Imaging PhysicsFree Access

Update on Multienergy CT: Physics, Principles, and Applications

Published Online:https://doi.org/10.1148/rg.2020200038

Abstract

Multienergy CT involves acquisition of two or more CT measurements with distinct energy spectra. Using the differential attenuation of tissues and materials at different x-ray energies, multienergy CT allows distinction of tissues and materials beyond that possible with conventional CT. Multienergy CT technologies can operate at the source or detector level. Dual-source, rapid tube-voltage switching, and dual-layer detector CT are the most commonly used multienergy CT technologies. Most of the currently available technologies typically use two energy levels, commonly referred to as dual-energy CT. With use of two or more energy bins, photon-counting detector CT can perform multienergy CT beyond current dual-energy CT technologies. Multienergy CT postprocessing can be performed in the projection or image domain using two-material or multimaterial decomposition. The most commonly used multienergy CT images are virtual monoenergetic images (VMIs), iodine maps, virtual noncontrast (VNC) images, and uric acid images. Low-energy VMIs are used to boost contrast signal and enhance lesion conspicuity. High-energy VMIs are used to decrease some artifacts. Iodine maps are used to evaluate perfusion, characterize lesions, and evaluate response to therapy. VNC images are used to characterize lesions and save radiation dose by eliminating true noncontrast images from multiphasic acquisitions. Uric acid images are used for characterization of renal calculi and gout.

Online supplemental material is available for this article.

©RSNA, 2020

SA-CME LEARNING OBJECTIVES

After completing this journal-based SA-CME activity, participants will be able to:

  • ■ Discuss the physics of multienergy CT and its different implementations.

  • ■ Describe postprocessing of multienergy CT and types of multienergy CT images.

  • ■ List the clinical applications of multienergy CT.

Introduction

Multienergy CT involves acquisition of CT measurements at two or more energy spectra, which allows material characterization beyond that possible with conventional CT. In conventional CT, the x-ray tissue attenuation (linear attenuation coefficient or CT number) depends on the effective energy of the polyenergetic x-ray beam, the material density, and the material effective atomic number (Zeff) (13).

X-ray tissue attenuation in the diagnostic range is primarily through Compton scattering and photoelectric absorption (4). Compton scattering is mainly dependent on the electron density (ρ) (which correlates with volumetric mass density) and is the main determinant of the soft-tissue contrast of low atomic number materials (hydrogen, carbon, nitrogen, and oxygen). The photoelectric effect is directly proportional to Zeff and inversely proportional to the x-ray energy (2,3). Both photoelectric and Compton interactions decrease with increase in x-ray energies. However, the decrease in photoelectric effect is more rapid (5), and there is increased photoelectric attenuation at x-ray energies just higher than the electron binding energy of materials (ie, k edge).

Teaching Point At conventional CT, materials with different atomic numbers may still have similar CT numbers (Hounsfield units) owing to their mass density, thus making material differentiation difficult (4).
The acquisition at different energies in multienergy CT allows material differentiation owing to differential attenuation characteristics of tissues at different x-ray energies.

In current clinical practice, this is achieved by performing two acquisitions at two different energy levels; hence, this is also called dual-energy CT. Spectral CT is another term to denote use of different x-ray spectra to distinguish tissue materials. The terms multienergy CT, dual-energy CT, and spectral CT are used interchangeably.

In this article, we review the principles, technologic implementations, postprocessing, and images of multienergy CT.

Multienergy CT Technologies

There are multiple multienergy CT technologies, some of which operate at the source level and others at the detector level (Table 1).

Table 1: Summary of Various Multienergy CT Techniques

Table 1:

Source-based Technologies

Dual-Source CT.—Dual-source CT has two x-ray tubes offset orthogonal (90°–95°) to each other with their independent detectors. In the dual-energy mode, these x-ray tubes are operated at two different voltages, one at low energy (70, 80, 90, or 100 kVp) and the other at high energy (140 or 150 kVp) (Movie 1). A tin filter can be applied to the high-energy beam to filter out low-energy photons and increase the spectral separation (6).

Movie 1: Dual-source CT. There are two x-ray tubes orthogonal to each other, each operated at a different tube potential.

The main advantage of this technique is that the tube parameters (voltage and current) and filters can be adjusted independently to customize radiation dose, photon flux, and spectra according to the patient’s body habitus and clinical indication. For example, in large patients, 100 kVp/tin filter 140 kVp may be preferred to improve the photon flux and reduce image noise, while in pediatric and extremity imaging, 70 kVp/tin filter 150 kVp may be preferred. The tube current can also be modulated along the x, y, and z axis. Experimental triple- and quadruple-beam configurations with additional filters allow multienergy binning (Tables 1, E1) (9).

The offset of low- and high-energy projection data by approximately a quarter of the rotation time does not allow projection-domain spectral reconstruction (discussed later). Image-domain spectral reconstruction is associated with higher beam hardening and imperfect material separation (10), especially for fast-moving structures. The temporal resolution of the dual-energy mode is lower in the early-generation scanners. Cross-scatter from one x-ray source to the other detector increases the spectral overlap and noise. This is combated by using only a small part of the detector element or by implementing scatter-correction algorithms during image reconstruction (11).

The limited space for the two detectors results in a constrained multienergy CT field of view (FOV) for one of the x-ray tubes. Hence, proper patient positioning is required for obtaining spectral information, especially in large patients. Single-energy CT information is still available outside the smaller FOV for the full 50 cm; but these images are noisier, since only one of the two composite images is used, and are prone to artifacts, since the energy of the beam from the larger tube is not the usual 120-kVp beam. Since the scanner can be operated in multiple modes (also single-energy, high-pitch helical modes), prospective selection of the multienergy mode is required.

Rapid kVp Switching.—Rapid kilovolt peak (kVp) switching CT has a single x-ray tube that alternates rapidly (∼0.25 msec) between low (80-kVp) and high (140-kVp) voltages for each x-ray projection (Movie 2). This technology requires a generator with high frequency and capacitance and a scintillator with fast sampling capabilities (∼50 μsec) (12). Near-simultaneous acquisition of projection data at both voltages results in negligible temporal misregistration and allows spectral reconstruction in both the projection and image domains (10). The single-detector design is also cost-efficient and allows larger multienergy CT FOV (50 cm) and z-axis coverage (up to 16 cm) than the dual-source and dual-layer designs without cross-scatter. Future designs include addition of a filter or filters at the x-ray source and switching the tube potential between more than two voltages (13).

Movie 2: Rapid kVp switching CT. There is a single x-ray tube that alternates rapidly between high and low energies for each x-ray projection.

A swift transition between the two energy states leads to difficulty in modulating the tube current for each kVp, resulting in high noise at the low kVp and high radiation dose at the high kVp. To balance the photon flux and radiation dose, a longer low-kVp state that lasts for 65% of the total exposure time is applied. This time delay also helps in avoiding possible reduction of signal from the low-kVp spectrum (6,14,15). A recent technology (Aquilion ONE/PRISM; Canon, Tokyo, Japan) uses deep learning networks to reduce noise on the basis of differences between output and reference data sets.

The single-tube design of rapid kVp switching precludes optimization of the spectral separation at kVp combinations other than 80 and 140, and placement of additional filters at the tube output is not yet feasible. Overlap of spectral data occurs owing to the voltage switch not being instantaneous and the same filter being used for both energy levels. The time required for voltage modulation between high- and low-energy scanning puts a limit on the gantry rotation speed, which is typically 0.5 seconds or longer (15). This can lead to motion-related artifacts, especially in cardiac imaging. Similar to dual-source scanners, prospective selection of multienergy mode is required.

Dual-Spin CT.—In the dual-spin (sequential) technology, the patient is scanned at two different tube potentials in two consecutive acquisitions (Movie 3). This is the earliest multienergy CT approach that can be performed on any scanner owing to minimal hardware requirements. A dedicated dual-spin mode is available on some volume or wide-detector scanners (Table E1).

Movie 3: Dual-spin CT. The patient volume is scanned consecutively at two different tube voltages.

Single-source sequential CT (Aquilion One and Prime, Revolution EVO; Canon) scans the same volume consecutively at two different tube voltages in axial or helical mode, with fast tube rotation (up to 0.27 msec). Single-source helical multienergy CT (Somatom Definition Edge; Siemens, Forchheim, Germany) performs two consecutive spiral acquisitions at different voltages (80 kVp at pitch of 0.6 and 130–140 kVp at fast pitch of 1.2), each at half the dose of a conventional 120-kVp acquisition (6,16). These latest-generation scanners are able to achieve large volume coverage (up to 16 cm) with minimal delay between the two acquisitions. Optimized tube current modulation and filters can be used for each voltage, and there is no cross-scatter or FOV restriction. Projection-domain spectral reconstruction can be performed for the axial acquisition mode (6).

The most significant limitation is the temporal delay between the two acquisitions, making this highly susceptible to image degradation, especially for highly mobile structures like the heart. Motion artifacts also cause false-positive results at bone edges in patients with gout (17). The temporal offset may also affect angiography, as the contrast medium may run off between the two acquisitions. Alternating the voltage at each gantry position instead of the scan volume may reduce misregistration.

When operated in spiral mode, only image-domain reconstruction is possible and usually involves image registration (16). Section-level decomposition is another challenge. Prospective selection of multienergy mode has to be made (6).

Split-Beam CT.—In split-beam (twin-beam) technology, a filter made of gold and tin is placed at the output of the x-ray tube. Any conventional CT scanner can be upgraded to multienergy CT by using this filter. As the 120-kVp x-ray beam passes through the filter, two beams with different mean energies of 67.5 keV and 85.3 keV emanate from the filter halves—with gold as a beam softener and tin as a beam hardener—before it reaches the patient. These low- and high-energy spectra are detected by the corresponding half of the detector system along the longitudinal (z) direction (Movie 4).

Movie 4: Split-beam CT. A filter composed of gold and tin at the output of the x-ray tube splits the x-ray beam into low- and high-energy components along the z-axis.

Since the low-energy and high-energy data are available at different anatomic regions, helical acquisition with pitch less than 0.5 is used to obtain data from all voxels at both energies (15). To compensate for the beam attenuation by the filter, a higher tube current–time product (milliampere-second [mAs] value) is needed to maintain adequate photon flux. Despite this, image noise and similar radiation dose are comparable to those of single-energy CT (18). A full 50-cm multienergy CT FOV is available.

This technique has suboptimal temporal registration of the low- and high-energy data, since each voxel is imaged with two different energy spectra at different times (15). Spectral separation is also inferior on this platform owing to cross-scatter, inherently lower energies, and indiscernible spectra in the central (owing to a finite 2–3-mm focal spot) and edge portions of the beam (19).

Detector-based Technologies

Dual-Layer CT.—The dual-layer (sandwich detector) technology has a single x-ray tube but two layers of detectors with maximal sensitivities to different photon energies. The top layer of yttrium-based garnet scintillator absorbs most of the low-energy photons, whereas the bottom layer of gadolinium oxysulfide absorbs most of the high-energy photons (Movie 5). The photodiodes and electronics for each detector layer are located on the side of the detectors. Thus, the polyenergetic x-ray beam is separated into two energies after it passes through the patient (20).

Movie 5: Dual-layer CT. There are two layers of detectors, with the top layer absorbing low-energy photons and the thick bottom layer absorbing high-energy photons.

Owing to detector-level processing, the spectral data are perfectly registered spatially and temporally, allowing both projection- and image-domain spectral postprocessing. The temporal resolution is good, and there is no need to correct for shifting or interpolation, which are often seen in source-based systems. No prior selection of multienergy CT protocols or change in workflow is necessary, as the acquisitions are always in multienergy CT mode. All dose modulation techniques can be used. There are no multienergy CT FOV or gantry time limitations. There is no cross-scatter from the source, but detector-level cross talk can occur.

The tube voltage has to be 120 or 140 kVp to maximize the spectral separation. To maintain radiation dose neutrality with conventional CT scanners, the tube current can be proportionally decreased (21). Since the milliampere-second (mAs) value is the same for both energies, noise levels may not be comparable; however, different detector thickness compensates for this effect. Owing to a fixed energy partition between the detector layers, only a single low/high peak tube voltage combination is possible. The spectral separation may not always be accurate, since there is no sharp distinction between low- and high-energy photons at the detector level.

A filter may be added between the detector layers to improve the energy separation, but this will decrease dose efficiency by absorbing low-energy photons that have passed through the patient and increase the overall radiation dose. As most of the low-energy photons are absorbed by the first detector system, there is effective filtration of the beam, resulting in hardening of the spectrum going into the second layer. Reduced spectral separation due to cross talk and close proximity between the detectors as well as use of the same filter for high and low energies is also a potential concern, which can potentially be minimized by a tungsten backbone along the detectors. The z-axis coverage is only 4 cm, which makes perfusion imaging challenging (20).

Photon-Counting Detector CT.—Photon-counting detector (PCD) CT uses a semiconductor detector material (cadmium telluride, cadmium zinc telluride, silicon, gallium arsenide) with direct conversion techniques (without scintillators) to count the individual photons and their associated energies (7,8). The output signal is proportional to the number of photons, and each photon is allocated to a specific energy bin according to its energy (Movie 6).

Teaching Point With two or more energy bins, PCD technology can perform multienergy CT beyond the current dual-energy CT technologies.

Movie 6: Photon-counting detector (PCD) CT. A semiconductor detector converts the x-ray photons directly into electrical signals instead of light photons. The output signal is proportional to the number of photons, and each photon is allocated to a specific energy bin according to its energy.

A major advantage of PCD CT is that the energy threshold (therefore beam spectrum) is configurable, enabling users to select appropriate thresholds to optimize specific imaging tasks. With this flexibility, k-edge imaging (ie, imaging materials with a detectable k edge in the diagnostic x-ray energy range) becomes feasible, benefiting multicontrast and molecular imaging (2224). K-edge imaging requires obtaining measurements below and above the k-edge energy and is limited in conventional multienergy CT owing to fixed spectra, which limit flexibility to adapt to different k edges of materials. Since the energy threshold and energy bin can be freely selected in PCD CT, energy thresholds according to the k edge of specific materials can be selected.

PCD technology also provides higher spatial resolution (eg, 150 μm), reduced or eliminated electronic noise, improved contrast-to-noise ratio, reduced radiation dose, and reduced metal artifacts compared with standard conventional CT (25). PCD CT has perfect spatial and temporal registration, and there are no limitations on FOV, tube current modulation, or tube potentials.

The high flux of CT (several hundred million photons per second) requires superfast electronics to avoid pulse pileup, which may cause two consecutive pulses to be registered as a single pulse or with an error of the registered energy, leading to noise and poor spectral resolution (7,26,27). In addition, nonideal physical effects such as charge sharing (a single photon causing a count event in more than one detector) and k-escape have a significant effect on the energy resolution and spectral separation (8). A combined PCD and dual-source technology can significantly improve the spectral separation and multienergy performance compared with those of both single-source PCD CT and dual-source CT (28). There are currently no commercial PCD CT scanners available, but there are several prototype systems being used for research activities (8).

Multienergy CT Postprocessing and Material Decomposition

Multienergy CT scanners provide CT data capable of generating material-specific images, which provide information about the presence, distribution, and concentration of a given material within the tissues. Given that the attenuation in the diagnostic range is primarily through photoelectric and Compton effects, the attenuation of any material without a measurable k edge can be modeled as a combination of photoelectric and Compton interactions. Since these interactions depend on the atomic number (Z) and mass density (ρ), respectively, a ρZ map can be generated; therefore, the attenuation of any material can also be modeled as the linear combination of attenuation coefficients of two underlying basis materials (4).

Multienergy CT postprocessing can be performed in the projection (prereconstruction) or image domain (Fig 1). Projection-based methods operate on the projection data directly obtained from the CT acquisition, whereas image-based methods operate on the images after they have been reconstructed. Projection-based spectral reconstruction requires spatial and temporal alignment of the low- and high-energy data, which is possible with the rapid kVp switching, dual-layer, and PCD technologies. The axial mode of the dual-spin technology has spatial alignment but not temporal alignment owing to delay between the two spins. Projection-based spectral reconstruction may be possible in the axial mode if there is no substantial motion between the two spins. This technique involves solving a system of two nonlinear equations to generate a photoelectric/Compton or ρ/Z sinogram from the high- and low-energy sinograms in the projection space (20) (Fig 1a).

Spectral reconstruction methods. For both of these techniques, the                    processes highlighted in brown occur in the pre-reconstruction space. (a)                    Projection-based technique. The high- and low-energy sinograms obtained from raw                    x-ray projections are converted to basis material density or                    photoelectric/Compton sinograms by using calibration methods. Material density                    images are then reconstructed in the image space; these can also be linearly                    combined into virtual monoenergetic images (VMIs). (b) Image-based technique.                    The high- and low-energy sinograms are separately reconstructed into high- and                    low-energy images. Material density images are derived from linear solution of a                    set of two or three unknown variables, and the measurements are obtained on                    high- and low-energy images. Material density images can be linearly combined                    into VMIs.

