Note: This copy is for your personal non-commercial use only. To order presentation-ready copies for distribution to your colleagues or clients, contact us at www.rsna.org/rsnarights.
New and Evolving Concepts in CT for Abdominal Vascular Imaging1 Jorge M. Fuentes-Orrego, MD Daniella Pinho, MD Naveen M. Kulkarni, MD Mukta Agrawal, MD Brian B. Ghoshhajra, MD Dushyant V. Sahani, MD Abbreviations: AAA = abdominal aortic aneurysm, DECT = dual-energy CT, dsDECT = double-source DECT, EVAR = endovascular aortic repair, MD = material decomposition, MIP = maximum intensity projection, MPR = multiplanar reconstruction, SECT = singleenergy CT, ssDECT = single-source DECT, 3D = three-dimensional, 2D = two-dimensional, VM = virtual monochromatic RadioGraphics 2014; 34:1363–1384 Published online 10.1148/rg.345130070 Content Codes: From the Division of Abdominal Imaging and Intervention, Department of Radiology, Massachusetts General Hospital, Harvard Medical School, 55 Fruit St, Boston, MA 02114. Presented as an education exhibit at the 2011 RSNA Annual Meeting. Received August 30, 2013; revision requested December 18; final revision received January 18, 2014; accepted March 24. For this journal-based SA-CME activity, the authors B.B.G. and D.V.S. have provided disclosures (see p 1382); all other authors, the editor, and the reviewers have disclosed no relevant relationships. Address correspondence to D.V.S. (e-mail: [email protected]). 1
SA-CME LEARNING OBJECTIVES After completing this journal-based SACME activity, participants will be able to: ■■Review practical and novel approaches to improving the quality of CT angiography. ■■Discuss
the applications of DECT in CT angiography. ■■Describe
strategies and newer technologic solutions for improving patient safety with radiation and contrast media exposure. See www.rsna.org/education/search/RG.
Computed tomographic (CT) angiography has become the standard of care, supplanting invasive angiography for comprehensive initial evaluation of acute and chronic conditions affecting the vascular system in the abdomen and elsewhere. Over the past decade, the capabilities of CT have improved substantially; simultaneously, the expectations of the referring physician and vascular surgeons have also evolved. Increasingly, CT angiography is used as an imaging biomarker for treatment selection and assessment of effectiveness. However, the growing use of CT angiography has also introduced some challenges, as potential radiation-associated and contrast media–induced risks need to be addressed. These concerns can be partly confronted by modifying scanning parameters (applying a low tube voltage) with or without using software-based solutions. Most recently, multienergy technology has endowed CT with new capabilities offering improved CT angiographic image quality and novel plaque characterization while decreasing radiation and iodine dose. In this article, we discuss current and new approaches using both conventional and multienergy CT for studying vascular disease in the abdomen. We propose various approaches to overcoming commonly encountered image quality challenges in CT angiography. In addition, we describe supplemental strategies for improving patient safety that leverage the available technology. ©
RSNA, 2014 • radiographics.rsna.org
Introduction
Vascular imaging has evolved rapidly during the past few decades, especially since the introduction of multidetector computed tomography (CT), with new hardware and software improvements allowing higher spatial and temporal resolution with implementation of three-dimensional (3D) volume-rendered images (1). These
VASCULAR/INTERVENTIONAL RADIOLOGY
1363
1364 September-October 2014
advancements have enhanced the capability of radiologists to assess vascular anatomy and disease involving not only large but also mediumsized and small vessels as well. Since CT images are acquired quickly and generated nearly in real time, there are no major workflow constraints; thus, CT angiography has become the imaging test of choice in most clinical settings (2,3). In addition, multidetector CT is widely available, cost-effective, reliable, and noninvasive. Despite these recognized benefits, use of multidetector CT has introduced new challenges for vascular imaging, as both radiation- and contrast media– related risks need to be addressed, particularly as utilization increases (4,5). The implementation of dual-energy CT (DECT) with novel applications such as material decomposition (MD) and virtual monochromatic (VM) imaging presents opportunities to assess vascular disease from a new perspective, expanding the role of CT angiography in vascular imaging from a traditional anatomic-morphologic tool to a functional and quantitative tool. This could significantly affect patient care as our knowledge of, experience with, and access to this innovative technology increase. In this article, we discuss (a) new approaches to improving the image quality of CT angiographic scans with current and new postprocessing techniques, (b) novel methods for functional and quantitative assessment of vascular disease and their potential clinical implications, and (c) various approaches to achieving safer vascular imaging acquisitions by reducing both radiation exposure and contrast media dose through protocol modifications that leverage available technologies.
