Detector-based spectral CT with a novel dual-layer technology: principles and applications
- 3.3k Downloads
Detector-based spectral computed tomography is a novel dual-energy CT technology that employs two layers of detectors to simultaneously collect low- and high-energy data in all patients using standard CT protocols. In addition to the conventional polyenergetic images created for each patient, projection-space decomposition is used to generate spectral basis images (photoelectric and Compton scatter) for creating multiple spectral images, including material decomposition (iodine-only, virtual non-contrast, effective atomic number) and virtual monoenergetic images, on-demand according to clinical need. These images are useful in multiple clinical applications, including- improving vascular contrast, improving lesion conspicuity, decreasing artefacts, material characterisation and reducing radiation dose. In this article, we discuss the principles of this novel technology and also illustrate the common clinical applications.
• The top and bottom layers of dual-layer CT absorb low- and high-energy photons, respectively.
• Multiple spectral images are generated by projection-space decomposition.
• Spectral images can be generated in all patients scanned in this scanner.
KeywordsCT Dual-energy Spectral Contrast Cardiac
Conventional computed tomography (CT) scanners utilising a single X-ray spectrum are able to discriminate several tissues based on their attenuation (Hounsfield units, HU), but the overlap in attenuation of many tissues limits discrimination. Tissues have a unique distribution of attenuation values as a function of X-ray energy and spectral CT (dual-energy/multi-energy CT) can further distinguish them by exploiting their energy-dependent attenuation properties. Data collected simultaneously from two energy levels can be used to determine the Compton scatter and photoelectric components of x-ray attenuation which have different energy dependencies. These components provide additional information about tissue density and atomic number that can be exploited to separate tissues with similar attenuation in a conventional image. Spectral CT technologies so far have been source-based, including dual-source, rapid kilovoltage (kVp) switching and dual-spin technologies .
Detector-based spectral CT (SDCT; Philips Healthcare, Cleveland, OH, USA) offers a novel approach to dual-energy imaging where the spectral separation occurs at the level of the detectors. In this article, we describe data acquisition and image generation with SDCT and the clinical applications of this technology.
Data acquisition hardware
Solving for the weights α i and α 2 for each ray projection acquired at two different energies decomposes the raw data into photoelectric and Compton scatter basis components .
Types of images from the detector-based spectral CT scanner
Mechanism of generation
Conventional (Polyenergetic/routine diagnostic)
Data from both layers considered as a single detector
Routine diagnostic use for all cases
Iodine density (iodine map)
Material decomposition with pixels representing iodine
Visualisation and quantification of iodine in vessels and organs of interest
Material decomposition and removal of iodine containing pixels
-Characterisation of lesions such as renal cysts/masses, adrenal nodules, lung nodules, etc.
-Radiation dose saving by eliminating need for true non contrast
Uric acid pair
Material decomposition; depiction of pixels containing uric acid
Urinary calculus characterisation
Effective atomic number
Material decomposition; colour coding depending on atomic number
Linear combination of basis pair images (40–200 keV)
-Low monoenergetic: enhanced vascular contrast
-High monoenergetic: decreased artefacts
Linear combination of basis pair, with attenuation values equivalent to conventional images
Higher image quality, with lower noise
Conventional images (routine diagnostic/polyenergetic images) -
These images are analogous to the images obtained from a single-energy scanner, and are utilised for routine diagnostic purposes. For every scan, the two detector layers are considered as a single detector and high- and low-energy data are combined. Filtered back projection or iterative reconstruction algorithms are used to reconstruct the combined raw data and create conventional images equivalent to those from a single-energy scanner. The image quality of these images have been shown comparable to images obtained from a single-energy scanner .
Spectral images -
Iodine density images- Iodine-containing pixels are assigned values equal to the concentration of the iodine in each pixel, expressed in milligrammes/ml (Fig. 2a). Pixels containing no iodine appear dark. Iodine-only images permit quantification of iodine in vessels and organs.
Virtual non-contrast images- Images where the contributions from iodine are removed, resulting in non-contrast-enhanced HU values of the tissue contained within the pixels (Fig. 2b). The resulting image mimics a true non-contrast-enhanced (pre-contrast) image.