Figure 1a. Spectral reconstruction methods. For both of these techniques, the processes highlighted in brown occur in the pre-reconstruction space. (a) Projection-based technique. The high- and low-energy sinograms obtained from raw x-ray projections are converted to basis material density or photoelectric/Compton sinograms by using calibration methods. Material density images are then reconstructed in the image space; these can also be linearly combined into virtual monoenergetic images (VMIs). (b) Image-based technique. The high- and low-energy sinograms are separately reconstructed into high- and low-energy images. Material density images are derived from linear solution of a set of two or three unknown variables, and the measurements are obtained on high- and low-energy images. Material density images can be linearly combined into VMIs.

Spectral reconstruction methods. For both of these techniques, the                    processes highlighted in brown occur in the pre-reconstruction space. (a)                    Projection-based technique. The high- and low-energy sinograms obtained from raw                    x-ray projections are converted to basis material density or                    photoelectric/Compton sinograms by using calibration methods. Material density                    images are then reconstructed in the image space; these can also be linearly                    combined into virtual monoenergetic images (VMIs). (b) Image-based technique.                    The high- and low-energy sinograms are separately reconstructed into high- and                    low-energy images. Material density images are derived from linear solution of a                    set of two or three unknown variables, and the measurements are obtained on                    high- and low-energy images. Material density images can be linearly combined                    into VMIs.

Figure 1b. Spectral reconstruction methods. For both of these techniques, the processes highlighted in brown occur in the pre-reconstruction space. (a) Projection-based technique. The high- and low-energy sinograms obtained from raw x-ray projections are converted to basis material density or photoelectric/Compton sinograms by using calibration methods. Material density images are then reconstructed in the image space; these can also be linearly combined into virtual monoenergetic images (VMIs). (b) Image-based technique. The high- and low-energy sinograms are separately reconstructed into high- and low-energy images. Material density images are derived from linear solution of a set of two or three unknown variables, and the measurements are obtained on high- and low-energy images. Material density images can be linearly combined into VMIs.

This is achieved with the aid of calibration measurements obtained empirically from the CT system (29,30). These sinograms are then reconstructed with usual methods to generate corresponding multienergy CT images, such as photoelectric/Compton, ρ/Z, or two basis material density maps. Virtual monoenergetic images (VMIs) may be obtained by linear combination of the basis material density maps (31).

Teaching Point Projection-domain material decomposition is associated with fewer beam-hardening artifacts;
however, such artifacts are not completely removed owing to imperfect calibrations (3,29,32).

Multienergy CT processing in the image domain is applicable in dual-source CT and the helical mode of dual-spin CT, which do not have spatial-temporal registration. In this approach, the low- and high-energy sinograms are separately reconstructed into low- and high-energy images (Fig 1b). This is followed by material decomposition to generate two basis material density maps. VMIs may be obtained by linear combination of the basis material density maps (33,34). Image-domain material decomposition has a higher amount of beam-hardening artifacts, which can be reduced by iterative beam-hardening correction in the image domain (35).

Material decomposition helps with classification of materials (ie, present or absent) and quantification of the concentration of a particular material in a mixture. Classification can be done on the basis of the Zeff information or the ratio of CT numbers at low and high energies, called the dual-energy ratio (DER) or CT number ratio (Fig 2). Quantification can be performed with two- or three-material decomposition techniques.

Material decomposition. In material decomposition, the CT numbers of the                    low-energy image (eg, 80 kVp) are plotted along the y-axis, and the CT numbers                    of the high-energy image (eg, 140 kVp) are plotted along the x-axis. Air and                    water are located along the identity line with fixed values of −1000 HU                    and 0 HU, respectively. Objects located apart from each other in this graph (ie,                    with different attenuation numbers at the different energies) can be separated                    by using a multienergy decomposition process. Note the locations of iodine and                    bone, which make them ideal for separation.

Figure 2. Material decomposition. In material decomposition, the CT numbers of the low-energy image (eg, 80 kVp) are plotted along the y-axis, and the CT numbers of the high-energy image (eg, 140 kVp) are plotted along the x-axis. Air and water are located along the identity line with fixed values of −1000 HU and 0 HU, respectively. Objects located apart from each other in this graph (ie, with different attenuation numbers at the different energies) can be separated by using a multienergy decomposition process. Note the locations of iodine and bone, which make them ideal for separation.

Two-material decomposition is used in the rapid kVp switching and dual-layer technologies. This method assumes that each voxel is represented by only two preselected materials in different proportions (3,4) (Fig 3a). Any two materials can be selected for two-material decomposition, depending on the clinical task, and these form the material basis pair (3,4). With this assumption, the amount of these two materials within a voxel is estimated on the basis of their attenuation characteristics at two different energies (Fig 3b, 3c) (3,32).

Two-material decomposition. (a) Graph shows the process of two-material                    decomposition. The entire imaged volume is assumed to be made of only two (pair)                    preselected materials in different proportions. The pixel composition can be                    described by a linear combination of equivalent amounts of the two materials                    (eg, α and β) with known attenuation profiles at the respective                    levels, essentially α cos(φ) + β sin(φ). (b) Iodine                    density map through the midchest, color coded in rainbow scale and obtained with                    two-material decomposition in the projection space using iodine and water as                    basis materials, shows perfusion defects in the lungs (arrows). (c) Water                    density map at the same level with color-coded values. Note that the algorithm                    does not account for calcium present in the bones (*); therefore, calcium                    is forcedly described as a mixture of iodine and water. (d) Conventional chest                    CT image at the same level shows bilateral pulmonary emboli (arrows) in the                    right and left pulmonary arteries.

Figure 3a. Two-material decomposition. (a) Graph shows the process of two-material decomposition. The entire imaged volume is assumed to be made of only two (pair) preselected materials in different proportions. The pixel composition can be described by a linear combination of equivalent amounts of the two materials (eg, α and β) with known attenuation profiles at the respective levels, essentially α cos(φ) + β sin(φ). (b) Iodine density map through the midchest, color coded in rainbow scale and obtained with two-material decomposition in the projection space using iodine and water as basis materials, shows perfusion defects in the lungs (arrows). (c) Water density map at the same level with color-coded values. Note that the algorithm does not account for calcium present in the bones (*); therefore, calcium is forcedly described as a mixture of iodine and water. (d) Conventional chest CT image at the same level shows bilateral pulmonary emboli (arrows) in the right and left pulmonary arteries.

Two-material decomposition. (a) Graph shows the process of two-material                    decomposition. The entire imaged volume is assumed to be made of only two (pair)                    preselected materials in different proportions. The pixel composition can be                    described by a linear combination of equivalent amounts of the two materials                    (eg, α and β) with known attenuation profiles at the respective                    levels, essentially α cos(φ) + β sin(φ). (b) Iodine                    density map through the midchest, color coded in rainbow scale and obtained with                    two-material decomposition in the projection space using iodine and water as                    basis materials, shows perfusion defects in the lungs (arrows). (c) Water                    density map at the same level with color-coded values. Note that the algorithm                    does not account for calcium present in the bones (*); therefore, calcium                    is forcedly described as a mixture of iodine and water. (d) Conventional chest                    CT image at the same level shows bilateral pulmonary emboli (arrows) in the                    right and left pulmonary arteries.

Figure 3b. Two-material decomposition. (a) Graph shows the process of two-material decomposition. The entire imaged volume is assumed to be made of only two (pair) preselected materials in different proportions. The pixel composition can be described by a linear combination of equivalent amounts of the two materials (eg, α and β) with known attenuation profiles at the respective levels, essentially α cos(φ) + β sin(φ). (b) Iodine density map through the midchest, color coded in rainbow scale and obtained with two-material decomposition in the projection space using iodine and water as basis materials, shows perfusion defects in the lungs (arrows). (c) Water density map at the same level with color-coded values. Note that the algorithm does not account for calcium present in the bones (*); therefore, calcium is forcedly described as a mixture of iodine and water. (d) Conventional chest CT image at the same level shows bilateral pulmonary emboli (arrows) in the right and left pulmonary arteries.

Two-material decomposition. (a) Graph shows the process of two-material                    decomposition. The entire imaged volume is assumed to be made of only two (pair)                    preselected materials in different proportions. The pixel composition can be                    described by a linear combination of equivalent amounts of the two materials                    (eg, α and β) with known attenuation profiles at the respective                    levels, essentially α cos(φ) + β sin(φ). (b) Iodine                    density map through the midchest, color coded in rainbow scale and obtained with                    two-material decomposition in the projection space using iodine and water as                    basis materials, shows perfusion defects in the lungs (arrows). (c) Water                    density map at the same level with color-coded values. Note that the algorithm                    does not account for calcium present in the bones (*); therefore, calcium                    is forcedly described as a mixture of iodine and water. (d) Conventional chest                    CT image at the same level shows bilateral pulmonary emboli (arrows) in the                    right and left pulmonary arteries.

Figure 3c. Two-material decomposition. (a) Graph shows the process of two-material decomposition. The entire imaged volume is assumed to be made of only two (pair) preselected materials in different proportions. The pixel composition can be described by a linear combination of equivalent amounts of the two materials (eg, α and β) with known attenuation profiles at the respective levels, essentially α cos(φ) + β sin(φ). (b) Iodine density map through the midchest, color coded in rainbow scale and obtained with two-material decomposition in the projection space using iodine and water as basis materials, shows perfusion defects in the lungs (arrows). (c) Water density map at the same level with color-coded values. Note that the algorithm does not account for calcium present in the bones (*); therefore, calcium is forcedly described as a mixture of iodine and water. (d) Conventional chest CT image at the same level shows bilateral pulmonary emboli (arrows) in the right and left pulmonary arteries.

Two-material decomposition. (a) Graph shows the process of two-material                    decomposition. The entire imaged volume is assumed to be made of only two (pair)                    preselected materials in different proportions. The pixel composition can be                    described by a linear combination of equivalent amounts of the two materials                    (eg, α and β) with known attenuation profiles at the respective                    levels, essentially α cos(φ) + β sin(φ). (b) Iodine                    density map through the midchest, color coded in rainbow scale and obtained with                    two-material decomposition in the projection space using iodine and water as                    basis materials, shows perfusion defects in the lungs (arrows). (c) Water                    density map at the same level with color-coded values. Note that the algorithm                    does not account for calcium present in the bones (*); therefore, calcium                    is forcedly described as a mixture of iodine and water. (d) Conventional chest                    CT image at the same level shows bilateral pulmonary emboli (arrows) in the                    right and left pulmonary arteries.

Figure 3d. Two-material decomposition. (a) Graph shows the process of two-material decomposition. The entire imaged volume is assumed to be made of only two (pair) preselected materials in different proportions. The pixel composition can be described by a linear combination of equivalent amounts of the two materials (eg, α and β) with known attenuation profiles at the respective levels, essentially α cos(φ) + β sin(φ). (b) Iodine density map through the midchest, color coded in rainbow scale and obtained with two-material decomposition in the projection space using iodine and water as basis materials, shows perfusion defects in the lungs (arrows). (c) Water density map at the same level with color-coded values. Note that the algorithm does not account for calcium present in the bones (*); therefore, calcium is forcedly described as a mixture of iodine and water. (d) Conventional chest CT image at the same level shows bilateral pulmonary emboli (arrows) in the right and left pulmonary arteries.

A three-material decomposition algorithm is commonly used in the dual-source scanner (36). This algorithm assumes that each voxel can be represented with three materials with known elemental compositions (Fig 4). To obtain more than two material–density images, other constraints have to be introduced. Liu et al (37) introduced an image domain–based method in which mass conservation was used as the additional physical constraint, where the sum of individual mass fractions of the three materials should add up to 1 (37).

Three-material decomposition. (a) Graph shows the process of                    three-material decomposition. The CT numbers of three basis materials at low and                    high energies are plotted along the y and x axes, respectively. Any unknown                    material can be mapped into this plot, and the percentage composition of each of                    the three basis materials can be calculated. (b–d) Triple-material                    decomposition applied in the image space in a patient with chronic                    thromboembolic pulmonary hypertension, using the volume conservation constraint                    and iodine, soft tissue, and air as basis materials. (b) Color-coded iodine                    image shows segmental decreased perfusion (arrows) in the lung bases. (c)                    Soft-tissue image shows homogeneity of the soft-tissue fraction distribution,                    which is consistent with the absence of parenchymal abnormality on anatomic                    images. (d) Air image shows lack of matched heterogeneity in lung aeration                    within the lung bases to support ventilation abnormalities. This finding was                    consistent with spirometry and ventilation–perfusion imaging                    findings.

Figure 4a. Three-material decomposition. (a) Graph shows the process of three-material decomposition. The CT numbers of three basis materials at low and high energies are plotted along the y and x axes, respectively. Any unknown material can be mapped into this plot, and the percentage composition of each of the three basis materials can be calculated. (b–d) Triple-material decomposition applied in the image space in a patient with chronic thromboembolic pulmonary hypertension, using the volume conservation constraint and iodine, soft tissue, and air as basis materials. (b) Color-coded iodine image shows segmental decreased perfusion (arrows) in the lung bases. (c) Soft-tissue image shows homogeneity of the soft-tissue fraction distribution, which is consistent with the absence of parenchymal abnormality on anatomic images. (d) Air image shows lack of matched heterogeneity in lung aeration within the lung bases to support ventilation abnormalities. This finding was consistent with spirometry and ventilation–perfusion imaging findings.

Three-material decomposition. (a) Graph shows the process of                    three-material decomposition. The CT numbers of three basis materials at low and                    high energies are plotted along the y and x axes, respectively. Any unknown                    material can be mapped into this plot, and the percentage composition of each of                    the three basis materials can be calculated. (b–d) Triple-material                    decomposition applied in the image space in a patient with chronic                    thromboembolic pulmonary hypertension, using the volume conservation constraint                    and iodine, soft tissue, and air as basis materials. (b) Color-coded iodine                    image shows segmental decreased perfusion (arrows) in the lung bases. (c)                    Soft-tissue image shows homogeneity of the soft-tissue fraction distribution,                    which is consistent with the absence of parenchymal abnormality on anatomic                    images. (d) Air image shows lack of matched heterogeneity in lung aeration                    within the lung bases to support ventilation abnormalities. This finding was                    consistent with spirometry and ventilation–perfusion imaging                    findings.

Figure 4b. Three-material decomposition. (a) Graph shows the process of three-material decomposition. The CT numbers of three basis materials at low and high energies are plotted along the y and x axes, respectively. Any unknown material can be mapped into this plot, and the percentage composition of each of the three basis materials can be calculated. (b–d) Triple-material decomposition applied in the image space in a patient with chronic thromboembolic pulmonary hypertension, using the volume conservation constraint and iodine, soft tissue, and air as basis materials. (b) Color-coded iodine image shows segmental decreased perfusion (arrows) in the lung bases. (c) Soft-tissue image shows homogeneity of the soft-tissue fraction distribution, which is consistent with the absence of parenchymal abnormality on anatomic images. (d) Air image shows lack of matched heterogeneity in lung aeration within the lung bases to support ventilation abnormalities. This finding was consistent with spirometry and ventilation–perfusion imaging findings.

Three-material decomposition. (a) Graph shows the process of                    three-material decomposition. The CT numbers of three basis materials at low and                    high energies are plotted along the y and x axes, respectively. Any unknown                    material can be mapped into this plot, and the percentage composition of each of                    the three basis materials can be calculated. (b–d) Triple-material                    decomposition applied in the image space in a patient with chronic                    thromboembolic pulmonary hypertension, using the volume conservation constraint                    and iodine, soft tissue, and air as basis materials. (b) Color-coded iodine                    image shows segmental decreased perfusion (arrows) in the lung bases. (c)                    Soft-tissue image shows homogeneity of the soft-tissue fraction distribution,                    which is consistent with the absence of parenchymal abnormality on anatomic                    images. (d) Air image shows lack of matched heterogeneity in lung aeration                    within the lung bases to support ventilation abnormalities. This finding was                    consistent with spirometry and ventilation–perfusion imaging                    findings.

Figure 4c. Three-material decomposition. (a) Graph shows the process of three-material decomposition. The CT numbers of three basis materials at low and high energies are plotted along the y and x axes, respectively. Any unknown material can be mapped into this plot, and the percentage composition of each of the three basis materials can be calculated. (b–d) Triple-material decomposition applied in the image space in a patient with chronic thromboembolic pulmonary hypertension, using the volume conservation constraint and iodine, soft tissue, and air as basis materials. (b) Color-coded iodine image shows segmental decreased perfusion (arrows) in the lung bases. (c) Soft-tissue image shows homogeneity of the soft-tissue fraction distribution, which is consistent with the absence of parenchymal abnormality on anatomic images. (d) Air image shows lack of matched heterogeneity in lung aeration within the lung bases to support ventilation abnormalities. This finding was consistent with spirometry and ventilation–perfusion imaging findings.

Three-material decomposition. (a) Graph shows the process of                    three-material decomposition. The CT numbers of three basis materials at low and                    high energies are plotted along the y and x axes, respectively. Any unknown                    material can be mapped into this plot, and the percentage composition of each of                    the three basis materials can be calculated. (b–d) Triple-material                    decomposition applied in the image space in a patient with chronic                    thromboembolic pulmonary hypertension, using the volume conservation constraint                    and iodine, soft tissue, and air as basis materials. (b) Color-coded iodine                    image shows segmental decreased perfusion (arrows) in the lung bases. (c)                    Soft-tissue image shows homogeneity of the soft-tissue fraction distribution,                    which is consistent with the absence of parenchymal abnormality on anatomic                    images. (d) Air image shows lack of matched heterogeneity in lung aeration                    within the lung bases to support ventilation abnormalities. This finding was                    consistent with spirometry and ventilation–perfusion imaging                    findings.