Overview: Technological Considerations
Applications of CT for vascular imaging emerged with the introduction of multidetector CT scanners that integrated multiple detector elements (16–320 detector rows), which transformed the imaging capability of CT from two-dimensional (2D) cross-sectional “slices” to rapid isotropic acquisitions and the generation of 3D volumetric images. These technological advances, together with improvements in image reconstruction software, increased the temporal and spatial resolution as well as the processing workflow, which enabled widespread use of CT angiography and the advent of multiphasic imaging. Furthermore, these advances laid the technical foundation for innovations such as functional imaging via perfusion CT and coronary CT angiography (6,7). Despite these advancements in single-energy CT (SECT), the scan protocols are often complex,
radiographics.rsna.org
necessitating multiphase scan acquisition, and the timing of initiating scanning and contrast media injection factors remain critical to the success of CT angiography. Radiation risks and contrast media–related renal injury are increasingly cited counterarguments to the use of CT (4,5). Technological innovations like DECT propose a unifying solution to the issues related to multiphase scanning, contrast medium volume, and radiation dose (8,9). Detailed descriptions of DECT have been published previously (10,11); however, to familiarize the reader with key concepts, we provide a brief description of the principles and operation of commercially available DECT scanners.
Principles of DECT
In contrast to conventional SECT, DECT encompasses the concomitant acquisition of low-energy (ie, 80–100-kVp tube potential) and high-energy (140-kVp) data (10). This technique plays on the advantage of differing interactions of x-ray photons with different materials. The Compton scatter and photoelectric effects represent the principal interactions that occur between materials and x-ray photons. The photoelectric effect describes the increased attenuation perceived when an incident photon expels an electron from the innermost orbit (k-shell) of the atom. Conversely, Compton scatter represents the interaction occurring when an incident photon ejects an electron from the outer shell of the atom, thus resulting in photon scattering, which serves no purpose for CT imaging (10). The probability of photons undergoing the photoelectric effect increases for materials with a high atomic number (Z) at a low range of energies. For instance, materials of interest for CT angiographic examinations include iodine (Z = 53) and calcium (Z = 20), which display distinctive attenuation values at differing low energies. This peculiarity allows “spectral separation” and creation of material-specific attenuation (ie, MD) images, such as iodine, water, fat, or calcium images. These images can serve different purposes. For instance, water images may substitute for noncontrast examinations by displaying only water densities (attenuations) rather than iodine densities; conversely, iodine images aid in detecting lesion enhancement and enable quantification of tissue iodine enhancement. Furthermore, through advanced postprocessing techniques, generation of virtual monochromatic (VM) images is possible. These are virtual images that simulate those hypothetically acquired using an x-ray beam containing photons of only a single energy level. These energy levels are described in terms of kilo–electron volts (keV) instead of peak kilovoltage (kVp), and
RG • Volume 34 Number 5
their upper limit is determined by the tube voltage (ie, 80, 100, or 140 kVp) used during the DECT acquisition. Images targeted to a broad spectrum of energies can be generated for VM imaging, ranging from 40 to 140 keV (11).
DECT Scanners
DECT images are acquired with (a) a single x-ray tube with rapid kilovoltage switching (single-source or ssDECT) (Discovery 750HD; GE Healthcare, Milwaukee, Wis) or (b) a configuration of two xray tubes (double-source or dsDECT) (Somatom Definition and Definition Flash; Siemens Medical Solutions, Forchheim, Germany). In the ssDECT system, fast switching alternates between 80 and 140 kVp in under 0.5 msec throughout the gantry rotation. This approach preserves the native temporal registration of the scan for both sets of images. In ssDECT, both tube voltages (80 and 140 kVp) have to be coupled with a fixed tube current– time product (12). This milliamperage restriction could have resulted in a slight increase in the radiation dose compared with SECT acquisition. However, recent software upgrades in ssDECT have enabled choosing various predefined milliampere settings