvis, MD, University of Maryland Medical Center.

The University of Maryland Medical Center in Baltimore also upgraded from a four-slice to a 16-slice CT recently-6 months ago, in fact-and Stuart Mirvis, MD, professor of diagnostic radiology and director of the section of trauma and emergency radiology, says he was impressed by the significant improvement made by that shift.

“It surprised me, because I thought there would not be as great an improvement as there was when we moved from single-slice to four-slice,” Mirvis says. “The image quality was fine on both, but with the 16-slice we are able to do a thin slice profile and get really good MPRs [multi-planar reconstructions].”

Mirvis was most impressed, however, with the impact MDCT has had on his section of the hospital.

“In my little world of trauma, getting through the scans as quickly as possible with few interruptions is an advantage,” Mirvis says. “We need speed, because we want anyone who is fairly sick to get out as quickly as possible and not deteriorate or arrest on the CT scanner. Before helical CT scans, it took an hour to get images of all the body parts, but now it is a 10-minute-or-less process.”

Loading and unloading the patient onto the scanner actually takes longer than doing the scan, Mirvis says, and so the hospital is looking at how to do that more efficiently.

“With MDCT, one picture is all you need to understand the pathology and how to fix it,” Mirvis says. “The modality is excellent for spine work and CT angiography, because the surgeon also can understand the information in a comfortable way before they operate.”

A number of protocols have been established within the trauma department in order to make use of the MDCT fairly streamlined, and Mirvis says the technologist rarely has to stop and figure out what to do.

“When doing a standard scan, we use 7-mm slice thickness and overlap by 3.2, with a pitch of 1.2 and 0.75 rotation speed,” Mirvis says. “We reconstruct the images every 1.5 mm, and we keep them in storage for about 24 hours. Then we fuse every three images from the dataset and send those to PACS. “A standard scan of the chest, abdomen, or pelvis for trauma uses 1.5 mm x 16-row collimation, a 1.2 pitch, and 0.75-second rotation speed with our system,” he says. “We initially reconstruct and review 5-mm-thick nonoverlapping axial scans. Raw data are saved for 48 hours for further analysis as needed. The raw data are reconstructed at 2 mm with 1-mm overlap for selected indications such as the diaphragm, vascular, or spine injury, or for any questionable finding arising from the original axial reconstructions. These thin slices are used to create all MPR, 3D, and volumetric images. Usually, reformatted images are 2-3 mm thick. We currently save only the original 5-mm axial reconstructions and all postprocessed images to the PACS network to avoid rapid filling of memory. System upgrades in the near future will allow greater flexibility in our image storage paradigm.

“The impact on treatment has been increasing steadily with each generation of CT scanner,” Mirvis continues, with each iteration becoming more sensitive for pseudoaneurysms, for example. ” In the case of the spleen, pseudo- aneurysms usually portend delayed bleeding or rupture, so if we see that, we know to preemptively strike and treat with embolization,” he says. “The whole process of managing injuries in the abdomen nonoperatively has changed with MDCT as well. We can now quantify the amount of free blood in abdomen and vascular injuries.”

As a result, the rate of nonoperative management has gone up steadily, according to Mirvis. “In the spleen alone, we are talking in the mid-90% success rate to get patients through nonoperative management,” Mirvis says. “There are far fewer exploratory surgeries. Better CT scanners have done a much better job at depicting bowel and mesenteric trauma. Diaphragmatic injuries are tough to diagnose, but MDCT has been very good at picking up even small injuries and allowing us to fix them ahead of time.”

Trauma is a polysystemic process and MDCT really lends itself to a multi-system problem, Mirvis points out.

“It provides so much more information than anything else available and it is extremely accurate in an experienced hand,” he says. “We sit and look at cases and sometimes shake our heads in awe at what we are able to see on the new scans. It is so much fun to explain pathology to the other physicians, and when they walk out shaking their heads, we know we’ve made an impression.

“If I were on a desert island and I was going to see a bunch of trauma patients, the MDCT would be the tool I’d want,” Mirvis continues. “It is the ultimate diagnostic tool for trauma, and nothing else is worthwhile to use in this application.”

Protocol No. 4: Angiography

The versatility of MDCT has made it a staple in the interventional neuroradiology department at Johns Hopkins as well, which has two 16-slice scanners-one with 0.5 rotation, the other with 0.4 rotation-and a four-slice scanner. Director Kieran Murphy, MD, notes that another fourth scanner is soon to be added, though “we haven’t yet decided what the layout of that one will be.” In any case, it will be located immediately next to the neurological intensive care unit.

“MDCT are becoming standard, and the better we make them the better we make health care delivery overall,” Murphy says. “CT is the way to go for stroke, cardiac care, acute brain injury and trauma, and MDCT does all the regular studies plus angiography, pulmonary, and pediatric.”

Within the realm of interventional neuroradiology, the MDCT fluoroscopy mode is used often for spinal interventions and for putting catheters into the brain ventricles. And the versatility of the modality has even decreased the number of people in the community who do cerebral angiography, and increased CT angiography (CTA), according to Murphy.

“I look for [techniques] that cross economic barriers, and smaller hospitals can afford MDCT,” Murphy says. “That means the protocols we develop with this will help people all over the world because it is an economically relevant tool.

“MDCT is one of the major events to occur in health care in the last 100 years,” he adds. “Think about it: you can walk into a hospital and have a $200 CTA instead of a $6,000 cerebral angiogram-plus there is no stroke risk, no IV, and it’s an outpatient procedure. That is a huge change.

“The downside perhaps is that this is a destructive technology, because it saves patients money and thus hospitals generate less revenue,” Murphy says. “But it’s the right thing to do for the patient and hopefully will lead to global reduction in procedures and less billing from catheter-based intervention.

“Simplicity, focus, and control,” Murphy says. “Everything we do in medicine is about those things, and these machines help with all of those issues.”

Elizabeth Finch is a contributing writer for Decisions in Axis Imaging News.

Elizabeth Finch is a contributing writer for Decisions in Axis Imaging News.

References:

  1. Angiography: single detector (SDCT) vs multidetector (MDCT) in CT angiography. Available at: www.ctisus.org/multidetector/syllabus/SINGLE_SDCT_MDCT.html.
  2. Marshall D, Simpson KN, Earle CC, Chu C. Potential cost-effectiveness of one-time screening for lung cancer (LC) in a high risk cohort. Lung Cancer. 2001;32:227-36.
  3. Wisnivesky P, Mushlin AI, Sicherman N, Henschke C. The cost-effectiveness of low-dose CT screening for lung cancer: preliminary results of baseline screening. Chest. 2003;124:614-621.
  4. Annual CT screening for lung cancer is cost-effective for high-risk persons, study finds. From the Weill Medical College of Cornell University. Available at: http://www.med.cornell.edu/news/press/2001/june_21_screening.html.