Figure 4d. Three-material decomposition. (a) Graph shows the process of three-material decomposition. The CT numbers of three basis materials at low and high energies are plotted along the y and x axes, respectively. Any unknown material can be mapped into this plot, and the percentage composition of each of the three basis materials can be calculated. (b–d) Triple-material decomposition applied in the image space in a patient with chronic thromboembolic pulmonary hypertension, using the volume conservation constraint and iodine, soft tissue, and air as basis materials. (b) Color-coded iodine image shows segmental decreased perfusion (arrows) in the lung bases. (c) Soft-tissue image shows homogeneity of the soft-tissue fraction distribution, which is consistent with the absence of parenchymal abnormality on anatomic images. (d) Air image shows lack of matched heterogeneity in lung aeration within the lung bases to support ventilation abnormalities. This finding was consistent with spirometry and ventilation–perfusion imaging findings.

CT numbers from the low- and high-energy images of three basis materials with known composition and density are plotted in a two-dimensional space (eg, high-energy CT numbers along the x-axis, low-energy CT numbers along the y-axis) to define two vectors. Three materials that are widely separated will form a triangle. Then, the CT numbers of the pixel at high and low energies can be mapped in the calibration diagram (Fig 4a). The presence of a material or its concentration is calculated by the position in this diagram (4,10,36).

Mendonca et al (38) proposed a multimaterial decomposition algorithm for obtaining three or more material density images. This is based on the assumption that material mixes in the human body behave as an ideal solution, where the principle of volume conservation is applicable (ie, the volume fractions of the three materials are equal to 1). Given the system with two inputs and three unknowns, this assumption ensures that the equations are solvable.

For more than three materials, the acquired dual-energy CT decomposition data are compared with a material triplet library to find the best fit to explain the acquired data. Then, it is assumed that only those three materials are present in the data, and the algorithm will provide percentages of the materials in the triplet library. Any material outside the selected triplet material will be shown as 0% (38).

Given the inherent ability of PCD CT systems to simultaneously measure more than two energy levels (usually four or five energy bins), multimaterial decomposition becomes facilitated, in both the projection and image domains (39). Images can be generated from each energy bin, with the number of bins dependent on the energy thresholds allowed by the circuit design. Energy weighting involves assigning a higher weight to a particular energy bin, depending on the clinical task and the material that needs to be highlighted. K-edge imaging is highly beneficial in imaging novel contrast media and using multiple contrast media in the same acquisition (Fig 5) (7,8,24).

K-edge imaging with PCD CT. A phantom study was performed using vials with                    different contrast media, including gadolinium (Gd), iodine (I), bismuth (Bi),                    and two mixtures (M1 and M2). (a) Threshold low image from PCD CT, which is                    equivalent to standard single-energy CT, does not show any significant                    difference between the contrast media. (b) Color image of the contrast media                    overlaid on standard gray-scale image shows different colors representing                    different k edges of the different contrast agents. Gadolinium is coded green,                    iodine red, and bismuth blue. The mixtures show up as different colors, with the                    yellow-coded one being a mix of green and red (gadolinium and iodine), whereas                    the other one is white owing to mixture of green, red, and blue (gadolinium,                    iodine, and bismuth).

Figure 5a. K-edge imaging with PCD CT. A phantom study was performed using vials with different contrast media, including gadolinium (Gd), iodine (I), bismuth (Bi), and two mixtures (M1 and M2). (a) Threshold low image from PCD CT, which is equivalent to standard single-energy CT, does not show any significant difference between the contrast media. (b) Color image of the contrast media overlaid on standard gray-scale image shows different colors representing different k edges of the different contrast agents. Gadolinium is coded green, iodine red, and bismuth blue. The mixtures show up as different colors, with the yellow-coded one being a mix of green and red (gadolinium and iodine), whereas the other one is white owing to mixture of green, red, and blue (gadolinium, iodine, and bismuth).

K-edge imaging with PCD CT. A phantom study was performed using vials with                    different contrast media, including gadolinium (Gd), iodine (I), bismuth (Bi),                    and two mixtures (M1 and M2). (a) Threshold low image from PCD CT, which is                    equivalent to standard single-energy CT, does not show any significant                    difference between the contrast media. (b) Color image of the contrast media                    overlaid on standard gray-scale image shows different colors representing                    different k edges of the different contrast agents. Gadolinium is coded green,                    iodine red, and bismuth blue. The mixtures show up as different colors, with the                    yellow-coded one being a mix of green and red (gadolinium and iodine), whereas                    the other one is white owing to mixture of green, red, and blue (gadolinium,                    iodine, and bismuth).

Figure 5b. K-edge imaging with PCD CT. A phantom study was performed using vials with different contrast media, including gadolinium (Gd), iodine (I), bismuth (Bi), and two mixtures (M1 and M2). (a) Threshold low image from PCD CT, which is equivalent to standard single-energy CT, does not show any significant difference between the contrast media. (b) Color image of the contrast media overlaid on standard gray-scale image shows different colors representing different k edges of the different contrast agents. Gadolinium is coded green, iodine red, and bismuth blue. The mixtures show up as different colors, with the yellow-coded one being a mix of green and red (gadolinium and iodine), whereas the other one is white owing to mixture of green, red, and blue (gadolinium, iodine, and bismuth).

Noise is a challenge in material decomposition. The spectral resolution is improved by using a narrow energy bin. However, this introduces noise owing to the presence of fewer photons in a narrow bin, which in turn decreases the accuracy of material decomposition. The same challenge exists for using more energy bins, such as in PCD CT. A higher radiation dose may be needed if similar noise in all energy bins is desired (40).

Energy-domain noise-reduction algorithms allow use of narrow and numerous bins without higher noise or radiation dose (40). Other methods of decreasing noise include the correlated noise-reduction technique (41), measurement-dependent filtering (42), negative correlation (30), and compressed sensing methods (43,44). Also, improved spectral separation from higher-energy bins reduces image noise (45). For mixed images that use all the photons, the noise is not affected.

Image Types from a Multienergy CT Scanner

Unlike conventional CT, multienergy CT generates numerous image sets. A summary of these images is provided in Table E2.

Routine Diagnostic Images

With all multienergy CT scanners, a set of images is sent to the picture archiving and communication system (PACS) for routine diagnostic purposes, analogous to a conventional single-energy CT scanner. In dual-source CT, mixed or combined images—which are obtained by linear blending of the low- and high-energy images—are used for this purpose. These are typically a combination of 50%–60% low-energy images and 50%–40% high-energy images, providing a good balance between contrast and noise at dose efficiency comparable to that of single-energy CT (46). In dual-layer CT, good-quality routine diagnostic images are generated with combined data from the top and bottom layers (47). In the rapid kVp switching technology, VMIs at 70 keV—which are considered as PACS equivalent to 120-kVp images—are used for diagnostic purposes (48).

Virtual Monoenergetic Images

VMIs are reconstructed gray-scale images that simulate the appearance of CT images that could theoretically be acquired by scanning a patient with a true monoenergetic x-ray source (49). VMIs are obtained by linear combination of basis pair images at different proportions and are denoted by kilo–electron volts (keV). Multienergy CT scanners are capable of generating VMIs from 35 to 200 keV. VMIs at 70 keV are considered the equivalent of 120-kVp conventional images, with similar attenuation values but with lower artifacts and noise (50).

Teaching Point Low-energy VMIs (<70 keV) accentuate the attenuation of iodine owing to their proximity to the k edge of iodine (33 keV),
with the highest attenuation seen at 40 keV (Fig 6). Hence, low-energy VMIs are mainly used in optimizing the vascular contrast in CT angiography studies (51,52). They are particularly beneficial for salvaging CT angiography studies with suboptimal vascular contrast and performing low contrast material dose CT angiography in patients with borderline renal dysfunction (49,53).

Low-energy VMIs. VMIs at the level of the main pulmonary artery show                        progressive increase in the attenuation of iodine as the energy level                        decreases from 90 keV to 40 keV. The CT number of the contrast                        material–opacified main pulmonary artery is shown in yellow. The                        highest attenuation is seen at 40 keV, demonstrating the utility of                        low-energy VMIs in contrast optimization.

Figure 6. Low-energy VMIs. VMIs at the level of the main pulmonary artery show progressive increase in the attenuation of iodine as the energy level decreases from 90 keV to 40 keV. The CT number of the contrast material–opacified main pulmonary artery is shown in yellow. The highest attenuation is seen at 40 keV, demonstrating the utility of low-energy VMIs in contrast optimization.

In addition, CT angiography–quality images can be obtained from routine CT, and small branches can be visualized (51,54). Low-energy VMIs can also be used to decrease the iodine load in CT arthrography (55). Owing to lower photons, noise is a challenge of low-energy VMIs, but this has been minimized using noise-reduction algorithms including energy-domain noise-reduction techniques (5658).

Low-energy VMIs can also be used to improve the conspicuity of hypervascular lesions, which appear brighter on these images (Fig 7). They also increase the conspicuity of iso- to mildly hypoattenuating lesions that do not alter the contour of the organ, owing to increased contrast at low energies between the enhancing parenchyma and nonenhancing lesion. The ideal energy of VMIs varies with the organ, such as 50 keV for the pancreas and 40 keV for isoattenuating gallstones (48,59). Low-energy VMIs, with or without color, allow evaluation of the mural integrity of the bowel, appendix, and gallbladder (60).

Improved lesion visualization with low-energy VMIs. (a) Axial                        conventional 120-kVp CT image shows vague lesions in the liver (arrows). (b)                        Axial 40-keV VMI at the same level shows significantly improved                        visualization of the hypervascular lesions (arrows).

Figure 7a. Improved lesion visualization with low-energy VMIs. (a) Axial conventional 120-kVp CT image shows vague lesions in the liver (arrows). (b) Axial 40-keV VMI at the same level shows significantly improved visualization of the hypervascular lesions (arrows).

Improved lesion visualization with low-energy VMIs. (a) Axial                        conventional 120-kVp CT image shows vague lesions in the liver (arrows). (b)                        Axial 40-keV VMI at the same level shows significantly improved                        visualization of the hypervascular lesions (arrows).

Figure 7b. Improved lesion visualization with low-energy VMIs. (a) Axial conventional 120-kVp CT image shows vague lesions in the liver (arrows). (b) Axial 40-keV VMI at the same level shows significantly improved visualization of the hypervascular lesions (arrows).

High-energy VMIs (>90 keV) reduce artifacts such as beam-hardening and metallic artifacts. Beam-hardening artifacts due to preferential attenuation of low-energy photons can be minimized with high-energy VMIs, especially in areas of dense contrast material, as in the axillary and subclavian veins and also adjacent to the myocardium for perfusion scans (61). Calcium blooming, which results in overestimation of luminal stenosis, can also be reduced (62).

Metallic artifacts from prostheses can be reduced with improved visualization of metallic prostheses and periprosthetic tissues without increasing radiation dose (55,63). The ideal energy for reducing such artifacts ranges from 108 to 149 keV, depending on the scanner, prosthesis, disease, and reader preference (55) (Fig 8). Note that reduction of metallic artifact depends on the atomic number and size of the metal.

Decreased artifact with high-energy VMIs. Axial CT images in a patient                        with extensive metallic artifacts show progressive decrease in artifacts                        (*, arrows in a) as the x-ray energy increases from 65 keV (a) to 100                        keV (b) to 150 keV (c).

Figure 8a. Decreased artifact with high-energy VMIs. Axial CT images in a patient with extensive metallic artifacts show progressive decrease in artifacts (*, arrows in a) as the x-ray energy increases from 65 keV (a) to 100 keV (b) to 150 keV (c).

Decreased artifact with high-energy VMIs. Axial CT images in a patient                        with extensive metallic artifacts show progressive decrease in artifacts                        (*, arrows in a) as the x-ray energy increases from 65 keV (a) to 100                        keV (b) to 150 keV (c).

Figure 8b. Decreased artifact with high-energy VMIs. Axial CT images in a patient with extensive metallic artifacts show progressive decrease in artifacts (*, arrows in a) as the x-ray energy increases from 65 keV (a) to 100 keV (b) to 150 keV (c).

Decreased artifact with high-energy VMIs. Axial CT images in a patient                        with extensive metallic artifacts show progressive decrease in artifacts                        (*, arrows in a) as the x-ray energy increases from 65 keV (a) to 100                        keV (b) to 150 keV (c).

Figure 8c. Decreased artifact with high-energy VMIs. Axial CT images in a patient with extensive metallic artifacts show progressive decrease in artifacts (*, arrows in a) as the x-ray energy increases from 65 keV (a) to 100 keV (b) to 150 keV (c).

Very high-energy VMIs (>140 keV) can mimic virtual noncontrast (VNC) images owing to absence of contrast material. These have been used to diagnose stones and characterize adrenal adenomas (64). These can also be used to remove contrast material from arthrography and for three-dimensional modeling, which is important for surgical planning (65). VMIs have been used in proton therapy to estimate the stopping power ratio (SPR), using empirical parameterization and calculating the water-equivalent path lengths from SPR maps (66).

Spectral attenuation curves plot the CT numbers of tissues at different energy levels of VMIs. Tissues or materials can be characterized on the basis of typical curves. At lower energies, the attenuation is higher for all the materials. Since CT number = (attenuationmaterial − attenuationwater)/attenuationwater × 1000, at lower energies, both attenuationmaterial and attenuationwater increase. For some materials, such as iodine, the increase in attenuationmaterial is more than the increase in attenuationwater; therefore, the CT number increases at lower energy. However, for other materials, such as fat, the increase in attenuationmaterial is less than that in attenuationwater; therefore, the CT number decreases (6769).

For example, in distinguishing hemorrhage from contrast enhancement in the brain, iodine shows a significant increase in attenuation at low energies, whereas the attenuation of hemorrhage is almost flat at all energies (70). In subtle hypoattenuating lesions at CT, an almost flat curve indicates the presence of water, whereas low numbers at low energies indicate the presence of fat (Fig 9).

Spectral attenuation curve. (a) Axial CT image in a patient with lower                        abdominal pain shows a hypoattenuating lesion (arrow) in the right adnexa.                        The attenuation value was −5 HU, indeterminate between a cyst and a                        fat-containing lesion. (b) Spectral attenuation curve from a region of                        interest (ROI) in the lesion shows that the CT number progressively                        decreases with the energy, indicating that it contains fat. The lesion was                        consistent with a dermoid cyst.

Figure 9a. Spectral attenuation curve. (a) Axial CT image in a patient with lower abdominal pain shows a hypoattenuating lesion (arrow) in the right adnexa. The attenuation value was −5 HU, indeterminate between a cyst and a fat-containing lesion. (b) Spectral attenuation curve from a region of interest (ROI) in the lesion shows that the CT number progressively decreases with the energy, indicating that it contains fat. The lesion was consistent with a dermoid cyst.

Spectral attenuation curve. (a) Axial CT image in a patient with lower                        abdominal pain shows a hypoattenuating lesion (arrow) in the right adnexa.                        The attenuation value was −5 HU, indeterminate between a cyst and a                        fat-containing lesion. (b) Spectral attenuation curve from a region of                        interest (ROI) in the lesion shows that the CT number progressively                        decreases with the energy, indicating that it contains fat. The lesion was                        consistent with a dermoid cyst.

Figure 9b. Spectral attenuation curve. (a) Axial CT image in a patient with lower abdominal pain shows a hypoattenuating lesion (arrow) in the right adnexa. The attenuation value was −5 HU, indeterminate between a cyst and a fat-containing lesion. (b) Spectral attenuation curve from a region of interest (ROI) in the lesion shows that the CT number progressively decreases with the energy, indicating that it contains fat. The lesion was consistent with a dermoid cyst.

Material Composition Images

In the following sections, we highlight the commonly used material composition images and their uses.

Iodine-only Images.—Iodine-only images (iodine maps) are multienergy CT images in which pixels containing iodine are highlighted. These can be generated with two-material decomposition using iodine and water as basis pairs. The substantially different attenuation properties at different energies—due to the k edge of iodine (33.2 keV)—allow good separation between iodine and water. These images can be displayed as gray-scale images that allow quantitation or as color overlays on top of anatomic images, which can be only for qualitative evaluation (Fig 10) (71). Measurements of iodine on iodine maps correlate well with iodine concentration in phantom studies (7274), making them good surrogates for tissue perfusion in organs and lesions (72).

Iodine decomposition. (a) Graph shows three-material decomposition of                        iodine using dual-source CT technology. A base material line is defined                        using the attenuation values of water and soft tissue at low and high                        energies. Addition of iodine to this base material line results in increased                        attenuation value with a characteristic slope of iodine. The attenuation of                        the voxel can now be decomposed into attenuation from iodine and residual                        virtual noncontrast (VNC) attenuation. (b) Axial iodine image through the                        midchest from a projection space–based material decomposition                        algorithm for two materials (iodine and water). (c) Iodine map at the same                        level from multimaterial decomposition using calcium as an additional basis                        material. Note the subtraction of calcium from the cortical bone. (d) Iodine                        map in c displayed in color and overlaid on a standard CT reconstruction in                        gray scale. The iodine map was windowed to display the lung perfusion                        heterogeneity in this patient with pulmonary hypertension.