ranging from 275 to 640 mA. Therefore, doses are lower and can be customized to patient body size and body part. The updated scanning parameters deliver a volume CT dose index (CTDIvol) ranging from 8.96 to 19.1 mGy for data acquisition in the entire abdomen and pelvis (ACR Dose Index Registry: https://nrdr.acr.org/Portal/DIR/ Main/page.aspx), which mirrors the doses reported for SECT acquisitions. In dsDECT scanners, high- and low-energy images are acquired simultaneously by scanning with two perpendicular x-ray tubes, set at different tube voltages. The dsDECT system permits more flexibility for selecting tube voltages for the low-energy tube, as it can be set at 80 or 100 kVp. The determining factor for choosing the tube voltage is patient body size (80 kVp for patients 90 kg]), since image quality can be inadequate at low–kilovolt peak acquisitions in heavier people. In addition, dsDECT scanners integrate tube current modulation on both x-ray sources, thus achieving radiation exposures similar to those of conventional SECT examinations at 120 kVp. Despite its many attributes, DECT poses some limitations. First, the cost of DECT scanners is substantially high compared with that of conventional SECT, which might restrict its availability. DECT introduces some operational complexity, as the scanner can be used for both DECT and SECT acquisition, therefore increasing the number of scan protocols available to the technologist for scanning each body part. Patient
Fuentes-Orrego et al 1365
body size or presence of metallic hardware can compromise the image quality, since large body habitus (total body weight >350 lb [>157.5 kg]) or metal hardware is likely to produce streak artifacts on the images. The field of view (FOV) on dsDECT scanners is restricted to 26–33 cm because of the fan angle of the high-energy x-ray tube. This limitation may preclude acquisition of DECT data of clinically relevant lesions located beyond the FOV on dsDECT scanners (12). Additional image datasets such as MD pairs or VM series can add to the processing time and some delay in scanning the subsequent patient. DECT acquisitions provide rich datasets for lesion and tissue analysis. Besides the generation of MD and VM images with both DECT approaches, ssDECT creates a quality series using the 140-kVp scan, which constitutes conventional SECT images. On the other hand, dsDECT scanners produce low-energy (80 or 100 kVp) and fusion images in addition to the high-energy datasets (140 kVp), thus creating an additional image series as compared with the ssDECT approach. The fusion images from dsDECT result from linear blending of both high- and low-energy series; the contribution of each energy to the blended image is modifiable but generally involves 60% and 40% weighting from each tube energy (Fig 1). The protocol for CT angiographic studies at the authors’ institution is detailed in Table 1.
Improving Image Contrast in CT Angiography Low–Tube Voltage Approach: SECT High intravascular attenuation with the least amount of image noise is a desirable expectation from a diagnostic CT angiographic examination to generate aesthetically pleasing 2D and 3D images. Similarly, high spatial resolution is essential to reliably assess pathologic conditions affecting medium-sized and small vessels at CT angiography. SECT CT angiographic images are typically acquired at 120–140 kV to ensure optimal image quality in patients of various body sizes and to facilitate an easier workflow (13). On the other hand, scanning at low kilovoltage (70–100 kV) leads to higher iodine attenuation, since the energy of the incident x-ray beam is closer to the k-edge of iodine (14,15) (Fig 2). In general, the CT attenuation value of iodine at 80 kV is almost twice that at 140 kV (16); each stepwise increase in kilovoltage results in a 25%–35% decrease in attenuation but a 50% decrease in noise. By applying low kilovoltage, substantial dose reduction can be achieved due to the quadratic relationship between kilovoltage and radiation dose (17,18). In high-contrast studies like CT
1366 September-October 2014
radiographics.rsna.org
RG • Volume 34 Number 5
Fuentes-Orrego et al 1367
Table 1: Scanning Protocol for DECT Angiography at the Authors’ Institution dsDECT
ssDECT
Patient Weight (kg)
Tube Voltage (kV)
Tube Current (mA)
Pitch
Rotation Time (sec)
Tube Voltage (kV)
Tube Current (mA)
Pitch
Rotation Time (sec)
90
80/140* 100/140
ATCM† ATCM
1.375 1.375
0.5 0.8
80/140 80/140
600 fixed 600 fixed
1.375 1.375
0.5 0.8
*Tube A/tube B. Tube A = low-energy x-ray source, tube B = high-energy x-ray source with limited field of
view (26–33 cm). †ATCM = automated tube current modulation.
