The Stanford International Symposium on Multidetector-Row CT has gained such momentum in its nine-year history that the next meeting will be in a larger venue in Las Vegas. The growing importance of multislice CT is reflected in the wide range of topics presented this year.
The Stanford International Symposium on Multidetector-Row CT has gained such momentum in its nine-year history that the next meeting will be in a larger venue in Las Vegas. The growing importance of multislice CT is reflected in the wide range of topics presented this year. Several themes stood out as important:
At the end of the four-day symposium, the team of lecturers provided a comprehensive update for mid-2007 and offered a preview of what to expect for MSCT technology and clinical practice in the next six to 12 months.
While technological advances are opening new opportunities, radiation safety and data overload present serious challenges.
The dual-source scanner has two x-ray tubes, resulting in a half-gantry rotation time of 83 msec. The two tubes can combine their individual 80-kW output at similar kVp and mA settings to increase the effective mA. Cynthia McCollough, Ph.D., a medical physicist at the Mayo Clinic in Rochester, MN, pointed out that this may allow use of higher pitch values, which could lead to shorter scans, improved enhancement, or thinner collimation.
Alternatively, the x-ray tubes can be set at different kVp and mA settings to generate dual-energy data sets. Data sets can be separated using postprocessing algorithms. Applications include CTA bone and calcium removal (i.e., peripheral arterial disease), generation of virtual unenhanced images, and characterization of renal calculi (Figures 1 and 2).
Dr. Kazuhiro Katada, a professor of radiology at Fujita Health University in Toyoake, presented his experience with Toshiba's 256-slice CT system. Johns Hopkins University Medical Center has a prototype of this scanner installed, and Toshiba anticipates commercial availability next year. The 256-slice system can scan a maximum of 128 mm. Depending on the distance to be covered, no more than a single rotation may be necessary, obviating the need for helical scanning. Complete coverage of the head or heart is possible with a single rotation. Scan time is significantly reduced, which is advantageous for critically ill patients and those who may require sedation. Scans are acquired with isophasic imaging, which permits more uniform delivery of contrast medium. Alternatively, the same organ can be scanned sequentially without table motion, allowing for dynamic functional evaluation, including perfusion and time-resolved CT angiography. This technique can also be used to generate unenhanced and contrast-enhanced images (subtraction). An interesting adjunct is its ability to perform scans at two different energies within 0.5 seconds of each other for tissue characterization.
The clear message was ALARA. Achieving ALARA begins with patient selection, adherence to exam indications, and eliminating unnecessary repeat exams. Consideration must be given to a patient's age, sex, and radiosensitivity, as well as the possibility that clinical problems could be addressed with a nonionizing modality such as ultrasound or MRI. Risks of radiation should always be weighed against the benefits of the exam. Multiple lectures stressed the importance of consistently using radiation reduction strategies when acquiring a study: using lead shielding, limiting the scan range, decreasing the tube voltage (kVp), and using attenuation-tube current (mA) modulation.
When using ECG gating for a cardiac or thoracic aortic study, ECG tube-current modulation should also be employed and a low heart rate targeted to ensure effective modulation. However, very low may necessitate increased scan redundancy and therefore increase dose. In discussions on ECG-gated radiation exposure reduction, the liveliest interest centered on multisource versus prospectively gated step-and-shoot scanning. Speakers presented data showing radiation exposure reduction with both approaches. Which approach will prove most effective remains an area of active research.
MSCT exams may be viewed on PACS, workstation, or both. Assessment on a workstation can be performed by a thin-client, thick-client, or stand-alone 3D system. With the increasing data load in imaging practices, stand-alone systems and thick-client solutions can put a significant strain on productivity. On the one hand, there is redundant data transmission and archiving. On the other, these workstations may not be readily accessible to all who need to interrogate an exam with advanced tools. The coming solution will be full integration of 3D workstation applications (e.g., vessel analysis, lung nodule analysis, lumen fly-through, CT perfusion, etc.) into PACS, using a thin-client model. Imagers and physicians will be able to view MSCT studies and perform advanced processing with a single hospital enterprise system, onsite or remotely, by intranet or over the Internet, depending on the available user bandwidth. TeraRecon, the first vendor to enter the market with a thin-client solution, leads this drive with the planned release of its iNtuition workflow architecture at the 2007 RSNA meeting.
Continued advancements in automated intelligent preprocessing, clinical analysis tools, multimodality integration, and time-resolved study analysis are all actively being pursued. It is anticipated that only thin-section source data will be stored and diagnostic images reconstructed in real-time on PACS to the desired section thickness. On-demand or automatically launched computer-aided detection algorithms (e.g., lung and colon MSCT exams) will be more widely used with PACS to improve reader performance in high-volume practices. Finally, as the technology places greater demand on clinical workflow, the role of a 3D laboratory for quantitative analysis and treatment planning will likely gain more prevalence in imaging practices.
Many of these issues came to light during the symposium's workstation face-off. Veteran participants Siemens, GE, Vital Images, TeraRecon, and Philips Medical Systems were joined by Carestream and Barco. Each vendor presented four cases: an aortic stent-graft CT, pancreas protocol CT, cardiac CTA, and renal artery CTA. The aortic stent-graft was a noncontrast follow-up to a prior CTA. The challenge was to demonstrate study coregistration with image processing. TeraRecon seemed to be the fastest, using automatic segmentation.