Figure 10a. Iodine decomposition. (a) Graph shows three-material decomposition of iodine using dual-source CT technology. A base material line is defined using the attenuation values of water and soft tissue at low and high energies. Addition of iodine to this base material line results in increased attenuation value with a characteristic slope of iodine. The attenuation of the voxel can now be decomposed into attenuation from iodine and residual virtual noncontrast (VNC) attenuation. (b) Axial iodine image through the midchest from a projection space–based material decomposition algorithm for two materials (iodine and water). (c) Iodine map at the same level from multimaterial decomposition using calcium as an additional basis material. Note the subtraction of calcium from the cortical bone. (d) Iodine map in c displayed in color and overlaid on a standard CT reconstruction in gray scale. The iodine map was windowed to display the lung perfusion heterogeneity in this patient with pulmonary hypertension.

Iodine decomposition. (a) Graph shows three-material decomposition of                        iodine using dual-source CT technology. A base material line is defined                        using the attenuation values of water and soft tissue at low and high                        energies. Addition of iodine to this base material line results in increased                        attenuation value with a characteristic slope of iodine. The attenuation of                        the voxel can now be decomposed into attenuation from iodine and residual                        virtual noncontrast (VNC) attenuation. (b) Axial iodine image through the                        midchest from a projection space–based material decomposition                        algorithm for two materials (iodine and water). (c) Iodine map at the same                        level from multimaterial decomposition using calcium as an additional basis                        material. Note the subtraction of calcium from the cortical bone. (d) Iodine                        map in c displayed in color and overlaid on a standard CT reconstruction in                        gray scale. The iodine map was windowed to display the lung perfusion                        heterogeneity in this patient with pulmonary hypertension.

Figure 10b. Iodine decomposition. (a) Graph shows three-material decomposition of iodine using dual-source CT technology. A base material line is defined using the attenuation values of water and soft tissue at low and high energies. Addition of iodine to this base material line results in increased attenuation value with a characteristic slope of iodine. The attenuation of the voxel can now be decomposed into attenuation from iodine and residual virtual noncontrast (VNC) attenuation. (b) Axial iodine image through the midchest from a projection space–based material decomposition algorithm for two materials (iodine and water). (c) Iodine map at the same level from multimaterial decomposition using calcium as an additional basis material. Note the subtraction of calcium from the cortical bone. (d) Iodine map in c displayed in color and overlaid on a standard CT reconstruction in gray scale. The iodine map was windowed to display the lung perfusion heterogeneity in this patient with pulmonary hypertension.

Iodine decomposition. (a) Graph shows three-material decomposition of                        iodine using dual-source CT technology. A base material line is defined                        using the attenuation values of water and soft tissue at low and high                        energies. Addition of iodine to this base material line results in increased                        attenuation value with a characteristic slope of iodine. The attenuation of                        the voxel can now be decomposed into attenuation from iodine and residual                        virtual noncontrast (VNC) attenuation. (b) Axial iodine image through the                        midchest from a projection space–based material decomposition                        algorithm for two materials (iodine and water). (c) Iodine map at the same                        level from multimaterial decomposition using calcium as an additional basis                        material. Note the subtraction of calcium from the cortical bone. (d) Iodine                        map in c displayed in color and overlaid on a standard CT reconstruction in                        gray scale. The iodine map was windowed to display the lung perfusion                        heterogeneity in this patient with pulmonary hypertension.

Figure 10c. Iodine decomposition. (a) Graph shows three-material decomposition of iodine using dual-source CT technology. A base material line is defined using the attenuation values of water and soft tissue at low and high energies. Addition of iodine to this base material line results in increased attenuation value with a characteristic slope of iodine. The attenuation of the voxel can now be decomposed into attenuation from iodine and residual virtual noncontrast (VNC) attenuation. (b) Axial iodine image through the midchest from a projection space–based material decomposition algorithm for two materials (iodine and water). (c) Iodine map at the same level from multimaterial decomposition using calcium as an additional basis material. Note the subtraction of calcium from the cortical bone. (d) Iodine map in c displayed in color and overlaid on a standard CT reconstruction in gray scale. The iodine map was windowed to display the lung perfusion heterogeneity in this patient with pulmonary hypertension.

Iodine decomposition. (a) Graph shows three-material decomposition of                        iodine using dual-source CT technology. A base material line is defined                        using the attenuation values of water and soft tissue at low and high                        energies. Addition of iodine to this base material line results in increased                        attenuation value with a characteristic slope of iodine. The attenuation of                        the voxel can now be decomposed into attenuation from iodine and residual                        virtual noncontrast (VNC) attenuation. (b) Axial iodine image through the                        midchest from a projection space–based material decomposition                        algorithm for two materials (iodine and water). (c) Iodine map at the same                        level from multimaterial decomposition using calcium as an additional basis                        material. Note the subtraction of calcium from the cortical bone. (d) Iodine                        map in c displayed in color and overlaid on a standard CT reconstruction in                        gray scale. The iodine map was windowed to display the lung perfusion                        heterogeneity in this patient with pulmonary hypertension.

Figure 10d. Iodine decomposition. (a) Graph shows three-material decomposition of iodine using dual-source CT technology. A base material line is defined using the attenuation values of water and soft tissue at low and high energies. Addition of iodine to this base material line results in increased attenuation value with a characteristic slope of iodine. The attenuation of the voxel can now be decomposed into attenuation from iodine and residual virtual noncontrast (VNC) attenuation. (b) Axial iodine image through the midchest from a projection space–based material decomposition algorithm for two materials (iodine and water). (c) Iodine map at the same level from multimaterial decomposition using calcium as an additional basis material. Note the subtraction of calcium from the cortical bone. (d) Iodine map in c displayed in color and overlaid on a standard CT reconstruction in gray scale. The iodine map was windowed to display the lung perfusion heterogeneity in this patient with pulmonary hypertension.

A dedicated algorithm for lung perfusion is available from one vendor (Siemens Healthineers, Erlangen, Germany), called the pulmonary blood volume (PBV), which is a color-coded map of normalized iodine density restricted to only the lungs using a threshold-based algorithm between −960 and −600 HU (Fig 11). These images are useful in diagnosis and risk stratification of pulmonary embolism and pulmonary hypertension (75,76). Iodine-only images of the heart acquired both at stress and at rest allow evaluation of myocardial ischemia, enabling assessment of the hemodynamic significance of stenosis detected at CT angiography (Fig 12) (54). Iodine maps can also be used to evaluate delayed myocardial enhancement as an alternative to MRI (Fig 13) or to quantify the extracellular volume (77).

Pulmonary perfused blood volume at multienergy CT. (a) Axial CT image                        shows a questionable filling defect (arrow) in a subsegmental branch of the                        pulmonary artery in the right lower lobe, concerning for thromboembolism.                        (b) Coronal pulmonary perfused blood volume (PBV) map overlaid on a standard                        CT image shows a segmental perfusion defect (arrow) in the right lower lobe                        that corresponds to the territory supplied by the affected artery,                        increasing the sensitivity and diagnostic certainty.

Figure 11a. Pulmonary perfused blood volume at multienergy CT. (a) Axial CT image shows a questionable filling defect (arrow) in a subsegmental branch of the pulmonary artery in the right lower lobe, concerning for thromboembolism. (b) Coronal pulmonary perfused blood volume (PBV) map overlaid on a standard CT image shows a segmental perfusion defect (arrow) in the right lower lobe that corresponds to the territory supplied by the affected artery, increasing the sensitivity and diagnostic certainty.

Pulmonary perfused blood volume at multienergy CT. (a) Axial CT image                        shows a questionable filling defect (arrow) in a subsegmental branch of the                        pulmonary artery in the right lower lobe, concerning for thromboembolism.                        (b) Coronal pulmonary perfused blood volume (PBV) map overlaid on a standard                        CT image shows a segmental perfusion defect (arrow) in the right lower lobe                        that corresponds to the territory supplied by the affected artery,                        increasing the sensitivity and diagnostic certainty.

Figure 11b. Pulmonary perfused blood volume at multienergy CT. (a) Axial CT image shows a questionable filling defect (arrow) in a subsegmental branch of the pulmonary artery in the right lower lobe, concerning for thromboembolism. (b) Coronal pulmonary perfused blood volume (PBV) map overlaid on a standard CT image shows a segmental perfusion defect (arrow) in the right lower lobe that corresponds to the territory supplied by the affected artery, increasing the sensitivity and diagnostic certainty.

Myocardial perfusion assessed with multienergy CT in a 68-year-old man                        with worsening exertional chest pain. (a) Curved multiplanar reconstruction                        from coronary CT angiography shows multiple calcified plaques along the                        right coronary artery (RCA) with severe (≥70%) ostial stenosis                        (arrowhead). (b) Colored map overlaid on a standard two-chamber reformation                        shows the iodine attenuation obtained with multienergy CT. Note the                        decreased perfusion at the basal inferior left ventricular wall (arrowhead).                        The significant stenosis at the RCA ostium was confirmed with conventional                        angiography (not shown).

Figure 12a. Myocardial perfusion assessed with multienergy CT in a 68-year-old man with worsening exertional chest pain. (a) Curved multiplanar reconstruction from coronary CT angiography shows multiple calcified plaques along the right coronary artery (RCA) with severe (≥70%) ostial stenosis (arrowhead). (b) Colored map overlaid on a standard two-chamber reformation shows the iodine attenuation obtained with multienergy CT. Note the decreased perfusion at the basal inferior left ventricular wall (arrowhead). The significant stenosis at the RCA ostium was confirmed with conventional angiography (not shown).

Myocardial perfusion assessed with multienergy CT in a 68-year-old man                        with worsening exertional chest pain. (a) Curved multiplanar reconstruction                        from coronary CT angiography shows multiple calcified plaques along the                        right coronary artery (RCA) with severe (≥70%) ostial stenosis                        (arrowhead). (b) Colored map overlaid on a standard two-chamber reformation                        shows the iodine attenuation obtained with multienergy CT. Note the                        decreased perfusion at the basal inferior left ventricular wall (arrowhead).                        The significant stenosis at the RCA ostium was confirmed with conventional                        angiography (not shown).

Figure 12b. Myocardial perfusion assessed with multienergy CT in a 68-year-old man with worsening exertional chest pain. (a) Curved multiplanar reconstruction from coronary CT angiography shows multiple calcified plaques along the right coronary artery (RCA) with severe (≥70%) ostial stenosis (arrowhead). (b) Colored map overlaid on a standard two-chamber reformation shows the iodine attenuation obtained with multienergy CT. Note the decreased perfusion at the basal inferior left ventricular wall (arrowhead). The significant stenosis at the RCA ostium was confirmed with conventional angiography (not shown).

Myocardial delayed enhancement (MDE) in a patient with sarcoidosis.                        (a) Iodine map (mid short-axis view) overlaid on a standard CT image shows                        linear MDE in the basal interventricular septum (black arrowheads) and along                        the subepicardial aspect of the basal anterolateral wall (white arrowheads).                        (b) Corresponding MR image shows late gadolinium enhancement (LGE) with a                        similar distribution (arrowheads).

Figure 13a. Myocardial delayed enhancement (MDE) in a patient with sarcoidosis. (a) Iodine map (mid short-axis view) overlaid on a standard CT image shows linear MDE in the basal interventricular septum (black arrowheads) and along the subepicardial aspect of the basal anterolateral wall (white arrowheads). (b) Corresponding MR image shows late gadolinium enhancement (LGE) with a similar distribution (arrowheads).

Myocardial delayed enhancement (MDE) in a patient with sarcoidosis.                        (a) Iodine map (mid short-axis view) overlaid on a standard CT image shows                        linear MDE in the basal interventricular septum (black arrowheads) and along                        the subepicardial aspect of the basal anterolateral wall (white arrowheads).                        (b) Corresponding MR image shows late gadolinium enhancement (LGE) with a                        similar distribution (arrowheads).

Figure 13b. Myocardial delayed enhancement (MDE) in a patient with sarcoidosis. (a) Iodine map (mid short-axis view) overlaid on a standard CT image shows linear MDE in the basal interventricular septum (black arrowheads) and along the subepicardial aspect of the basal anterolateral wall (white arrowheads). (b) Corresponding MR image shows late gadolinium enhancement (LGE) with a similar distribution (arrowheads).

Iodine maps can be used for lesion characterization, either on their own or along with VNC images. Hyperattenuating lesions at contrast-enhanced CT could be either contrast enhancement or hemorrhage. The presence of iodine in the lesion confirms contrast enhancement versus hemorrhage, which is useful in distinguishing hemorrhagic cyst versus neoplasm in the kidney or brain, active bleeding versus hematoma, or contrast material staining after angiography versus hemorrhage in the brain (54,70). The presence of iodine in a hypoattenuating lesion raises the possibility of neoplasm (64).

Specific iodine thresholds have been identified for detection and characterization of several pathologic conditions. The threshold varies between different organs, pathologic conditions, scanners, vendors, and acquisition parameters, ranging from 0.5 to 2.0 mg/mL (78,79). For example, an iodine cutoff of 1.74 mg/mL allows distinction between thrombus and slow flow in the left atrial appendage (sensitivity 97%, specificity 100%) (80), a cutoff of 1.58 mg/mL allows distinction between malignant and benign mediastinal mass (sensitivity 100%, specificity 80%) (81) (Fig 14), and a cutoff of 0.9 mg/mL allows distinction between neoplastic and bland thrombus (sensitivity 100%, specificity 95.2%) (82).

Tumor perfusion assessed with multienergy CT in a 65-year-old man. (a)                        Axial CT image shows a large heterogeneous superior mediastinal mass                        (arrowheads). (b) Iodine map at the same level shows iodine uptake in the                        lesion (arrow) of 1.5 mg/mL, which favors malignancy.

Figure 14a. Tumor perfusion assessed with multienergy CT in a 65-year-old man. (a) Axial CT image shows a large heterogeneous superior mediastinal mass (arrowheads). (b) Iodine map at the same level shows iodine uptake in the lesion (arrow) of 1.5 mg/mL, which favors malignancy.

Tumor perfusion assessed with multienergy CT in a 65-year-old man. (a)                        Axial CT image shows a large heterogeneous superior mediastinal mass                        (arrowheads). (b) Iodine map at the same level shows iodine uptake in the                        lesion (arrow) of 1.5 mg/mL, which favors malignancy.

Figure 14b. Tumor perfusion assessed with multienergy CT in a 65-year-old man. (a) Axial CT image shows a large heterogeneous superior mediastinal mass (arrowheads). (b) Iodine map at the same level shows iodine uptake in the lesion (arrow) of 1.5 mg/mL, which favors malignancy.

The iodine concentration correlates with the metabolic activity at fluorodeoxyglucose (FDG) PET/CT (83) and allows evaluation of response to therapy, including that of gastrointestinal tumors, which may not show immediate reduction in size after successful treatment (84). In liver tumors after transcatheter arterial embolization, absence or decrease of iodine uptake indicates successful therapy (60).

Iodine maps have also been shown to improve the conspicuity of lesions, including in the pancreas and kidneys (85), and evaluation of the mural integrity of the bowel, appendix, and gallbladder (86). In CT arthrography, iodine allows distinction of contrast material in a labral tear from a calcified body (55).

Virtual Noncontrast Images.—VNC images, also called water-only images, are those in which the iodine component in each pixel has been removed using iodine-water decomposition. These images resemble a true noncontrast (TNC) conventional CT acquisition, with similar attenuation numbers (87). Hence, VNC images can replace TNC images in multiphasic studies such as CT angiography for evaluation of aneurysmal endoleaks, CT urography, or liver CT, providing radiation dose savings of up to 50% (88,89). Additional dose savings of up to 64% can be achieved by acquiring only venous phase images and using VMIs for arterial phases (90,91).

VNC images can be used for lesion characterization, either alone or with iodine maps, particularly in a hyperattenuating lesion at postcontrast CT. In a hyperattenuating renal cyst at postcontrast CT, presence of high attenuation on VNC images indicates a hemorrhagic or proteinaceous cyst (Fig 15), while absence of high attenuation on VNC images indicates contrast enhancement of a neoplasm (64) (Fig 16).

Lesion characterization using VNC images and iodine maps. (a) Axial                        contrast-enhanced CT image shows a hyperattenuating lesion (arrow) in the                        upper pole of the left kidney. This could be due to a complicated cyst or an                        enhancing neoplasm. (b) Corresponding VNC image shows that the lesion is                        hyperattenuating (arrow). (c) Corresponding iodine map does not show any                        iodine uptake in the lesion (arrow). These multienergy CT findings are                        consistent with a hemorrhagic cyst.

Figure 15a. Lesion characterization using VNC images and iodine maps. (a) Axial contrast-enhanced CT image shows a hyperattenuating lesion (arrow) in the upper pole of the left kidney. This could be due to a complicated cyst or an enhancing neoplasm. (b) Corresponding VNC image shows that the lesion is hyperattenuating (arrow). (c) Corresponding iodine map does not show any iodine uptake in the lesion (arrow). These multienergy CT findings are consistent with a hemorrhagic cyst.