angiography, the contrast-to-noise ratio takes precedence over image noise and relatively higher image noise can be accepted in comparison with that of other diagnostic abdominal CT examinations. Therefore, low-kilovoltage imaging yields a dual benefit of improving image contrast and decreasing radiation dose. To make the low-kilovoltage technique effective without any degradation of image quality (ie, noise penalty), appropriate modifications (increases) in the tube current (milliamperage or milliampere-second value) are essential. This compensates for the increased image noise often experienced due to the photon-starvation phenomenon, which is responsible for degradation of image quality (17). Despite these desirable gains, the low-kilovoltage technique is not frequently applied at most CT departments due to lack of familiarity with appropriate kilovoltage and milliamperage selections and other parameter adjustments that negatively affect the CT workflow (19). The kilovoltage selections can be based on body size (body mass index, abdominal girth, or total weight) and require end-user adjustments in other scanning parameters (Table 2) (19–21). To overcome the challenge of selecting the optimal settings, two CT manufacturers (Siemens Healthcare, Erlangen, Germany; Toshiba Medical Systems, Tokyo, Japan) have developed software solutions (22,23). One software algorithm (CARE kV; Siemens Medical Solutions) automates the kilovoltage selection based on the attenuation profile of the scout image of the relevant body part, the
desired contrast-to-noise ratio, and the relevance of iodine attenuation (noncontrast examination vs routine examination vs CT angiography). The software also accounts for the necessary milliamperage adjustments to maintain diagnostic-quality images (22). At our institution, we have found that for most (>50%) of our patients undergoing routine CT of the abdomen, a lower kilovoltage (80–100 kV) was selected by the software over our default protocol of 120 kV (24). Two other groups have reported that in more than 50% of patients undergoing CT angiography, a lower kilovoltage (80–100 kV) was prescribed over that in their followed protocols (25,26). On the other hand, Toshiba’s approach (Adaptive Iterative Dose Reduction [AIDR3D]) guides the user to select the correct kilovoltage for the patient. After the scout image is acquired, the system shows the modulation planned for the patient. If the milliamperage planned reaches the maximum level in most of the scanning region, that means the kilovoltage selected is too low for the examination; therefore, one incremental step in the kilovoltage is needed. When the selected kilovoltage is optimal, the milliamperage modulation calculated would be low (23). At our institution, a low–tube voltage approach for CT angiography is followed based on the patient’s total body weight (Table 3).
Application of DECT DECT offers many advantages for vascular imaging. The concurrently postprocessed VM
Figure 1. Axial CT angiographic images obtained with ssDECT (a–d) and dsDECT (e–g) in two patients after endovascular aortic repair (EVAR) who were followed up for endoleak assessment. (a) Conventional SECT image obtained at 140 kVp during the arterial phase shows a type II endoleak (arrow). (b, c) Iodine (b) and water (c) MD images generated from the arterial phase. Arrow in b = type II endoleak. (d) VM image (65 keV) derived from the delayed phase. Note that the endoleak can be confidently detected in the delayed phase on a low-keV image. (e–g) dsDECT images obtained at 140 kVp (e), 100 kVp (f), and 120 kVp (fusion image) (g) during the arterial phase.
1368 September-October 2014
radiographics.rsna.org
Figure 2. Axial (a, b) and coronal (c, d) CT angiographic images in a 53-year-old man after EVAR. The attenuation of iodine in the abdominal aorta on the 80-kVp images (a, c) (average, 479 HU) is decreased to approximately 52% on the 140-kVp images (b, d) (average, 251 HU).
Table 2: Strategies for Use of Low Kilovoltage in Abdominal CT Angiography BMI ≤25 26–30 31–35 ≥36
Abdominal Width (cm)
Body Weight (kg)
Tube Voltage (kV)
≤30 31–45 46–50 ≥51
136
80–100 100–120 120 140
Note.—Strategies are based on body mass index (BMI), abdominal width from the anteroposterior scout image, and the weight-based approach.
images of lower energies (40–65 keV) and MD images from DECT datasets (80/140 kV) can provide substantially higher intravascular contrast over 120- or 140-kV SECT images. This observation was confirmed in a phantom experiment establishing correlation between kilovoltage and kilo–electron voltage (27). The investigators found that the attenuation of iodine at 80 kV approximates that attained at 60 keV; furthermore, they demonstrated that subsequent
incremental steps of 10 keV from this level approximate the attenuation of iodine observed at 100 kV (70 keV), 120 kV (≈77 keV), and 140 kV (≈86 keV). Therefore, the image contrast of VM images increases when images are created using low energies (200 HU (28,68). Typically, intravascular enhancement of >200 HU enables more consistent and reliable visualization of large and small arterial branches in the abdomen, and the quality of 2D and 3D images is also aesthetically pleasing (71). The relationship between the iodine concentration in the contrast medium and intravascular attenuation is linear when the scans are performed at a constant kilovoltage and using similar injection rates. However, the intravascular attenuation almost doubles at lower kilovoltage (80 kV) for similar iodine concentrations (67). The substantial gain in the intravascular enhancement with low kilovoltage can enable iodine dose reduction for CT angiography. However, for a diagnostic-quality CT angiographic examination, it is essential that the duration of contrast medium injection match the scan duration and that the injection rate be >2.5 mL/sec (67). When scanning more than one body part (CT angiography of the abdomen and pelvis), an injection volume of 50–70 mL is needed to ensure uniform intravascular enhancement. When CT angiographic examination is performed with an iodine dose of