Uric acid pair images- A uric acid image displays only uric acid pixels with original HU values while all others appear dark. The uric acid-removed image is a complement to the uric acid image These images are utilised in the evaluation of urinary calculi composition and gout.
Z-effective images- are colour-coded based on the effective atomic number of tissues (Fig. 2c). The coefficients of the photoelectric and Compton scatter components (α1 and α2, respectively) computed during the spectral decomposition process are functions of the spatial distribution of tissue [i.e., α1(x, y) and α2(x, y)]. The ratio of these coefficients is proportional to the effective atomic number, Z:
Virtual monoenergetic images (VMI) mimic images generated from application of a monoenergetic beam at a single kiloelectron voltage (keV) level. VMI are generated between 40 and 200 keV (Fig. 2d) by a linear combination of basis pair images. The attenuation, μ, at any energy can be determined from
Equivalent virtual monoenergetic images mimic those generated by application of a monoenergetic beam at a single keV with attenuation values equivalent to conventional images but with lower artefacts and noise. For SDCT, the equivalent VMI levels are 70 keV for body, 66 keV for head and 64 keV for extremities .
A major advantage of SDCT is elimination of the requirement to prospectively select patients who need dual energy CT. All other currently available commercial solutions—dual-source CT, rapid tube potential switching and dual-spin technology—require selection of a special dual-energy protocol prior to scanning . Spectral data are available for all patients imaged with SDCT without a change in clinical workflow, permitting evaluation of incidentally discovered findings and reduction of artifacts. The complete spatial and temporal alignment is ideal for cardiovascular studies and also allows projection-space decomposition with more accurate artefact correction and better noise removal in VMI compared to image space decomposition . All radiation dose-reduction strategies available with conventional CT can be employed with SDCT. There are no restrictions related to field of view, gantry rotation time or cross-scatter.
Sufficient spectral separation requires a tube potential of at least 120 kVp. Therefore, radiation dose reduction for smaller patients relies on lowering the tube current. Due to higher kVp, contrast is lowered in conventional images, but low-energy VMI may be helpful in such instances. The z coverage of the currently available scanner configuration is 4 cm, which makes some dynamic studies, such as myocardial perfusion, challenging.
Enhanced visualisation of vascular contrastThe attenuation of iodine is higher at lower energies due to the increased photoelectric attenuation at energies approaching the K-edge of iodine (33.2 keV). VMI reconstructions at lower energies (40–70 keV) can be used to enhance the visualisation of intravascular contrast, which can be clinically used in several settings [10, 11]. Suboptimal enhanced vascular studies are frequently encountered in imaging, either due to technical or patient factors. Usually, a suboptimal study requires a repeat study with additional bolus of contrast. However, with SDCT, due to the availability of low-energy VMIs on demand, suboptimal vascular studies can be salvaged, thus obviating the need for additional contrast injection or alternate imaging tests (Fig. 3). Although other currently available DECT scanners also have VMI capabilities, the dual-energy mode is not always switched on in these scanners and hence it may not always be available in a given patient. Vascular studies can be performed with low contrast dose, especially in patients with severe renal dysfunction and the contrast signal can be boosted by using VMI (Fig. 4). In addition, CT angiographic quality studies can be generated from routine contrast-enhanced CT scans, which may help in evaluation of incidentally seen vascular lesions. VMI, particularly at low energy or equivalent energy scans, can be used in improving the conspicuity of several lesions (Fig. 5). Hypervascular lesions may be more prominent at low-keV VMI due to improved iodine visualisation. Hypovascular lesions such as pancreatic adenocarcinomas are also well seen in low- or equivalent-keV VMI (Fig 5) .
Artefact reductionSeveral artefacts are commonly seen in CT, including metallic artefacts, beam hardening and calcium blooming . Metals seen in strategic locations can limit the diagnostic confidence by obscuring vital structures. If the presence of metal is known ahead of time, metal reconstruction algorithms can be used. Retrospectively generated high-energy VMI from SDCT can be used to reduce/eliminate incidentally encountered metallic artefacts, particularly metals with low atomic number such as stainless steel and aluminium (Fig 6). Artefacts from high atomic number metals such as embolisation coils are improved with iodine maps. Beam hardening artefacts which are caused by the polyenergetic nature of the x-ray energy can also be minimised/eliminated by using the high-energy VMI (Fig 7). Calcium blooming is an important artefact in vascular imaging, which causes overestimation of stenosis and inappropriate classification of the lesion. This can also be reduced by using high-energy VMI . This ability of the SDCT to minimise artefacts in critical areas, particularly retrospectively, may improve the diagnostic confidence.