Lesion characterization using VNC images and iodine maps. (a) Axial                        contrast-enhanced CT image shows a hyperattenuating lesion (arrow) in the                        upper pole of the left kidney. This could be due to a complicated cyst or an                        enhancing neoplasm. (b) Corresponding VNC image shows that the lesion is                        hyperattenuating (arrow). (c) Corresponding iodine map does not show any                        iodine uptake in the lesion (arrow). These multienergy CT findings are                        consistent with a hemorrhagic cyst.

Figure 15b. Lesion characterization using VNC images and iodine maps. (a) Axial contrast-enhanced CT image shows a hyperattenuating lesion (arrow) in the upper pole of the left kidney. This could be due to a complicated cyst or an enhancing neoplasm. (b) Corresponding VNC image shows that the lesion is hyperattenuating (arrow). (c) Corresponding iodine map does not show any iodine uptake in the lesion (arrow). These multienergy CT findings are consistent with a hemorrhagic cyst.

Lesion characterization using VNC images and iodine maps. (a) Axial                        contrast-enhanced CT image shows a hyperattenuating lesion (arrow) in the                        upper pole of the left kidney. This could be due to a complicated cyst or an                        enhancing neoplasm. (b) Corresponding VNC image shows that the lesion is                        hyperattenuating (arrow). (c) Corresponding iodine map does not show any                        iodine uptake in the lesion (arrow). These multienergy CT findings are                        consistent with a hemorrhagic cyst.

Figure 15c. Lesion characterization using VNC images and iodine maps. (a) Axial contrast-enhanced CT image shows a hyperattenuating lesion (arrow) in the upper pole of the left kidney. This could be due to a complicated cyst or an enhancing neoplasm. (b) Corresponding VNC image shows that the lesion is hyperattenuating (arrow). (c) Corresponding iodine map does not show any iodine uptake in the lesion (arrow). These multienergy CT findings are consistent with a hemorrhagic cyst.

Lesion characterization using VNC images and iodine maps. (a) Axial                        contrast-enhanced CT image in a patient with abdominal pain shows a                        hyperattenuating lesion (arrow) in the upper pole of the left kidney. This                        could be due to a complicated cyst or an enhancing neoplasm. (b)                        Corresponding VNC image shows no high attenuation in the lesion (arrow). (c)                        Corresponding iodine map shows significant iodine uptake in the lesion                        (arrow). This constellation of findings indicates that the hyperattenuation                        is due to contrast enhancement, consistent with a neoplasm.

Figure 16a. Lesion characterization using VNC images and iodine maps. (a) Axial contrast-enhanced CT image in a patient with abdominal pain shows a hyperattenuating lesion (arrow) in the upper pole of the left kidney. This could be due to a complicated cyst or an enhancing neoplasm. (b) Corresponding VNC image shows no high attenuation in the lesion (arrow). (c) Corresponding iodine map shows significant iodine uptake in the lesion (arrow). This constellation of findings indicates that the hyperattenuation is due to contrast enhancement, consistent with a neoplasm.

Lesion characterization using VNC images and iodine maps. (a) Axial                        contrast-enhanced CT image in a patient with abdominal pain shows a                        hyperattenuating lesion (arrow) in the upper pole of the left kidney. This                        could be due to a complicated cyst or an enhancing neoplasm. (b)                        Corresponding VNC image shows no high attenuation in the lesion (arrow). (c)                        Corresponding iodine map shows significant iodine uptake in the lesion                        (arrow). This constellation of findings indicates that the hyperattenuation                        is due to contrast enhancement, consistent with a neoplasm.

Figure 16b. Lesion characterization using VNC images and iodine maps. (a) Axial contrast-enhanced CT image in a patient with abdominal pain shows a hyperattenuating lesion (arrow) in the upper pole of the left kidney. This could be due to a complicated cyst or an enhancing neoplasm. (b) Corresponding VNC image shows no high attenuation in the lesion (arrow). (c) Corresponding iodine map shows significant iodine uptake in the lesion (arrow). This constellation of findings indicates that the hyperattenuation is due to contrast enhancement, consistent with a neoplasm.

Lesion characterization using VNC images and iodine maps. (a) Axial                        contrast-enhanced CT image in a patient with abdominal pain shows a                        hyperattenuating lesion (arrow) in the upper pole of the left kidney. This                        could be due to a complicated cyst or an enhancing neoplasm. (b)                        Corresponding VNC image shows no high attenuation in the lesion (arrow). (c)                        Corresponding iodine map shows significant iodine uptake in the lesion                        (arrow). This constellation of findings indicates that the hyperattenuation                        is due to contrast enhancement, consistent with a neoplasm.

Figure 16c. Lesion characterization using VNC images and iodine maps. (a) Axial contrast-enhanced CT image in a patient with abdominal pain shows a hyperattenuating lesion (arrow) in the upper pole of the left kidney. This could be due to a complicated cyst or an enhancing neoplasm. (b) Corresponding VNC image shows no high attenuation in the lesion (arrow). (c) Corresponding iodine map shows significant iodine uptake in the lesion (arrow). This constellation of findings indicates that the hyperattenuation is due to contrast enhancement, consistent with a neoplasm.

In suspected gastrointestinal bleeding, presence of pixels with high intensity on iodine maps with absence on VNC images indicates active bleeding; presence on both iodine maps and VNC images indicates calcium or pills; and absence on iodine map with presence on VNC images indicates hemorrhage (92). Similarly, VNC images allow differentiation between contrast material extravasation after intra-arterial revascularization and hemorrhage in a patient with ischemic stroke (93).

Since the attenuation values of VNC images correlate with those of TNC images, they can be used to characterize incidentally encountered lesions such as adrenal adenoma, avoiding the need for additional imaging, which saves radiation, cost, and patient anxiety (94). VNC images can be used for measuring calcium score from contrast-enhanced coronary CT angiography, thus saving the radiation dose of a TNC acquisition (95).

Uric Acid Images.—Tissues with uric acid are composed of light chemical elements (hydrogen, carbon, oxygen, and nitrogen) (35,9698). These can be separated with three-material decomposition of uric acid, calcium, and water or two-material decomposition of iodine and water (Fig 17). Uric acid imaging is useful in urinary stones larger than 3 mm (98) and gout. Uric acid stones are often treated with urinary alkalization as a first-line treatment, while non–uric acid stones often require urologic intervention. Conventional ROI-based attenuation measurements are not robust with variable Hounsfield unit values and challenging in mixed stones (97).

Uric acid decomposition. With a three-material decomposition                        technique, uric acid, calcium, and soft tissue can be accurately                        distinguished from each other on the basis of differential attenuation at                        the two energy levels. Material above the line is color coded differently                        from material below the line.

Figure 17. Uric acid decomposition. With a three-material decomposition technique, uric acid, calcium, and soft tissue can be accurately distinguished from each other on the basis of differential attenuation at the two energy levels. Material above the line is color coded differently from material below the line.

In the three-material decomposition technique, uric acid stones are coded with color different from that of non–uric acid stones (Fig 18). In the two-material technique, uric acid stones are visualized only on the water images, whereas non–uric acid stones are seen on both water and iodine images (35,97,98) (Fig 19). Multienergy CT also allows qualitative and quantitative determination of monosodium urate crystal deposition in tissues, which is useful in diagnosing gout (Fig 20). Multienergy CT is useful in challenging or atypical scenarios, to determine the severity of disease, and to monitor response to therapy (55,99).

Urinary stone composition at dual-source CT. (a) Axial CT image in a                        patient with renal colic shows a stone (arrow) in the right kidney. (b) Uric                        acid–decomposed image shows that the stone (arrow) is color coded                        red, indicating a uric acid calculus. A calcium stone will be color coded                        blue.

Figure 18a. Urinary stone composition at dual-source CT. (a) Axial CT image in a patient with renal colic shows a stone (arrow) in the right kidney. (b) Uric acid–decomposed image shows that the stone (arrow) is color coded red, indicating a uric acid calculus. A calcium stone will be color coded blue.

Urinary stone composition at dual-source CT. (a) Axial CT image in a                        patient with renal colic shows a stone (arrow) in the right kidney. (b) Uric                        acid–decomposed image shows that the stone (arrow) is color coded                        red, indicating a uric acid calculus. A calcium stone will be color coded                        blue.

Figure 18b. Urinary stone composition at dual-source CT. (a) Axial CT image in a patient with renal colic shows a stone (arrow) in the right kidney. (b) Uric acid–decomposed image shows that the stone (arrow) is color coded red, indicating a uric acid calculus. A calcium stone will be color coded blue.

Urinary stone composition at rapid kVp switching CT. (a) Axial CT                        image shows a stone (arrow) in the left kidney. (b) Iodine image at the same                        level does not show the stone (arrow). (c) Water image at the same level                        shows the stone (arrow), indicating that it is a uric acid calculus. A                        calcium stone would be seen on both iodine and water images.

Figure 19a. Urinary stone composition at rapid kVp switching CT. (a) Axial CT image shows a stone (arrow) in the left kidney. (b) Iodine image at the same level does not show the stone (arrow). (c) Water image at the same level shows the stone (arrow), indicating that it is a uric acid calculus. A calcium stone would be seen on both iodine and water images.

Urinary stone composition at rapid kVp switching CT. (a) Axial CT                        image shows a stone (arrow) in the left kidney. (b) Iodine image at the same                        level does not show the stone (arrow). (c) Water image at the same level                        shows the stone (arrow), indicating that it is a uric acid calculus. A                        calcium stone would be seen on both iodine and water images.

Figure 19b. Urinary stone composition at rapid kVp switching CT. (a) Axial CT image shows a stone (arrow) in the left kidney. (b) Iodine image at the same level does not show the stone (arrow). (c) Water image at the same level shows the stone (arrow), indicating that it is a uric acid calculus. A calcium stone would be seen on both iodine and water images.

Urinary stone composition at rapid kVp switching CT. (a) Axial CT                        image shows a stone (arrow) in the left kidney. (b) Iodine image at the same                        level does not show the stone (arrow). (c) Water image at the same level                        shows the stone (arrow), indicating that it is a uric acid calculus. A                        calcium stone would be seen on both iodine and water images.

Figure 19c. Urinary stone composition at rapid kVp switching CT. (a) Axial CT image shows a stone (arrow) in the left kidney. (b) Iodine image at the same level does not show the stone (arrow). (c) Water image at the same level shows the stone (arrow), indicating that it is a uric acid calculus. A calcium stone would be seen on both iodine and water images.

Gout at multienergy CT. (a) Axial CT image in a 57-year-old man with                        ankle pain shows mild soft tissue (arrow) surrounding the tibia. (b) Axial                        multienergy CT image shows monosodium urate crystals color coded green                        (arrow) around the tibia. (c) Three-dimensional volume-rendered image of the                        ankle shows the gout crystals color coded green.

Figure 20a. Gout at multienergy CT. (a) Axial CT image in a 57-year-old man with ankle pain shows mild soft tissue (arrow) surrounding the tibia. (b) Axial multienergy CT image shows monosodium urate crystals color coded green (arrow) around the tibia. (c) Three-dimensional volume-rendered image of the ankle shows the gout crystals color coded green.

Gout at multienergy CT. (a) Axial CT image in a 57-year-old man with                        ankle pain shows mild soft tissue (arrow) surrounding the tibia. (b) Axial                        multienergy CT image shows monosodium urate crystals color coded green                        (arrow) around the tibia. (c) Three-dimensional volume-rendered image of the                        ankle shows the gout crystals color coded green.

Figure 20b. Gout at multienergy CT. (a) Axial CT image in a 57-year-old man with ankle pain shows mild soft tissue (arrow) surrounding the tibia. (b) Axial multienergy CT image shows monosodium urate crystals color coded green (arrow) around the tibia. (c) Three-dimensional volume-rendered image of the ankle shows the gout crystals color coded green.

Gout at multienergy CT. (a) Axial CT image in a 57-year-old man with                        ankle pain shows mild soft tissue (arrow) surrounding the tibia. (b) Axial                        multienergy CT image shows monosodium urate crystals color coded green                        (arrow) around the tibia. (c) Three-dimensional volume-rendered image of the                        ankle shows the gout crystals color coded green.

Figure 20c. Gout at multienergy CT. (a) Axial CT image in a 57-year-old man with ankle pain shows mild soft tissue (arrow) surrounding the tibia. (b) Axial multienergy CT image shows monosodium urate crystals color coded green (arrow) around the tibia. (c) Three-dimensional volume-rendered image of the ankle shows the gout crystals color coded green.

Calcium Decomposition Algorithms.—Calcium-noncalcium pairs can be obtained by modifying the iodine-water decomposition curves to that of bone, that is, by changing the dual-energy ratio (DER) to 1.44 from 2.24 for iodine (at 100 kVp/tin 140 kVp) (70). Calcium-noncalcium pairs allow distinction of calcium from hemorrhage in a noncontrast brain CT study. Calcium is seen on calcium-overlay images, whereas hemorrhage is seen on the noncalcium images (70). This avoids the need for follow-up CT, which is the standard of care with conventional CT. This technique also allows distinction between calcium and hemorrhage within a mass (100).

Calcium can be separated from iodine by using calcium-iodine basis pair two-material decomposition or calcium-iodine-blood three-material decomposition (101). Calcium-iodine pairs have been used to detect and quantify calcium in plaques in the abdominal aorta (102). Material-specific calcium images can help in analyzing the composition and morphologic features of the vascular plaque (103).

Subtraction of bone is essential for three-dimensional reconstructions in vascular CT angiography, particularly in the extremities and neck. Multienergy CT allows subtraction of bone using a three-material decomposition algorithm, where bone-containing voxels are located below a virtual separation line between bone and iodine and assigned a low attenuation value such as −1024 HU. This generates high-quality bone-free maximum intensity projection (MIP) and volume-rendered images (Fig 21a) (70). The multienergy CT technique is more accurate with less erosion of vessels near bones and less labor-intensive than the conventional threshold-based bone subtraction techniques. However, some of the calcium may be artifactually removed (104,105). Bone subtraction also improves detection of subtle subdural hematomas or masses near the skull bones (70).

Bone subtraction and calcified plaque subtraction using multienergy                        CT. (a) Coronal CT image after bone subtraction with multienergy CT. The                        bones are assigned a negative attenuation value. (b) Coronal maximum                        intensity projection (MIP) image after bone subtraction provides a                        comprehensive evaluation of the peripheral arteries. Owing to extensive                        plaques, the lumen cannot be adequately visualized. (c) Coronal MIP image                        after removal of calcified plaques shows a luminogram, which gives an                        estimate of the extent and distribution of the disease.

Figure 21a. Bone subtraction and calcified plaque subtraction using multienergy CT. (a) Coronal CT image after bone subtraction with multienergy CT. The bones are assigned a negative attenuation value. (b) Coronal maximum intensity projection (MIP) image after bone subtraction provides a comprehensive evaluation of the peripheral arteries. Owing to extensive plaques, the lumen cannot be adequately visualized. (c) Coronal MIP image after removal of calcified plaques shows a luminogram, which gives an estimate of the extent and distribution of the disease.

Bone subtraction and calcified plaque subtraction using multienergy                        CT. (a) Coronal CT image after bone subtraction with multienergy CT. The                        bones are assigned a negative attenuation value. (b) Coronal maximum                        intensity projection (MIP) image after bone subtraction provides a                        comprehensive evaluation of the peripheral arteries. Owing to extensive                        plaques, the lumen cannot be adequately visualized. (c) Coronal MIP image                        after removal of calcified plaques shows a luminogram, which gives an                        estimate of the extent and distribution of the disease.

Figure 21b. Bone subtraction and calcified plaque subtraction using multienergy CT. (a) Coronal CT image after bone subtraction with multienergy CT. The bones are assigned a negative attenuation value. (b) Coronal maximum intensity projection (MIP) image after bone subtraction provides a comprehensive evaluation of the peripheral arteries. Owing to extensive plaques, the lumen cannot be adequately visualized. (c) Coronal MIP image after removal of calcified plaques shows a luminogram, which gives an estimate of the extent and distribution of the disease.

Bone subtraction and calcified plaque subtraction using multienergy                        CT. (a) Coronal CT image after bone subtraction with multienergy CT. The                        bones are assigned a negative attenuation value. (b) Coronal maximum                        intensity projection (MIP) image after bone subtraction provides a                        comprehensive evaluation of the peripheral arteries. Owing to extensive                        plaques, the lumen cannot be adequately visualized. (c) Coronal MIP image                        after removal of calcified plaques shows a luminogram, which gives an                        estimate of the extent and distribution of the disease.

Figure 21c. Bone subtraction and calcified plaque subtraction using multienergy CT. (a) Coronal CT image after bone subtraction with multienergy CT. The bones are assigned a negative attenuation value. (b) Coronal maximum intensity projection (MIP) image after bone subtraction provides a comprehensive evaluation of the peripheral arteries. Owing to extensive plaques, the lumen cannot be adequately visualized. (c) Coronal MIP image after removal of calcified plaques shows a luminogram, which gives an estimate of the extent and distribution of the disease.

After bone subtraction, calcium in the atherosclerotic plaques can be removed using morphologic criteria to generate an MIP luminogram, which improves luminal evaluation (Fig 21b, 21c). This technique is more successful in the proximal pelvic and thigh arteries (106). However, it is less reliable near the skull base and below the knee and is limited by calcium blooming (106). The accuracy of calcium subtraction can be improved using a three-material decomposition algorithm that uses calcium as one base material and a mixture of soft tissue and iodine (using attenuation constraints) as the other base material while maintaining CT values of air and fat (107109). The calcium separation with this technique is more complete than using iodine and water algorithms (calcium is present in both these images) (107) and also removes calcium blooming (108,109).