New and Evolving Concepts in CT for Abdominal Vascular Imaging1 Jorge M. Fuentes-Orrego, MD Daniella Pinho, MD Naveen M. Kulkarni, MD Mukta Agrawal, MD Brian B. Ghoshhajra, MD Dushyant V. Sahani, MD Abbreviations: AAA = abdominal aortic aneurysm, DECT = dual-energy CT, dsDECT = double-source DECT, EVAR = endovascular aortic repair, MD = material decomposition, MIP = maximum intensity projection, MPR = multiplanar reconstruction, SECT = singleenergy CT, ssDECT = single-source DECT, 3D = three-dimensional, 2D = two-dimensional, VM = virtual monochromatic RadioGraphics 2014; 34:1363–1384 Published online 10.1148/rg.345130070 Content Codes: From the Division of Abdominal Imaging and Intervention, Department of Radiology, Massachusetts General Hospital, Harvard Medical School, 55 Fruit St, Boston, MA 02114. Presented as an education exhibit at the 2011 RSNA Annual Meeting. Received August 30, 2013; revision requested December 18; final revision received January 18, 2014; accepted March 24. For this journal-based SA-CME activity, the authors B.B.G. and D.V.S. have provided disclosures (see p 1382); all other authors, the editor, and the reviewers have disclosed no relevant relationships. Address correspondence to D.V.S. (e-mail: [email protected]). 1
SA-CME LEARNING OBJECTIVES After completing this journal-based SACME activity, participants will be able to: ■■Review practical and novel approaches to improving the quality of CT angiography. ■■Discuss
the applications of DECT in CT angiography. ■■Describe
strategies and newer technologic solutions for improving patient safety with radiation and contrast media exposure. See www.rsna.org/education/search/RG.
Computed tomographic (CT) angiography has become the standard of care, supplanting invasive angiography for comprehensive initial evaluation of acute and chronic conditions affecting the vascular system in the abdomen and elsewhere. Over the past decade, the capabilities of CT have improved substantially; simultaneously, the expectations of the referring physician and vascular surgeons have also evolved. Increasingly, CT angiography is used as an imaging biomarker for treatment selection and assessment of effectiveness. However, the growing use of CT angiography has also introduced some challenges, as potential radiation-associated and contrast media–induced risks need to be addressed. These concerns can be partly confronted by modifying scanning parameters (applying a low tube voltage) with or without using software-based solutions. Most recently, multienergy technology has endowed CT with new capabilities offering improved CT angiographic image quality and novel plaque characterization while decreasing radiation and iodine dose. In this article, we discuss current and new approaches using both conventional and multienergy CT for studying vascular disease in the abdomen. We propose various approaches to overcoming commonly encountered image quality challenges in CT angiography. In addition, we describe supplemental strategies for improving patient safety that leverage the available technology. ©
RSNA, 2014 • radiographics.rsna.org
Introduction
Vascular imaging has evolved rapidly during the past few decades, especially since the introduction of multidetector computed tomography (CT), with new hardware and software improvements allowing higher spatial and temporal resolution with implementation of three-dimensional (3D) volume-rendered images (1). These
VASCULAR/INTERVENTIONAL RADIOLOGY
1363
1364 September-October 2014
advancements have enhanced the capability of radiologists to assess vascular anatomy and disease involving not only large but also mediumsized and small vessels as well. Since CT images are acquired quickly and generated nearly in real time, there are no major workflow constraints; thus, CT angiography has become the imaging test of choice in most clinical settings (2,3). In addition, multidetector CT is widely available, cost-effective, reliable, and noninvasive. Despite these recognized benefits, use of multidetector CT has introduced new challenges for vascular imaging, as both radiation- and contrast media– related risks need to be addressed, particularly as utilization increases (4,5). The implementation of dual-energy CT (DECT) with novel applications such as material decomposition (MD) and virtual monochromatic (VM) imaging presents opportunities to assess vascular disease from a new perspective, expanding the role of CT angiography in vascular imaging from a traditional anatomic-morphologic tool to a functional and quantitative tool. This could significantly affect patient care as our knowledge of, experience with, and access to this innovative technology increase. In this article, we discuss (a) new approaches to improving the image quality of CT angiographic scans with current and new postprocessing techniques, (b) novel methods for functional and quantitative assessment of vascular disease and their potential clinical implications, and (c) various approaches to achieving safer vascular imaging acquisitions by reducing both radiation exposure and contrast media dose through protocol modifications that leverage available technologies.