The purpose of the pancreas case was to use advanced tools to segment out the pancreas, pancreatic duct, and vasculature. Variability between presenters was wide, with some producing oblique minimum intensity projections (MinIPs) and others generating curved planar reformations (CPRs). This was also an opportunity for some vendors to demonstrate labeling tools, with TeraRecon and GE using a menu option system and Vitrea using manual labeling. The cardiac CT case was a standard multiphase 4D data set. However, the task was to analyze morphology and function using a thin-client solution at 4 MHz bandwidth. For this case, GE's Advantage Windows had no thin-client server, and Siemens' Leonardo was operated at a 10-MHz connection, as it could not perform at the mandated modem speed. While load times varied, TeraRecon, Carestream, and Vital Images were noticeably faster than Barco and Philips. TeraRecon was the obvious leader in the cardiac CT demonstration, with near seamless processing of wall motion, ejection fraction, coronary artery autosegmentation, and vessel analysis.
Applications in many areas of clinical practice were discussed.
Lectures on cardiac CT use for coronary artery disease stressed the importance of patient selection and preparation, particularly for patients presenting to the emergency department with acute chest pain. Calcium scoring, plaque morphology, and CT performance for coronary artery disease were all reviewed. Lectures on noncoronary artery disease addressed aortic valve assessment, pulmonary vein mapping, and noncardiac anatomy and findings on a cardiac CT exam.
The vascular session focused on aortic, renal, peripheral arterial, and venous MSCT applications. The key take-home messages from these vascular lectures were that high clinical utility is achieved with CTA and CT venography in the vascular system, and that clinical success relies on technical success. Aorta topics included practical reviews of dissection, acute inflammatory conditions of the aorta, ulcer like lesions of the aorta, and endoleaks. The renal CTA lecture emphasized broad application, including renal donor evaluation, renal artery stenosis (ostial and fibromuscular dysplasia), dissection, vasculitis, and aneurysms. Discussions of the peripheral arterial system and occlusive disease focused on MSCT technique, CTA pitfalls, the role of MRI, and dual-energy application with dual- and single-source 64-slice scanners for bone and calcium removal. The lecture on indirect and direct MSCT venography addressed clinical indications, technique, and pitfalls.
The second part of the abdominal lectures addressed virtual CT colonography and CT enterography. The main issues raised in CTC were the use of minimal, noncathartic preparation and CAD. Minimal-preparation CTC employs barium and iodine oral contrast for fecal tagging, which can then be electronically subtracted, although this results in artifacts that interfere with CAD. The overall consensus was that minimal preparation CTC with CAD is a technology that is very nearly ready for widespread clinical use, but some significant software innovations will be necessary first.
The main points addressed in CT enterography were the use of negative enteric contrast and the necessity of multiplanar reformats. Negative contrast, such as E-Z-EM's VoLumen or a suspension of mannitol, increases the conspicuity of the submucosa, vessels, and calcifications. The use of coronal imaging for improved characterization of jejunal folds and for mesenteric processes was stressed.
From the same isotropic acquisition, a soft-tissue data set can be reconstructed using a different filter. This second data set can then be applied to display and render skin, soft tissues, tendons, muscles, viscera, and vessels. This multireconstruction approach has revolutionized the assessment, triage, and management of polytrauma patients. As discussed by Dr. Jorge Soto, an associate professor of radiology at Boston Medical College, vascular, visceral, soft-tissue, and musculoskeletal structures can be evaluated using a single bolus of contrast.
PET/CT can be used to characterize nodules and assess tumor physiology. However, false-positive results due to infectious and inflammatory diseases are not uncommon. Additional topics in the thoracic section addressed reducing radiation in routine thoracic CT, advanced applications and strategies to ensure appropriate utilization and radiation protection for CT pulmonary angiography, and, finally, strategies to optimize CTPA enhancement when a patent foramen ovale may be present.
For more information regarding this symposium, or to order a symposium CD-ROM complete with the speakers' exam protocols, please go to mdct.stanford.edu.
The 10th Annual International Symposium on Multidetector-Row CT will be held May 13-16, 2008, at the Wynn Hotel in Las Vegas.
Dr. Hellinger is director of cardiovascular imaging and 3D imaging at The Children's Hospital of Philadelphia, and Dr. Margolis is an assistant professor of radiology at the University of California, Los Angeles.
Mammography Study Suggests DBT-Based AI May Help Reduce Disparities with Breast Cancer Screening
December 13th 2024New research suggests that AI-powered assessment of digital breast tomosynthesis (DBT) for short-term breast cancer risk may help address racial disparities with detection and shortcomings of traditional mammography in women with dense breasts.
FDA Clears Updated AI Platform for Digital Breast Tomosynthesis
November 12th 2024Employing advanced deep learning convolutional neural networks, ProFound Detection Version 4.0 reportedly offers a 50 percent improvement in detecting cancer in dense breasts in comparison to the previous version of the software.