Virtual noncalcium (VNCa) images are those in which the bone mineral is subtracted from the trabecular bone using a three-material decomposition of bone mineral, yellow marrow, and soft tissue. VNCa images are primarily used in musculoskeletal imaging to evaluate the bone marrow when MRI cannot be performed, particularly in an acute setting or owing to contraindications. VNCa images allow detection of bone marrow edema with high sensitivity (64%–96%) and specificity (40%–99%) (55,110,111). Bone marrow edema can be visualized on color-overlay images or can be quantified (Fig 22). This technique is useful in detection of subtle fractures (55), subtle bone marrow neoplasms such as osteolytic multiple myeloma (112,113), and subtle osteitis in inflammatory osteitis (114), which are not evident at conventional CT. VNCa images help in CT-guided biopsy of subtle bone neoplasms (115).

Use of virtual noncalcium (VNCa) images in a 69-year-old man with                        rectal cancer who underwent urinary diversion. (a) Axial CT image shows gas                        (arrow) and cortical irregularity at the superior border of the pubic                        symphysis. (b) VNCa image shows bone marrow edema (arrows) in the pubic                        symphysis. Note the higher attenuation compared with that of the femora and                        ischia. (c) Corresponding fat-suppressed T2-weighted MR image confirms the                        presence of edema (arrows) in the pubic symphysis.

Figure 22a. Use of virtual noncalcium (VNCa) images in a 69-year-old man with rectal cancer who underwent urinary diversion. (a) Axial CT image shows gas (arrow) and cortical irregularity at the superior border of the pubic symphysis. (b) VNCa image shows bone marrow edema (arrows) in the pubic symphysis. Note the higher attenuation compared with that of the femora and ischia. (c) Corresponding fat-suppressed T2-weighted MR image confirms the presence of edema (arrows) in the pubic symphysis.

Use of virtual noncalcium (VNCa) images in a 69-year-old man with                        rectal cancer who underwent urinary diversion. (a) Axial CT image shows gas                        (arrow) and cortical irregularity at the superior border of the pubic                        symphysis. (b) VNCa image shows bone marrow edema (arrows) in the pubic                        symphysis. Note the higher attenuation compared with that of the femora and                        ischia. (c) Corresponding fat-suppressed T2-weighted MR image confirms the                        presence of edema (arrows) in the pubic symphysis.

Figure 22b. Use of virtual noncalcium (VNCa) images in a 69-year-old man with rectal cancer who underwent urinary diversion. (a) Axial CT image shows gas (arrow) and cortical irregularity at the superior border of the pubic symphysis. (b) VNCa image shows bone marrow edema (arrows) in the pubic symphysis. Note the higher attenuation compared with that of the femora and ischia. (c) Corresponding fat-suppressed T2-weighted MR image confirms the presence of edema (arrows) in the pubic symphysis.

Use of virtual noncalcium (VNCa) images in a 69-year-old man with                        rectal cancer who underwent urinary diversion. (a) Axial CT image shows gas                        (arrow) and cortical irregularity at the superior border of the pubic                        symphysis. (b) VNCa image shows bone marrow edema (arrows) in the pubic                        symphysis. Note the higher attenuation compared with that of the femora and                        ischia. (c) Corresponding fat-suppressed T2-weighted MR image confirms the                        presence of edema (arrows) in the pubic symphysis.

Figure 22c. Use of virtual noncalcium (VNCa) images in a 69-year-old man with rectal cancer who underwent urinary diversion. (a) Axial CT image shows gas (arrow) and cortical irregularity at the superior border of the pubic symphysis. (b) VNCa image shows bone marrow edema (arrows) in the pubic symphysis. Note the higher attenuation compared with that of the femora and ischia. (c) Corresponding fat-suppressed T2-weighted MR image confirms the presence of edema (arrows) in the pubic symphysis.

Other Material Decomposition Images.—Numerous innovative decomposition algorithms have been developed for selectively highlighting or suppressing specific materials (Table 2). Iron maps allow quantification of iron in the liver and myocardium (116,117). Iron-containing entities such as pigmented villonodular synovitis, hemophiliac arthropathy, hemochromatosis arthropathy, and metallosis can be diagnosed (118,119). Virtual noniron images using iron, fat, and liver tissue allow quantification of liver fat (120). Material-specific fat images allow accurate quantification of hepatic steatosis, described as percentage fat fraction (121). They can also be used to characterize fat-containing lesions (Fig 23), including adrenal lesions (122).

Use of material-specific fat images in a 33-year-old woman with a                        history of trauma. (a) Axial CT image of the abdomen shows a band of                        hypoattenuation (arrows) in the liver, which could be due to laceration or                        focal fatty infiltration. (b) Material-specific fat image at the same level                        shows the presence of fat in the lesion with a fat fraction of 47.5%,                        consistent with focal fatty infiltration.

Figure 23a. Use of material-specific fat images in a 33-year-old woman with a history of trauma. (a) Axial CT image of the abdomen shows a band of hypoattenuation (arrows) in the liver, which could be due to laceration or focal fatty infiltration. (b) Material-specific fat image at the same level shows the presence of fat in the lesion with a fat fraction of 47.5%, consistent with focal fatty infiltration.

Use of material-specific fat images in a 33-year-old woman with a                        history of trauma. (a) Axial CT image of the abdomen shows a band of                        hypoattenuation (arrows) in the liver, which could be due to laceration or                        focal fatty infiltration. (b) Material-specific fat image at the same level                        shows the presence of fat in the lesion with a fat fraction of 47.5%,                        consistent with focal fatty infiltration.

Figure 23b. Use of material-specific fat images in a 33-year-old woman with a history of trauma. (a) Axial CT image of the abdomen shows a band of hypoattenuation (arrows) in the liver, which could be due to laceration or focal fatty infiltration. (b) Material-specific fat image at the same level shows the presence of fat in the lesion with a fat fraction of 47.5%, consistent with focal fatty infiltration.

Tagged fecal material images can provide artifact-free visualization of the colonic lumen in CT colonography (123). Material pair images, with one highlighting the brain gray-white matter difference and the other a flattened “brain edema” image without gray-white matter difference, can be used to detect brain edema (124). A two-material decomposition algorithm of trabecular bone and soft tissue allows quantification of trabecular bone density without phantom calibration (125) and measurement of density adjacent to a prosthesis (126). Collagen maps can improve visualization and characterization of tendons in the hand and foot and some ligaments of the knee (127). It is an alternative technique for evaluation of disk degeneration, bulge, and herniation, seen as loss of disk attenuation (128).

Xenon maps are used to evaluate lung ventilation in several pulmonary processes such as chronic obstructive pulmonary disease (COPD) and asthma on top of morphologic and perfusion information (129). Dark-blood images, by suppressing the signal of the vascular lumen, may increase contrast between the lumen and wall, improving evaluation of the vascular wall, particularly for intramural hematoma and vasculitis (130). Multienergy CT has been shown to allow evaluation of the accumulation of liposomal iodine and gold nanoparticles after nanoparticle-augmented radiation therapy in tumors (131).

Zeff, Electron Density, and Fingerprinting Images.—Zeff-weighted images are color coded on the basis of the Zeff of the materials. During the material decomposition process, the coefficients of photoelectric and Compton scatter components (α1 and α2) are functions of the spatial distribution of tissue [α1 (x,y) and α2 (x,y)]. The ratio of these coefficients is proportional to the effective atomic number, Z: α1 (x,y)/α2 (x,y) ∝ (Z2 = z3/z). From this ratio, Zeff can be estimated as a function of (x,y). These images provide a higher level of discrimination, since they provide the material makeup (1,132). In urinary stones, low Zeff is suggestive of uric acid calculi and high Zeff is suggestive of non–uric acid calculi (Fig 24).

Use of effective atomic number (Zeff)–weighted images in a                        patient with renal colic. (a) Axial CT image shows a nonobstructive calculus                        (arrow) in the left kidney. (b) Zeff–based image at the same level                        shows that the calculus is color coded blue (arrow), indicating the presence                        of calcium.

Figure 24a. Use of effective atomic number (Zeff)–weighted images in a patient with renal colic. (a) Axial CT image shows a nonobstructive calculus (arrow) in the left kidney. (b) Zeff–based image at the same level shows that the calculus is color coded blue (arrow), indicating the presence of calcium.

Use of effective atomic number (Zeff)–weighted images in a                        patient with renal colic. (a) Axial CT image shows a nonobstructive calculus                        (arrow) in the left kidney. (b) Zeff–based image at the same level                        shows that the calculus is color coded blue (arrow), indicating the presence                        of calcium.

Figure 24b. Use of effective atomic number (Zeff)–weighted images in a patient with renal colic. (a) Axial CT image shows a nonobstructive calculus (arrow) in the left kidney. (b) Zeff–based image at the same level shows that the calculus is color coded blue (arrow), indicating the presence of calcium.

Similarly, electron density (ρe)–weighted images can be generated (133). Zeff and ρe images allow discrimination of meniscal calcium pyrophosphate deposits from calcium hydroxyapatite deposits in subchondral and trabecular bone (134). Zeff and ρe images can provide superior stopping power ratio (SPR) relative to water determination in phantom materials and organic tissue samples, which is important to calculate the dose distribution and proton range (135).

CT fingerprinting involves a two-dimensional histogram of photoelectric and Compton attenuation along the y-axis and x-axis, respectively (136) (Fig 25). Materials occupy unique locations within this space, providing a fingerprint of the object composition. Using custom software, pixel clusters in the two-dimensional histogram corresponding to a particular CT image can be identified, which in turn can be used for identification and quantification of specific materials (Fig 25), such as isoattenuating gallstones (136).

CT fingerprinting. (a) The pixels of a chest CT image are plotted                        according to the photoelectric and Compton attenuation effects using                        multienergy decomposition. Note the particular distribution of certain                        materials in specific regions of the plot, with the tendency to follow a                        linear path (material fingerprint). (b) The calcium (green) and iodine                        (pink) clusters have been identified on the two-dimensional histogram, and                        an ROI has been drawn around them. Using custom software, corresponding                        areas on the conventional CT image can be highlighted and                        quantified.

Figure 25a. CT fingerprinting. (a) The pixels of a chest CT image are plotted according to the photoelectric and Compton attenuation effects using multienergy decomposition. Note the particular distribution of certain materials in specific regions of the plot, with the tendency to follow a linear path (material fingerprint). (b) The calcium (green) and iodine (pink) clusters have been identified on the two-dimensional histogram, and an ROI has been drawn around them. Using custom software, corresponding areas on the conventional CT image can be highlighted and quantified.

CT fingerprinting. (a) The pixels of a chest CT image are plotted                        according to the photoelectric and Compton attenuation effects using                        multienergy decomposition. Note the particular distribution of certain                        materials in specific regions of the plot, with the tendency to follow a                        linear path (material fingerprint). (b) The calcium (green) and iodine                        (pink) clusters have been identified on the two-dimensional histogram, and                        an ROI has been drawn around them. Using custom software, corresponding                        areas on the conventional CT image can be highlighted and                        quantified.

Figure 25b. CT fingerprinting. (a) The pixels of a chest CT image are plotted according to the photoelectric and Compton attenuation effects using multienergy decomposition. Note the particular distribution of certain materials in specific regions of the plot, with the tendency to follow a linear path (material fingerprint). (b) The calcium (green) and iodine (pink) clusters have been identified on the two-dimensional histogram, and an ROI has been drawn around them. Using custom software, corresponding areas on the conventional CT image can be highlighted and quantified.

Radiation Doses in Multienergy CT

A common misconception about multienergy CT is that its radiation dose is twice that of single-energy CT owing to the two acquisitions. However, it is well established that

Teaching Point the radiation doses of multienergy CT are comparable to or even lower than those of single-energy CT (21,137,138)
, often less than the dose reference level of body CT (139).

Several radiation dose–reduction strategies are used in multienergy CT scanners. In dual-source CT, optimized tube currents for the two tubes, tube current modulation, and a tin filter reduce radiation doses. In rapid kVp switching, a longer acquisition from the low-energy 80 kVp is used. In dual-layer CT, although a higher tube voltage of 120 or 140 kVp is used, dose neutrality with single-energy CT is maintained by proportionally reducing the tube current.

Additional radiation dose–reduction strategies such as automatic tube current modulation, adaptive dose shielding, and iterative reconstruction are available with several multienergy CT technologies (139). The availability of correction algorithms such as energy-domain noise reduction for multienergy CT allows dose efficiency (56,57). Replacement of true noncontrast (TNC) images by VNC images in multiphasic studies (88) and in characterization of incidental lesions, which obviates further CT imaging, also saves radiation doses (140).

Conclusion

Multienergy CT provides tissue characterization beyond that possible with a conventional CT scanner. There are several multienergy CT technologies at both the source and detector levels. Material decomposition is performed in the projection or image domain using two- or multimaterial decomposition algorithms. Iodine maps, VNC images, and VMIs are the most commonly used multienergy CT images in clinical practice.

Disclosures of Conflicts of Interest.—P.R. Activities related to the present article: disclosed no relevant relationships. Activities not related to the present article: royalties from Elsevier. Other activities: disclosed no relevant relationships. D.B. Activities related to the present article: disclosed no relevant relationships. Activities not related to the present article: speaker for Boehringer Ingelheim Pharmaceuticals. Other activities: disclosed no relevant relationships. A.R.K. Activities related to the present article: disclosed no relevant relationships. Activities not related to the present article: research grants from GE Healthcare and Philips. Other activities: disclosed no relevant relationships. S.L. Activities related to the present article: disclosed no relevant relationships. Activities not related to the present article: disclosed no relevant relationships. Other activities: intellectual property owned by Mayo Clinic; issued patent licensed by Siemens.

Recipient of a Certificate of Merit award for an education exhibit at the 2019 RSNA Annual Meeting.

For this journal-based SA-CME activity, the authors P.R., D.B., A.R.K., and S.L. have provided disclosures; all other authors, the editor, and the reviewers have disclosed no relevant relationships.