Overview: Technological Considerations
Applications of CT for vascular imaging emerged with the introduction of multidetector CT scanners that integrated multiple detector elements (16–320 detector rows), which transformed the imaging capability of CT from two-dimensional (2D) cross-sectional “slices” to rapid isotropic acquisitions and the generation of 3D volumetric images. These technological advances, together with improvements in image reconstruction software, increased the temporal and spatial resolution as well as the processing workflow, which enabled widespread use of CT angiography and the advent of multiphasic imaging. Furthermore, these advances laid the technical foundation for innovations such as functional imaging via perfusion CT and coronary CT angiography (6,7). Despite these advancements in single-energy CT (SECT), the scan protocols are often complex,
radiographics.rsna.org
necessitating multiphase scan acquisition, and the timing of initiating scanning and contrast media injection factors remain critical to the success of CT angiography. Radiation risks and contrast media–related renal injury are increasingly cited counterarguments to the use of CT (4,5). Technological innovations like DECT propose a unifying solution to the issues related to multiphase scanning, contrast medium volume, and radiation dose (8,9). Detailed descriptions of DECT have been published previously (10,11); however, to familiarize the reader with key concepts, we provide a brief description of the principles and operation of commercially available DECT scanners.
Principles of DECT
In contrast to conventional SECT, DECT encompasses the concomitant acquisition of low-energy (ie, 80–100-kVp tube potential) and high-energy (140-kVp) data (10). This technique plays on the advantage of differing interactions of x-ray photons with different materials. The Compton scatter and photoelectric effects represent the principal interactions that occur between materials and x-ray photons. The photoelectric effect describes the increased attenuation perceived when an incident photon expels an electron from the innermost orbit (k-shell) of the atom. Conversely, Compton scatter represents the interaction occurring when an incident photon ejects an electron from the outer shell of the atom, thus resulting in photon scattering, which serves no purpose for CT imaging (10). The probability of photons undergoing the photoelectric effect increases for materials with a high atomic number (Z) at a low range of energies. For instance, materials of interest for CT angiographic examinations include iodine (Z = 53) and calcium (Z = 20), which display distinctive attenuation values at differing low energies. This peculiarity allows “spectral separation” and creation of material-specific attenuation (ie, MD) images, such as iodine, water, fat, or calcium images. These images can serve different purposes. For instance, water images may substitute for noncontrast examinations by displaying only water densities (attenuations) rather than iodine densities; conversely, iodine images aid in detecting lesion enhancement and enable quantification of tissue iodine enhancement. Furthermore, through advanced postprocessing techniques, generation of virtual monochromatic (VM) images is possible. These are virtual images that simulate those hypothetically acquired using an x-ray beam containing photons of only a single energy level. These energy levels are described in terms of kilo–electron volts (keV) instead of peak kilovoltage (kVp), and
RG • Volume 34 Number 5
their upper limit is determined by the tube voltage (ie, 80, 100, or 140 kVp) used during the DECT acquisition. Images targeted to a broad spectrum of energies can be generated for VM imaging, ranging from 40 to 140 keV (11).
DECT Scanners
DECT images are acquired with (a) a single x-ray tube with rapid kilovoltage switching (single-source or ssDECT) (Discovery 750HD; GE Healthcare, Milwaukee, Wis) or (b) a configuration of two xray tubes (double-source or dsDECT) (Somatom Definition and Definition Flash; Siemens Medical Solutions, Forchheim, Germany). In the ssDECT system, fast switching alternates between 80 and 140 kVp in under 0.5 msec throughout the gantry rotation. This approach preserves the native temporal registration of the scan for both sets of images. In ssDECT, both tube voltages (80 and 140 kVp) have to be coupled with a fixed tube current– time product (12). This milliamperage restriction could have resulted in a slight increase in the radiation dose compared with SECT acquisition. However, recent software upgrades in ssDECT have enabled choosing various predefined milliampere settings