Material decompositionZ-effective images, iodine maps and calcium-uric acid images are currently used for the evaluation of urinary calculi, gout, lesion/plaque characterisation, perfusion of organs/tumour, tumour response to therapy, bone removal for CTA head, and virtual colonoscopy without bowel preparation. Increased utility may be possible with SDCT because of the opportunity to access spectral data for incidental findings. Characterisation of urinary calculi is important since management is different for uric acid and calcium calculi. Urinary calculi composition can be characterised by the z-effective images, and uric acid pair images (Fig 8) [14, 15]. Similarly, gout crystals can also be detected in joints. Iodine maps can be used to evaluate perfusion in several organs, including heart and lungs, the former in the diagnosis of myocardial ischemia and the latter in the evaluation of small pulmonary embolism (both acute and chronic) (Fig 9). Lesions involving the adrenal, kidney, liver and lung can be characterised with SDCT by a combination of VNC and iodine map, without the need for additional tests such as CT and MRI and possible additional radiation. For example, using VNC, the attenuation of an adrenal lesion from a contrast-enhanced image can be obtained and a lipid-rich adenoma can be diagnosed if attenuation is <10 HU . A hyperattenuating renal lesion in contrast CT is likely to be a complicated cyst if there is high attenuation in VNC and no iodine uptake in the iodine map, and likely to be a solid tumour if there is no high attenuation in VNC, but there is iodine uptake (Fig 10). Some of these lesions usually require multi-phasic images for adequate characterisation.
Radiation dose reductionThe attenuation values of the VNC images from SDCT have been shown to be similar to true unenhanced images in all tissues except for subcutaneous fat . Hence, VNC images may eliminate the need for true non-contrast scans in multi-phasic studies, thus reducing radiation dose (Fig. 11). Radiation dose savings of up to 40% has been shown by eliminating the true non-contrast phase and up to 64% by eliminating the arterial phase [18, 19]. The ability to salvage suboptimal enhanced studies and characterise incidental findings (e.g., adrenal, renal lesions) may also obviate the need for an additional CT scan, thereby reducing radiation dose.
SDCT is a unique technology utilising dual-layer detectors which provides spectral data on demand for all patients. There are several applications of this technology in several organ systems. The major applications of this technology are to improve the vascular contrast using low-energy VMI images; decrease several artefacts using high-energy VMI images; material composition of several tissues and lesions using an iodine map, VNC and effective-z images; and radiation dose savings using VNC images instead of true non-contrast images as well as saving additional CT examinations for lesion characterisation.
Compliance with ethical standards
Institutional research support from Philips Healthcare, not related to this paper.
Prabhakar Rajiah had received honoraria from Philips in the past (> 2 years).
Amar Dhanantwari is an employee of Philips.
- 3.Pelgrim GJ, van Hamservelt RW, Willemink MJ, Schmidt BT, Flohr T, Schillam A et al (2017) Accuracy of iodine quantification using dual energy CT in latest generation dual source and dual layer CT. Eur Radiol. https://doi.org/10.1007/s00330-017-472-9.PMID-28168368
- 4.Kalendar WA, Perman WH, Vetter JR, Kotz E (1986) Evaluation of a prototype dual energy computed tomography apparatus. I. Phantom studies. Med Phys 1986 13(3):334–339Google Scholar
- 6.Hojjati M, Van Hedent S, Rassouli N, Tasuoka C, Jordan D, Dhanantwari A, 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 2017 May 10. https://doi.org/10.1007/s00261-017-1170-z
- 7.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 poly- chromatic 120-kVp images. Abdominal Radiology 2017 Apr 18. https://doi.org/10.1007/s00261-017-1151-2
- 15.Kaza RK, Ananthakrishnan L, Kambadakone A, Platt JF (2017) Update of dual-energy CT applications in the genitourinary tract. AJR Am J RoerngenolGoogle Scholar
Open Access This article is distributed under the terms of the Creative Commons Attribution 4.0 International License (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made.