References

  • 1. Alvarez RE, Macovski A. Energy-selective reconstructions in X-ray computerized tomography. Phys Med Biol 1976;21(5):733–744. Crossref, MedlineGoogle Scholar
  • 2. Millner MR, McDavid WD, Waggener RG, Dennis MJ, Payne WH, Sank VJ. Extraction of information from CT scans at different energies. Med Phys 1979;6(1):70–71. Crossref, MedlineGoogle Scholar
  • 3. Patino M, Prochowski A, Agrawal MD, et al. Material Separation Using Dual-Energy CT: Current and Emerging Applications. RadioGraphics 2016;36(4):1087–1105. LinkGoogle Scholar
  • 4. McCollough CH, Leng S, Yu L, Fletcher JG. Dual- and Multi-Energy CT: Principles, Technical Approaches, and Clinical Applications. Radiology 2015;276(3):637–653. LinkGoogle Scholar
  • 5. McKetty MH. The AAPM/RSNA physics tutorial for residents: x-ray attenuation. RadioGraphics 1998;18(1):151–163; quiz 149. LinkGoogle Scholar
  • 6. Siegel MJ, Kaza RK, Bolus DN, et al. White Paper of the Society of Computed Body Tomography and Magnetic Resonance on Dual-Energy CT. I. Technology and Terminology. J Comput Assist Tomogr 2016;40(6):841–845. Crossref, MedlineGoogle Scholar
  • 7. Leng S, Bruesewitz M, Tao S, et al. Photon-counting detector CT: system design and clinical applications of an emerging technology. RadioGraphics 2019;39(3):729–743. LinkGoogle Scholar
  • 8. Willemink MJ, Persson M, Pourmorteza A, Pelc NJ, Fleischmann D. Photon-counting CT: technical principles and clinical prospects. Radiology 2018;289(2):293–312. LinkGoogle Scholar
  • 9. Yu L, Ren L, Li Z, Leng S, McCollough CH. Dual-source multienergy CT with triple or quadruple x-ray beams. J Med Imaging (Bellingham) 2018;5(3):033502. MedlineGoogle Scholar
  • 10. Faby S, Kuchenbecker S, Sawall S, et al. Performance of today’s dual energy CT and future multi energy CT in virtual non-contrast imaging and in iodine quantification: a simulation study. Med Phys 2015;42(7):4349–4366. Crossref, MedlineGoogle Scholar
  • 11. Petersilka M, Bruder H, Krauss B, Stierstorfer K, Flohr TG. Technical principles of dual source CT. Eur J Radiol 2008;68(3):362–368. Crossref, MedlineGoogle Scholar
  • 12. Chandra N, Langan D. Gemstone Detector: Dual Energy Imaging via Fast kVp Switching. In: Johnson T, Fink C, Schonberg S, Reiser M, eds. Dual Energy CT in Clinical Practice: Medical Radiology. Berlin, Germany: Springer, 2011; 35–41. CrossrefGoogle Scholar
  • 13. Zhao W, Vernekohl D, Han F, et al. A unified material decomposition framework for quantitative dual- and triple-energy CT imaging. Med Phys 2018;45(7):2964–2977. Crossref, MedlineGoogle Scholar
  • 14. Zhang D, Li X, Liu B. Objective characterization of GE Discovery CT750 HD scanner: gemstone spectral imaging mode. Med Phys 2011;38(3):1178–1188. Crossref, MedlineGoogle Scholar
  • 15. Goo HW, Goo JM. Dual-energy CT: new horizon in medical imaging. Korean J Radiol 2017;18(4):555–569. Crossref, MedlineGoogle Scholar
  • 16. Leng S, Shiung M, Ai S, et al. Feasibility of discriminating uric acid from non-uric acid renal stones using consecutive spatially registered low- and high-energy scans obtained on a conventional CT scanner. AJR Am J Roentgenol 2015;204(1):92–97. Crossref, MedlineGoogle Scholar
  • 17. Diekhoff T, Ziegeler K, Feist E, et al. First experience with single-source dual-energy computed tomography in six patients with acute arthralgia: a feasibility experiment using joint aspiration as a reference. Skeletal Radiol 2015;44(11):1573–1577. Crossref, MedlineGoogle Scholar
  • 18. Euler A, Obmann MM, Szucs-Farkas Z, et al. Comparison of image quality and radiation dose between split-filter dual-energy images and single-energy images in single-source abdominal CT. Eur Radiol 2018;28(8):3405–3412. Crossref, MedlineGoogle Scholar
  • 19. Almeida IP, Schyns LE, Öllers MC, et al. Dual-energy CT quantitative imaging: a comparison study between twin-beam and dual-source CT scanners. Med Phys 2017;44(1):171–179. Crossref, MedlineGoogle Scholar
  • 20. Rassouli N, Etesami M, Dhanantwari A, Rajiah P. Detector-based spectral CT with a novel dual-layer technology: principles and applications. Insights Imaging 2017;8(6):589–598. Crossref, MedlineGoogle Scholar
  • 21. van Ommen F, de Jong HWAM, Dankbaar JW, Bennink E, Leiner T, Schilham AMR. Dose of CT protocols acquired in clinical routine using a dual-layer detector CT scanner: a preliminary report. Eur J Radiol 2019;112:65–71. Crossref, MedlineGoogle Scholar
  • 22. Muenzel D, Bar-Ness D, Roessl E, et al. Spectral Photon-counting CT: Initial Experience with Dual-Contrast Agent K-Edge Colonography. Radiology 2017;283(3):723–728. LinkGoogle Scholar
  • 23. Dangelmaier J, Bar-Ness D, Daerr H, et al. Experimental feasibility of spectral photon-counting computed tomography with two contrast agents for the detection of endoleaks following endovascular aortic repair. Eur Radiol 2018;28(8):3318–3325. Crossref, MedlineGoogle Scholar
  • 24. Symons R, Cork TE, Lakshmanan MN, et al. Dual-contrast agent photon-counting computed tomography of the heart: initial experience. Int J Cardiovasc Imaging 2017;33(8):1253–1261. Crossref, MedlineGoogle Scholar
  • 25. Yu Z, Leng S, Kappler S, et al. Noise performance of low-dose CT: comparison between an energy integrating detector and a photon counting detector using a whole-body research photon counting CT scanner. J Med Imaging (Bellingham) 2016;3(4):043503. Crossref, MedlineGoogle Scholar
  • 26. Taguchi K, Frey EC, Wang X, Iwanczyk JS, Barber WC. An analytical model of the effects of pulse pileup on the energy spectrum recorded by energy resolved photon counting x-ray detectors. Med Phys 2010;37(8):3957–3969. Crossref, MedlineGoogle Scholar
  • 27. Wang AS, Harrison D, Lobastov V, Tkaczyk JE. Pulse pileup statistics for energy discriminating photon counting x-ray detectors. Med Phys 2011;38(7):4265–4275. Crossref, MedlineGoogle Scholar
  • 28. Tao A, Huang R, Tao S, Michalak GJ, McCollough CH, Leng S. Dual-source photon counting detector CT with a tin filter: a phantom study on iodine quantification performance. Phys Med Biol 2019;64(11):115019. Crossref, MedlineGoogle Scholar
  • 29. Alvarez RE, Seppi E. A Comparison of Noise and Dose in Conventional and Energy Selective Computed Tomography. IEEE Trans Nucl Sci 1979;26(2):2853–2856. CrossrefGoogle Scholar
  • 30. Kalender WA, Klotz E, Kostaridou L. An algorithm for noise suppression in dual energy CT material density images. IEEE Trans Med Imaging 1988;7(3):218–224. Crossref, MedlineGoogle Scholar
  • 31. Lehmann LA, Alvarez RE, Macovski A, et al. Generalized image combinations in dual KVP digital radiography. Med Phys 1981;8(5):659–667. Crossref, MedlineGoogle Scholar
  • 32. Megibow AJ, Kambadakone A, Ananthakrishnan L. Dual-Energy Computed Tomography: Image Acquisition, Processing, and Workflow. Radiol Clin North Am 2018;56(4):507–520. Crossref, MedlineGoogle Scholar
  • 33. Yu L, Leng S, McCollough CH. Dual-energy CT-based monochromatic imaging. AJR Am J Roentgenol 2012;199(5 suppl):S9–S15. Crossref, MedlineGoogle Scholar
  • 34. Lambert JW, Sun Y, Gould RG, Ohliger MA, Li Z, Yeh BM. An Image-Domain Contrast Material Extraction Method for Dual-Energy Computed Tomography. Invest Radiol 2017;52(4):245–254. Crossref, MedlineGoogle Scholar
  • 35. Kyriakou Y, Meyer E, Prell D, Kachelriess M. Empirical beam hardening correction (EBHC) for CT. Med Phys 2010;37(10):5179–5187. Crossref, MedlineGoogle Scholar
  • 36. Johnson TR, Krauss B, Sedlmair M, et al. Material differentiation by dual energy CT: initial experience. Eur Radiol 2007;17(6):1510–1517. Crossref, MedlineGoogle Scholar
  • 37. Liu X, Yu L, Primak AN, McCollough CH. Quantitative imaging of element composition and mass fraction using dual-energy CT: three-material decomposition. Med Phys 2009;36(5):1602–1609. Crossref, MedlineGoogle Scholar
  • 38. Mendonca PR, Lamb P, Sahani DV. A Flexible Method for Multi-Material Decomposition of Dual-Energy CT Images. IEEE Trans Med Imaging 2014;33(1):99–116. Crossref, MedlineGoogle Scholar
  • 39. Wang X, Meier D, Taguchi K, Wagenaar DJ, Patt BE, Frey EC. Material separation in x-ray CT with energy resolved photon-counting detectors. Med Phys 2011;38(3):1534–1546. Crossref, MedlineGoogle Scholar
  • 40. Leng S, Yu L, Wang J, Fletcher JG, Mistretta CA, McCollough CH. Noise reduction in spectral CT: reducing dose and breaking the trade-off between image noise and energy bin selection. Med Phys 2011;38(9):4946–4957. Crossref, MedlineGoogle Scholar
  • 41. McCollough CH, Van Lysel MS, Peppler WW, Mistretta CA. A correlated noise reduction algorithm for dual-energy digital subtraction angiography. Med Phys 1989;16(6):873–880. Crossref, MedlineGoogle Scholar
  • 42. Macovski A. Physical problems of computerized tomography. Proc IEEE 1983;71(3):373–378. CrossrefGoogle Scholar
  • 43. Yu Z, Leng S, Li Z, et al. Spectral prior image constrained compressed sensing (spectral PICCS) for photon-counting computed tomography. Phys Med Biol 2016; 61(18):6707–6732. Crossref, MedlineGoogle Scholar
  • 44. Chen GH, Tang J, Leng S. Prior image constrained compressed sensing (PICCS): a method to accurately reconstruct dynamic CT images from highly undersampled projection data sets. Med Phys 2008;35(2):660–663. Crossref, MedlineGoogle Scholar
  • 45. Wang AS, Pelc NJ. Sufficient statistics as a generalization of binning in spectral x-ray imaging. IEEE Trans Med Imaging 2011;30(1):84–93. Crossref, MedlineGoogle Scholar
  • 46. Albrecht MH, Trommer J, Wichmann JL, et al. Comprehensive Comparison of Virtual Monoenergetic and Linearly Blended Reconstruction Techniques in Third-Generation Dual-Source Dual-Energy Computed Tomography Angiography of the Thorax and Abdomen. Invest Radiol 2016;51(9):582–590. Crossref, MedlineGoogle Scholar
  • 47. Hojjati M, Van Hedent S, Rassouli N, et al. Quality of routine diagnostic abdominal images generated from a novel detector-based spectral CT scanner: a technical report on a phantom and clinical study. Abdom Radiol (NY) 2017;42(11):2752–2759. Crossref, MedlineGoogle Scholar
  • 48. Patel BN, Thomas JV, Lockhart ME, Berland LL, Morgan DE. Single-source dual-energy spectral multidetector CT of pancreatic adenocarcinoma: optimization of energy level viewing significantly increases lesion contrast. Clin Radiol 2013;68(2):148–154. Crossref, MedlineGoogle Scholar
  • 49. Albrecht MH, Vogl TJ, Martin SS, et al. Review of Clinical Applications for Virtual Monoenergetic Dual-Energy CT. Radiology 2019;293(2):260–271. LinkGoogle Scholar
  • 50. Rassouli N, Chalian H, Rajiah P, Dhanantwari A, Landeras L. Assessment of 70-keV virtual monoenergetic spectral images in abdominal CT imaging: a comparison study to conventional polychromatic 120-kVp images. Abdom Radiol (NY) 2017;42(10):2579–2586. Crossref, MedlineGoogle Scholar
  • 51. Chalian H, Kalisz K, Rassouli N, Dhanantwari A, Rajiah P. Utility of virtual monoenergetic images derived from a dual-layer detector-based spectral CT in the assessment of aortic anatomy and pathology: a retrospective case control study. Clin Imaging 2018;52:292–301. Crossref, MedlineGoogle Scholar
  • 52. Ghandour A, Sher A, Rassouli N, Dhanantwari A, Rajiah P. Evaluation of Virtual Monoenergetic Images on Pulmonary Vasculature Using the Dual-Layer Detector-Based Spectral Computed Tomography. J Comput Assist Tomogr 2018;42(6):858–865. Crossref, MedlineGoogle Scholar
  • 53. Cavallo AU, Patterson AJ, Thomas R, et al. Low dose contrast CT for transcatheter aortic valve replacement assessment: results from the prospective SPECTACULAR study (spectral CT assessment prior to TAVR). J Cardiovasc Comput Tomogr 2020;14(1):68–74. Crossref, MedlineGoogle Scholar
  • 54. Kalisz K, Halliburton S, Abbara S, et al. Update on Cardiovascular Applications of Multienergy CT. RadioGraphics 2017;37(7):1955–1974. LinkGoogle Scholar
  • 55. Rajiah P, Sundaram M, Subhas N. Dual-energy CT in musculoskeletal imaging: what is the role beyond gout? AJR Am J Roentgenol 2019;213(3):493–505. Crossref, MedlineGoogle Scholar
  • 56. Leng S, Yu L, Fletcher JG, McCollough CH. Maximizing Iodine Contrast-to-Noise Ratios in Abdominal CT Imaging through Use of Energy Domain Noise Reduction and Virtual Monoenergetic Dual-Energy CT. Radiology 2015;276(2):562–570. LinkGoogle Scholar
  • 57. Grant KL, Flohr TG, Krauss B, Sedlmair M, Thomas C, Schmidt B. Assessment of an advanced image-based technique to calculate virtual monoenergetic computed tomographic images from a dual-energy examination to improve contrast-to-noise ratio in examinations using iodinated contrast media. Invest Radiol 2014;49(9):586–592. Crossref, MedlineGoogle Scholar
  • 58. Kalisz K, Rassouli N, Dhanantwari A, Jordan D, Rajiah P. Noise characteristics of virtual monoenergetic images from a novel detector-based spectral CT scanner. Eur J Radiol 2018;98:118–125. Crossref, MedlineGoogle Scholar
  • 59. Uyeda JW, Richardson IJ, Sodickson AD. Making the invisible visible: improving conspicuity of noncalcified gallstones using dual-energy CT. Abdom Radiol (NY) 2017;42(12):2933–2939. Crossref, MedlineGoogle Scholar
  • 60. Morgan DE. Dual-energy CT of the abdomen. Abdom Imaging 2014;39(1):108–134. Crossref, MedlineGoogle Scholar
  • 61. Laukamp KR, Gupta A, Große Hokamp N, et al. Role of spectral-detector CT in reduction of artifacts from contrast media in axillary and subclavian veins: single institution study in 50 patients. Acta Radiol 2020;61(4):450–460. Crossref, MedlineGoogle Scholar
  • 62. Van Hedent S, Große Hokamp N, Kessner R, Gilkeson R, Ros PR, Gupta A. Effect of virtual monoenergetic images from spectral detector computed tomography on coronary calcium blooming. J Comput Assist Tomogr 2018;42(6):912–918. Crossref, MedlineGoogle Scholar
  • 63. Kosmas C, Hojjati M, Young PC, Abedi A, Gholamrezanezhad A, Rajiah P. Dual-layer spectral computerized tomography for metal artifact reduction: small versus large orthopedic devices. Skeletal Radiol 2019;48(12):1981–1990. Crossref, MedlineGoogle Scholar
  • 64. Fulton N, Rajiah P. Abdominal applications of a novel detector-based spectral CT. Curr Probl Diagn Radiol 2018;47(2):110–118. Crossref, MedlineGoogle Scholar
  • 65. Wong WD, Shah S, Murray N, Walstra F, Khosa F, Nicolaou S. Advanced musculoskeletal applications of dual-energy computed tomography. Radiol Clin North Am 2018;56(4):587–600. Crossref, MedlineGoogle Scholar
  • 66. Taasti VT, Muren LP, Jensen K, et al. Comparison of single and dual energy CT for stopping power determination in proton therapy of head and neck cancer. Phys Imaging Radiat Oncol 2018;6:14–19. CrossrefGoogle Scholar
  • 67. https://physics.nist.gov/PhysRefData/XrayMassCoef/ComTab/adipose.html. Accessed March 19, 2020. Google Scholar
  • 68. Machida H, Tanaka I, Fukui R, et al. Dual-energy spectral CT: various clinical vascular applications. RadioGraphics 2016;36(4):1215–1232. LinkGoogle Scholar
  • 69. Wang B, Gao Z, Zou Q, Li L. Quantitative diagnosis of fatty liver with dual-energy CT: an experimental study in rabbits. Acta Radiol 2003;44(1):92–97. Crossref, MedlineGoogle Scholar
  • 70. Potter CA, Sodickson AD. Dual-energy CT in emergency neuroimaging: added value and novel applications. RadioGraphics 2016;36(7):2186–2198. LinkGoogle Scholar
  • 71. Kay FU, Oz OK, Abbara S, Mortani Barbosa EJ Jr, Agarwal PP, Rajiah P. Translation of Quantitative Imaging Biomarkers into Clinical Chest CT. RadioGraphics 2019;39(4):957–976. LinkGoogle Scholar
  • 72. Fuld MK, Halaweish AF, Haynes SE, Divekar AA, Guo J, Hoffman EA. Pulmonary perfused blood volume with dual-energy CT as surrogate for pulmonary perfusion assessed with dynamic multidetector CT. Radiology 2013;267(3):747–756. LinkGoogle Scholar
  • 73. Jacobsen MC, Schellingerhout D, Wood CA, et al. Intermanufacturer comparison of dual-energy CT iodine quantification and monochromatic attenuation: a phantom study. Radiology 2018;287(1):224–234. LinkGoogle Scholar