ranging from 275 to 640 mA. Therefore, doses are lower and can be customized to patient body size and body part. The updated scanning parameters deliver a volume CT dose index (CTDIvol) ranging from 8.96 to 19.1 mGy for data acquisition in the entire abdomen and pelvis (ACR Dose Index Registry: https://nrdr.acr.org/Portal/DIR/ Main/page.aspx), which mirrors the doses reported for SECT acquisitions. In dsDECT scanners, high- and low-energy images are acquired simultaneously by scanning with two perpendicular x-ray tubes, set at different tube voltages. The dsDECT system permits more flexibility for selecting tube voltages for the low-energy tube, as it can be set at 80 or 100 kVp. The determining factor for choosing the tube voltage is patient body size (80 kVp for patients 90 kg]), since image quality can be inadequate at low–kilovolt peak acquisitions in heavier people. In addition, dsDECT scanners integrate tube current modulation on both x-ray sources, thus achieving radiation exposures similar to those of conventional SECT examinations at 120 kVp. Despite its many attributes, DECT poses some limitations. First, the cost of DECT scanners is substantially high compared with that of conventional SECT, which might restrict its availability. DECT introduces some operational complexity, as the scanner can be used for both DECT and SECT acquisition, therefore increasing the number of scan protocols available to the technologist for scanning each body part. Patient
Fuentes-Orrego et al 1365
body size or presence of metallic hardware can compromise the image quality, since large body habitus (total body weight >350 lb [>157.5 kg]) or metal hardware is likely to produce streak artifacts on the images. The field of view (FOV) on dsDECT scanners is restricted to 26–33 cm because of the fan angle of the high-energy x-ray tube. This limitation may preclude acquisition of DECT data of clinically relevant lesions located beyond the FOV on dsDECT scanners (12). Additional image datasets such as MD pairs or VM series can add to the processing time and some delay in scanning the subsequent patient. DECT acquisitions provide rich datasets for lesion and tissue analysis. Besides the generation of MD and VM images with both DECT approaches, ssDECT creates a quality series using the 140-kVp scan, which constitutes conventional SECT images. On the other hand, dsDECT scanners produce low-energy (80 or 100 kVp) and fusion images in addition to the high-energy datasets (140 kVp), thus creating an additional image series as compared with the ssDECT approach. The fusion images from dsDECT result from linear blending of both high- and low-energy series; the contribution of each energy to the blended image is modifiable but generally involves 60% and 40% weighting from each tube energy (Fig 1). The protocol for CT angiographic studies at the authors’ institution is detailed in Table 1.
Improving Image Contrast in CT Angiography Low–Tube Voltage Approach: SECT High intravascular attenuation with the least amount of image noise is a desirable expectation from a diagnostic CT angiographic examination to generate aesthetically pleasing 2D and 3D images. Similarly, high spatial resolution is essential to reliably assess pathologic conditions affecting medium-sized and small vessels at CT angiography. SECT CT angiographic images are typically acquired at 120–140 kV to ensure optimal image quality in patients of various body sizes and to facilitate an easier workflow (13). On the other hand, scanning at low kilovoltage (70–100 kV) leads to higher iodine attenuation, since the energy of the incident x-ray beam is closer to the k-edge of iodine (14,15) (Fig 2). In general, the CT attenuation value of iodine at 80 kV is almost twice that at 140 kV (16); each stepwise increase in kilovoltage results in a 25%–35% decrease in attenuation but a 50% decrease in noise. By applying low kilovoltage, substantial dose reduction can be achieved due to the quadratic relationship between kilovoltage and radiation dose (17,18). In high-contrast studies like CT
1366 September-October 2014
radiographics.rsna.org
RG • Volume 34 Number 5
Fuentes-Orrego et al 1367
Table 1: Scanning Protocol for DECT Angiography at the Authors’ Institution dsDECT
ssDECT
Patient Weight (kg)
Tube Voltage (kV)
Tube Current (mA)
Pitch
Rotation Time (sec)
Tube Voltage (kV)
Tube Current (mA)
Pitch
Rotation Time (sec)
90
80/140* 100/140
ATCM† ATCM
1.375 1.375
0.5 0.8
80/140 80/140
600 fixed 600 fixed
1.375 1.375
0.5 0.8
*Tube A/tube B. Tube A = low-energy x-ray source, tube B = high-energy x-ray source with limited field of
view (26–33 cm). †ATCM = automated tube current modulation.
angiography, the contrast-to-noise ratio takes precedence over image noise and relatively higher image noise can be accepted in comparison with that of other diagnostic abdominal CT examinations. Therefore, low-kilovoltage imaging yields a dual benefit of improving image contrast and decreasing radiation dose. To make the low-kilovoltage technique effective without any degradation of image quality (ie, noise penalty), appropriate modifications (increases) in the tube current (milliamperage or milliampere-second value) are essential. This compensates for the increased image noise often experienced due to the photon-starvation phenomenon, which is responsible for degradation of image quality (17). Despite these desirable gains, the low-kilovoltage technique is not frequently applied at most CT departments due to lack of familiarity with appropriate kilovoltage and milliamperage selections and other parameter adjustments that negatively affect the CT workflow (19). The kilovoltage selections can be based on body size (body mass index, abdominal girth, or total weight) and require end-user adjustments in other scanning parameters (Table 2) (19–21). To overcome the challenge of selecting the optimal settings, two CT manufacturers (Siemens Healthcare, Erlangen, Germany; Toshiba Medical Systems, Tokyo, Japan) have developed software solutions (22,23). One software algorithm (CARE kV; Siemens Medical Solutions) automates the kilovoltage selection based on the attenuation profile of the scout image of the relevant body part, the
desired contrast-to-noise ratio, and the relevance of iodine attenuation (noncontrast examination vs routine examination vs CT angiography). The software also accounts for the necessary milliamperage adjustments to maintain diagnostic-quality images (22). At our institution, we have found that for most (>50%) of our patients undergoing routine CT of the abdomen, a lower kilovoltage (80–100 kV) was selected by the software over our default protocol of 120 kV (24). Two other groups have reported that in more than 50% of patients undergoing CT angiography, a lower kilovoltage (80–100 kV) was prescribed over that in their followed protocols (25,26). On the other hand, Toshiba’s approach (Adaptive Iterative Dose Reduction [AIDR3D]) guides the user to select the correct kilovoltage for the patient. After the scout image is acquired, the system shows the modulation planned for the patient. If the milliamperage planned reaches the maximum level in most of the scanning region, that means the kilovoltage selected is too low for the examination; therefore, one incremental step in the kilovoltage is needed. When the selected kilovoltage is optimal, the milliamperage modulation calculated would be low (23). At our institution, a low–tube voltage approach for CT angiography is followed based on the patient’s total body weight (Table 3).
Application of DECT DECT offers many advantages for vascular imaging. The concurrently postprocessed VM
Figure 1. Axial CT angiographic images obtained with ssDECT (a–d) and dsDECT (e–g) in two patients after endovascular aortic repair (EVAR) who were followed up for endoleak assessment. (a) Conventional SECT image obtained at 140 kVp during the arterial phase shows a type II endoleak (arrow). (b, c) Iodine (b) and water (c) MD images generated from the arterial phase. Arrow in b = type II endoleak. (d) VM image (65 keV) derived from the delayed phase. Note that the endoleak can be confidently detected in the delayed phase on a low-keV image. (e–g) dsDECT images obtained at 140 kVp (e), 100 kVp (f), and 120 kVp (fusion image) (g) during the arterial phase.
1368 September-October 2014
radiographics.rsna.org
Figure 2. Axial (a, b) and coronal (c, d) CT angiographic images in a 53-year-old man after EVAR. The attenuation of iodine in the abdominal aorta on the 80-kVp images (a, c) (average, 479 HU) is decreased to approximately 52% on the 140-kVp images (b, d) (average, 251 HU).
Table 2: Strategies for Use of Low Kilovoltage in Abdominal CT Angiography BMI ≤25 26–30 31–35 ≥36
Abdominal Width (cm)
Body Weight (kg)
Tube Voltage (kV)
≤30 31–45 46–50 ≥51
136
80–100 100–120 120 140
Note.—Strategies are based on body mass index (BMI), abdominal width from the anteroposterior scout image, and the weight-based approach.
images of lower energies (40–65 keV) and MD images from DECT datasets (80/140 kV) can provide substantially higher intravascular contrast over 120- or 140-kV SECT images. This observation was confirmed in a phantom experiment establishing correlation between kilovoltage and kilo–electron voltage (27). The investigators found that the attenuation of iodine at 80 kV approximates that attained at 60 keV; furthermore, they demonstrated that subsequent
incremental steps of 10 keV from this level approximate the attenuation of iodine observed at 100 kV (70 keV), 120 kV (≈77 keV), and 140 kV (≈86 keV). Therefore, the image contrast of VM images increases when images are created using low energies (200 HU (28,68). Typically, intravascular enhancement of >200 HU enables more consistent and reliable visualization of large and small arterial branches in the abdomen, and the quality of 2D and 3D images is also aesthetically pleasing (71). The relationship between the iodine concentration in the contrast medium and intravascular attenuation is linear when the scans are performed at a constant kilovoltage and using similar injection rates. However, the intravascular attenuation almost doubles at lower kilovoltage (80 kV) for similar iodine concentrations (67). The substantial gain in the intravascular enhancement with low kilovoltage can enable iodine dose reduction for CT angiography. However, for a diagnostic-quality CT angiographic examination, it is essential that the duration of contrast medium injection match the scan duration and that the injection rate be >2.5 mL/sec (67). When scanning more than one body part (CT angiography of the abdomen and pelvis), an injection volume of 50–70 mL is needed to ensure uniform intravascular enhancement. When CT angiographic examination is performed with an iodine dose of