  • 74. Leng S, Zhou W, Yu Z, et al. Spectral performance of a whole-body research photon counting detector CT: quantitative accuracy in derived image sets. Phys Med Biol 2017;62(17):7216–7232. Crossref, MedlineGoogle Scholar
  • 75. Rajiah P, Tanabe Y, Partovi S, Moore A. State of the art: utility of multi-energy CT in the evaluation of pulmonary vasculature. Int J Cardiovasc Imaging 2019;35(8):1509–1524. Crossref, MedlineGoogle Scholar
  • 76. Ameli-Renani S, Rahman F, Nair A, et al. Dual-energy CT for imaging of pulmonary hypertension: challenges and opportunities. RadioGraphics 2014;34(7):1769–1790. LinkGoogle Scholar
  • 77. Ohta Y, Kitao S, Yunaga H, et al. Myocardial Delayed Enhancement CT for the Evaluation of Heart Failure: Comparison to MRI. Radiology 2018;288(3):682–691. LinkGoogle Scholar
  • 78. Kaza RK, Caoili EM, Cohan RH, Platt JF. Distinguishing enhancing from nonenhancing renal lesions with fast kilovoltage-switching dual-energy CT. AJR Am J Roentgenol 2011;197(6):1375–1381. Crossref, MedlineGoogle Scholar
  • 79. Ascenti G, Mileto A, Krauss B, et al. Distinguishing enhancing from nonenhancing renal masses with dual-source dual-energy CT: iodine quantification versus standard enhancement measurements. Eur Radiol 2013;23(8):2288–2295. Crossref, MedlineGoogle Scholar
  • 80. Hur J, Kim YJ, Lee HJ, et al. Cardioembolic stroke: dual-energy cardiac CT for differentiation of left atrial appendage thrombus and circulatory stasis. Radiology 2012;263(3):688–695. LinkGoogle Scholar
  • 81. Lee SH, Hur J, Kim YJ, Lee HJ, Hong YJ, Choi BW. Additional value of dual-energy CT to differentiate between benign and malignant mediastinal tumors: an initial experience. Eur J Radiol 2013;82(11):2043–2049. Crossref, MedlineGoogle Scholar
  • 82. Ascenti G, Sofia C, Mazziotti S, et al. Dual-energy CT with iodine quantification in distinguishing between bland and neoplastic portal vein thrombosis in patients with hepatocellular carcinoma. Clin Radiol 2016;71(9):938.e1–938.e9. Crossref, MedlineGoogle Scholar
  • 83. Schmid-Bindert G, Henzler T, Chu TQ, et al. Functional imaging of lung cancer using dual energy CT: how does iodine related attenuation correlate with standardized uptake value of 18FDG-PET-CT? Eur Radiol 2012;22(1):93–103. Crossref, MedlineGoogle Scholar
  • 84. Meyer M, Hohenberger P, Apfaltrer P, et al. CT-based response assessment of advanced gastrointestinal stromal tumor: dual energy CT provides a more predictive imaging biomarker of clinical benefit than RECIST or Choi criteria. Eur J Radiol 2013;82(6):923–928. Crossref, MedlineGoogle Scholar
  • 85. McNamara MM, Little MD, Alexander LF, Carroll LV, Beasley TM, Morgan DE. Multireader evaluation of lesion conspicuity in small pancreatic adenocarcinomas: complimentary value of iodine material density and low keV simulated monoenergetic images using multiphasic rapid kVp-switching dual energy CT. Abdom Imaging 2015;40(5):1230–1240. Crossref, MedlineGoogle Scholar
  • 86. Elbanna KY, Mohammed MF, Chahal T, et al. Dual-Energy CT in Differentiating Nonperforated Gangrenous Appendicitis From Uncomplicated Appendicitis. AJR Am J Roentgenol 2018;211(4):776–782. Crossref, MedlineGoogle Scholar
  • 87. Ananthakrishnan L, Rajiah P, Ahn R, et al. Spectral detector CT-derived virtual non-contrast images: comparison of attenuation values with unenhanced CT. Abdom Radiol (NY) 2017;42(3):702–709. Crossref, MedlineGoogle Scholar
  • 88. Chandarana H, Godoy MCB, Vlahos I, et al. Abdominal aorta: evaluation with dual-source dual-energy multidetector CT after endovascular repair of aneurysms—initial observations. Radiology 2008;249(2):692–700. LinkGoogle Scholar
  • 89. Stolzmann P, Frauenfelder T, Pfammatter T, et al. Endoleaks after endovascular abdominal aortic aneurysm repair: detection with dual-energy dual-source CT. Radiology 2008;249(2):682–691. LinkGoogle Scholar
  • 90. Maturen KE, Kleaveland PA, Kaza RK, et al. Aortic endograft surveillance: use of fast-switch kVp dual-energy computed tomography with virtual noncontrast imaging. J Comput Assist Tomogr 2011;35(6):742–746. Crossref, MedlineGoogle Scholar
  • 91. Patel AA, Sutphin PD, Xi Y, Abbara S, Kalva SP. Arterial phase CTA replacement by a virtual arterial phase reconstruction from a venous phase CTA: preliminary results using detector-based spectral CT. Cardiovasc Intervent Radiol 2019;42(2):250–259. Crossref, MedlineGoogle Scholar
  • 92. Trabzonlu TA, Mozaffary A, Kim D, Yaghmai V. Dual-energy CT evaluation of gastrointestinal bleeding. Abdom Radiol (NY) 2020;45(1):1–14. Crossref, MedlineGoogle Scholar
  • 93. Tijssen MP, Hofman PA, Stadler AA, et al. The role of dual energy CT in differentiating between brain haemorrhage and contrast medium after mechanical revascularisation in acute ischaemic stroke. Eur Radiol 2014;24(4):834–840. Crossref, MedlineGoogle Scholar
  • 94. Ho LM, Marin D, Neville AM, et al. Characterization of adrenal nodules with dual-energy CT: can virtual unenhanced attenuation values replace true unenhanced attenuation values? AJR Am J Roentgenol 2012;198(4):840–845. Crossref, MedlineGoogle Scholar
  • 95. Yamada Y, Jinzaki M, Okamura T, et al. Feasibility of coronary artery calcium scoring on virtual unenhanced images derived from single-source fast kVp-switching dual-energy coronary CT angiography. J Cardiovasc Comput Tomogr 2014;8(5):391–400. Crossref, MedlineGoogle Scholar
  • 96. Primak AN, Fletcher JG, Vrtiska TJ, et al. Noninvasive differentiation of uric acid versus non-uric acid kidney stones using dual-energy CT. Acad Radiol 2007;14(12):1441–1447. Crossref, MedlineGoogle Scholar
  • 97. Kambadakone AR, Eisner BH, Catalano OA, Sahani DV. New and evolving concepts in the imaging and management of urolithiasis: urologists’ perspective. RadioGraphics 2010;30(3):603–623. LinkGoogle Scholar
  • 98. Kulkarni NM, Eisner BH, Pinho DF, Joshi MC, Kambadakone AR, Sahani DV. Determination of renal stone composition in phantom and patients using single-source dual-energy computed tomography. J Comput Assist Tomogr 2013;37(1):37–45. Crossref, MedlineGoogle Scholar
  • 99. Gamala M, Jacobs JWG, van Laar JM. The diagnostic performance of dual energy CT for diagnosing gout: a systematic literature review and meta-analysis. Rheumatology (Oxford) 2019;58(12):2117–2121. Crossref, MedlineGoogle Scholar
  • 100. Wiggins WF, Potter CA, Sodickson AD. Dual-energy CT to differentiate small foci of intracranial hemorrhage from calcium. Radiology 2020;294(1):129–138. LinkGoogle Scholar
  • 101. Vlahos I, Chung R, Nair A, Morgan R. Dual-energy CT: vascular applications. AJR Am J Roentgenol 2012;199(5 suppl):S87–S97. Crossref, MedlineGoogle Scholar
  • 102. Fuentes-Orrego JM, Pinho D, Kulkarni NM, Agrawal M, Ghoshhajra BB, Sahani DV. New and evolving concepts in CT for abdominal vascular imaging. RadioGraphics 2014;34(5):1363–1384. LinkGoogle Scholar
  • 103. Korn A, Bender B, Thomas C, et al. Dual energy CTA of the carotid bifurcation: advantage of plaque subtraction for assessment of grade of the stenosis and morphology. Eur J Radiol 2011;80(2):e120–e125. Crossref, MedlineGoogle Scholar
  • 104. Meyer BC, Werncke T, Hopfenmüller W, Raatschen HJ, Wolf KJ, Albrecht T. Dual energy CT of peripheral arteries: effect of automatic bone and plaque removal on image quality and grading of stenoses. Eur J Radiol 2008;68(3):414–422. Crossref, MedlineGoogle Scholar
  • 105. Brockmann C, Jochum S, Sadick M, et al. Dual-energy CT angiography in peripheral arterial occlusive disease. Cardiovasc Intervent Radiol 2009;32(4):630–637. Crossref, MedlineGoogle Scholar
  • 106. Sommer WH, Johnson TR, Becker CR, et al. The value of dual-energy bone removal in maximum intensity projections of lower extremity computed tomography angiography. Invest Radiol 2009;44(5):285–292. Crossref, MedlineGoogle Scholar
  • 107. Mannil M, Ramachandran J, Vittoria de Martini I, et al. Modified Dual-Energy Algorithm for Calcified Plaque Removal: Evaluation in Carotid Computed Tomography Angiography and Comparison With Digital Subtraction Angiography. Invest Radiol 2017;52(11):680–685. Crossref, MedlineGoogle Scholar
  • 108. De Santis D, Jin KN, Schoepf UJ, et al. Heavily Calcified Coronary Arteries: Advanced Calcium Subtraction Improves Luminal Visualization and Diagnostic Confidence in Dual-Energy Coronary Computed Tomography Angiography. Invest Radiol 2018;53(2):103–109. Crossref, MedlineGoogle Scholar
  • 109. De Santis D, De Cecco CN, Schoepf UJ, et al. Modified calcium subtraction in dual-energy CT angiography of the lower extremity runoff: impact on diagnostic accuracy for stenosis detection. Eur Radiol 2019;29(9):4783–4793. Crossref, MedlineGoogle Scholar
  • 110. Ali IT, Wong WD, Liang T, et al. Clinical Utility of Dual-Energy CT Analysis of Bone Marrow Edema in Acute Wrist Fractures. AJR Am J Roentgenol 2018;210(4):842–847. Crossref, MedlineGoogle Scholar
  • 111. Guggenberger R, Gnannt R, Hodler J, et al. Diagnostic performance of dual-energy CT for the detection of traumatic bone marrow lesions in the ankle: comparison with MR imaging. Radiology 2012;264(1):164–173. LinkGoogle Scholar
  • 112. Mallinson PI, Coupal TM, McLaughlin PD, Nicolaou S, Munk PL, Ouellette HA. Dual-energy CT for the musculoskeletal system. Radiology 2016;281(3):690–707. LinkGoogle Scholar
  • 113. Kosmala A, Weng AM, Heidemeier A, et al. Multiple Myeloma and Dual-Energy CT: Diagnostic Accuracy of Virtual Noncalcium Technique for Detection of Bone Marrow Infiltration of the Spine and Pelvis. Radiology 2018;286(1):205–213. LinkGoogle Scholar
  • 114. Wu H, Zhang G, Shi L, et al. Axial Spondyloarthritis: Dual-Energy Virtual Noncalcium CT in the Detection of Bone Marrow Edema in the Sacroiliac Joints. Radiology 2019;290(1):157–164. LinkGoogle Scholar
  • 115. Burke MC, Garg A, Youngner JM, Deshmukh SD, Omar IM. Initial experience with dual-energy computed tomography-guided bone biopsies of bone lesions that are occult on monoenergetic CT. Skeletal Radiol 2019;48(4):605–613. Crossref, MedlineGoogle Scholar
  • 116. Werner S, Krauss B, Haberland U, et al. Dual-energy CT for liver iron quantification in patients with haematological disorders. Eur Radiol 2019;29(6):2868–2877. Crossref, MedlineGoogle Scholar
  • 117. Tsai YS, Chen JS, Wang CK, et al. Quantitative assessment of iron in heart and liver phantoms using dual-energy computed tomography. Exp Ther Med 2014;8(3):907–912. Crossref, MedlineGoogle Scholar
  • 118. Nicolaou S, Liang T, Murphy DT, Korzan JR, Ouellette H, Munk P. Dual-energy CT: a promising new technique for assessment of the musculoskeletal system. AJR Am J Roentgenol 2012;199(5 suppl):S78–S86. Crossref, MedlineGoogle Scholar
  • 119. Becce F, Federau C, Letovanec I, Grandjean A, So AK, Rüdiger HA. Dual-energy computed tomography molecular imaging of pigmented villonodular synovitis. Rheumatology (Oxford) 2015;54(3):457. Crossref, MedlineGoogle Scholar
  • 120. Fischer MA, Gnannt R, Raptis D, et al. Quantification of liver fat in the presence of iron and iodine: an ex-vivo dual-energy CT study. Invest Radiol 2011;46(6):351–358. Crossref, MedlineGoogle Scholar
  • 121. Patel BN, Kumbla RA, Berland LL, Fineberg NS, Morgan DE. Material density hepatic steatosis quantification on intravenous contrast-enhanced rapid kilovolt (peak)-switching single-source dual-energy computed tomography. J Comput Assist Tomogr 2013;37(6):904–910. Crossref, MedlineGoogle Scholar
  • 122. Mileto A, Nelson RC, Marin D, Roy Choudhury K, Ho LM. Dual-energy multidetector CT for the characterization of incidental adrenal nodules: diagnostic performance of contrast-enhanced material density analysis. Radiology 2015;274(2):445–454. LinkGoogle Scholar
  • 123. Cai W, Kim SH, Lee JG, Yoshida H. Informatics in radiology: dual-energy electronic cleansing for fecal-tagging CT colonography. RadioGraphics 2013;33(3):891–912. LinkGoogle Scholar
  • 124. Mohammed MF, Marais O, Min A, et al. Unenhanced dual-energy computed tomography: visualization of brain edema. Invest Radiol 2018;53(2):63–69. Crossref, MedlineGoogle Scholar
  • 125. Van Hedent S, Su KH, Jordan DW, et al. Improving bone mineral density assessment using spectral detector CT. J Clin Densitom 2019;22(3):374–381. Crossref, MedlineGoogle Scholar
  • 126. Mussmann B, Overgaard S, Torfing T, et al. Intra- and inter-observer agreement and reliability of bone mineral density measurements around acetabular cup: a porcine ex-vivo study using single- and dual-energy computed tomography. Acta Radiol Open 2017;6(7):2058460117719746. MedlineGoogle Scholar
  • 127. Deng K, Sun C, Liu C, Ma R. Initial experience with visualizing hand and foot tendons by dual-energy computed tomography. Clin Imaging 2009;33(5):384–389. Crossref, MedlineGoogle Scholar
  • 128. Murray N, Le M, Ebrahimzadeh O, et al. Imaging the spine with dual-energy CT. Curr Radiol Rep 2017;5(9):44. CrossrefGoogle Scholar
  • 129. Kong X, Sheng HX, Lu GM, et al. Xenon-enhanced dual-energy CT lung ventilation imaging: techniques and clinical applications. AJR Am J Roentgenol 2014;202(2):309–317. Crossref, MedlineGoogle Scholar
  • 130. Rotzinger DC, Si-Mohamed SA, Shapira N, Douek PC, Meuli RA, Boussel L. “Dark-blood” dual-energy computed tomography angiography for thoracic aortic wall imaging. Eur Radiol 2020;30(1):425–431. Crossref, MedlineGoogle Scholar
  • 131. Ashton JR, Castle KD, Qi Y, Kirsch DG, West JL, Badea CT. Dual-energy CT imaging of tumor liposome delivery after gold nanoparticle-augmented radiation therapy. Theranostics 2018;8(7):1782–1797. Crossref, MedlineGoogle Scholar
  • 132. Kalender WA, Perman WH, Vetter JR, Klotz E. Evaluation of a prototype dual-energy computed tomographic apparatus. I. Phantom studies. Med Phys 1986;13(3):334–339. Crossref, MedlineGoogle Scholar
  • 133. Bazalova M, Carrier JF, Beaulieu L, Verhaegen F. Dual-energy CT-based material extraction for tissue segmentation in Monte Carlo dose calculations. Phys Med Biol 2008;53(9):2439–2456. Crossref, MedlineGoogle Scholar
  • 134. Pascart T, Norberciak L, Legrand J, Becce F, Budzik JF. Dual-energy computed tomography in calcium pyrophosphate deposition: initial clinical experience. Osteoarthritis Cartilage 2019;27(9):1309–1314. Crossref, MedlineGoogle Scholar
  • 135. Hudobivnik N, Schwarz F, Johnson T, et al. Comparison of proton therapy treatment planning for head tumors with a pencil beam algorithm on dual and single energy CT images. Med Phys 2016;43(1):495–504. Crossref, MedlineGoogle Scholar
  • 136. Soesbe TC, Lewis MA, Xi Y, et al. A Technique to Identify Isoattenuating Gallstones with Dual-Layer Spectral CT: An ex Vivo Phantom Study. Radiology 2019;292(2):400–406. LinkGoogle Scholar
  • 137. Jepperson MA, Cernigliaro JG, Ibrahim SH, Morin RL, Haley WE, Thiel DD. In vivo comparison of radiation exposure of dual-energy CT versus low-dose CT versus standard CT for imaging urinary calculi. J Endourol 2015;29(2):141–146. Crossref, MedlineGoogle Scholar
  • 138. Duan X, Ananthakrishnan L, Guild JB, Xi Y, Rajiah P. Radiation doses and image quality of abdominal CT scans at different patient sizes using spectral detector CT scanner: a phantom and clinical study. Abdom Radiol (NY) 2019. 10.1007/s00261-019-02247-1. Published October 5, 2019. Crossref, MedlineGoogle Scholar
  • 139. Foley WD, Shuman WP, Siegel MJ, et al. White paper of the Society of Computed Body Tomography and Magnetic Resonance on dual-energy CT. II. Radiation dose and iodine sensitivity. J Comput Assist Tomogr 2016;40(6):846–850. Crossref, MedlineGoogle Scholar
  • 140. Wortman JR, Bunch PM, Fulwadhva UP, Bonci GA, Sodickson AD. Dual-Energy CT of Incidental Findings in the Abdomen: Can We Reduce the Need for Follow-Up Imaging? AJR Am J Roentgenol 2016;207(4):W58–W68. Crossref, MedlineGoogle Scholar

Article History

Received: Mar 14 2020
Revision requested: Apr 29 2020
Revision received: June 9 2020
Accepted: June 17 2020
Published online: Aug 21 2020
Published in print: Sept 2020