Potential clinical usefulness and limitation of fast kv switch Dual energy CT

Transcription

1 Potential clinical usefulness and limitation of fast kv switch Dual energy CT Poster No.: C-1859 Congress: ECR 2011 Type: Educational Exhibit Authors: M. Jinzaki, Y. Tanami, H. Sugiura, M. inoue, S. Koga, Y Yamada, S. Kuribayashi ; 8582/JP, Tokyo/JP, shinjuku-ku, 4 Tokyo/JP, Tkyo/JP Keywords: Calcifications / Calculi, CT-Quantitative, CT, Cardiovascular system, Abdomen DOI: /ecr2011/C-1859 Any information contained in this pdf file is automatically generated from digital material submitted to EPOS by third parties in the form of scientific presentations. References to any names, marks, products, or services of third parties or hypertext links to thirdparty sites or information are provided solely as a convenience to you and do not in any way constitute or imply ECR's endorsement, sponsorship or recommendation of the third party, information, product or service. ECR is not responsible for the content of these pages and does not make any representations regarding the content or accuracy of material in this file. As per copyright regulations, any unauthorised use of the material or parts thereof as well as commercial reproduction or multiple distribution by any traditional or electronically based reproduction/publication method ist strictly prohibited. You agree to defend, indemnify, and hold ECR harmless from and against any and all claims, damages, costs, and expenses, including attorneys' fees, arising from or related to your use of these pages. Please note: Links to movies, ppt slideshows and any other multimedia files are not available in the pdf version of presentations. Page 1 of 37

2 Learning objectives To review the difference between fast kv switch dual energy CT system and dual source CT system. To understand the two component of dual energy CT: material decomposition and virtual monochromatic spectral imaging. To understand the clinical usefulness of fast kv switch dual energy CT imaging. Background 1) Principles of Dual Energy Dual Energy image is composed of material decomposition and virtual monochromatic image. # Material decomposition The differentiation of material in computed tomography (CT) is based on their X-ray attenuation as quantified in Hounsfield Units. Attenuation is caused by absorption and scattering of radiation by the material under investigation. The two main mechanisms responsible for these effects in the photon energy range used in CT are the Compton scatter and the photo effect (Fig. 1). The total cross section of the Compton effect is almost independent of photon energy, whereas the total cross section of the photo effect is strongly energy-dependent. Page 2 of 37

3 Fig.: 1: Photoelectric effect and Compton scattering References: M. Jinzaki; Diagnostic Radiology, Keio University, 8582, JAPAN CT numbers do not vary very much with beam energy for soft tissues, but do vary noticeably for high z materials. Thus, materials can be differentiated further by applying different X-ray spectra (Fig. 2) and analyzing the differences in attenuation. Fig.: 2: CT Number of Water, Iodine and Calcium on different kvp The CT attenuation of a material is different in 140kV and 80kV. Each material has its own characteristic changing pattern in 140 and 80kV. Thus, the change in measurement of CT attenuation between 80 and 140kV will enable differentiating iodine and calcium. References: M. Jinzaki; Diagnostic Radiology, Keio University, 8582, JAPAN This process is referred to as material decomposition or material separation. This works especially well in materials with large atomic numbers due to the photo effect. One of these materials is iodine, which is commonly used in CT as a contrast material and is generally known to have stronger enhancement at low tube voltage settings. This effect makes it beneficial to use the spectral information to differentiate iodine from other materials that do not show this behavior. Page 3 of 37

4 # Virtual monochromatic spectrum image Virtual monochromatic (VMS) image is also an important component of dual energy CT as well as material decomposition. VMS images may be synthesized from the material decomposition images. VMS image depicts how the imaged object would look if the X-ray source produced only X-ray photons at a single energy (Fig. 3). In a CT system, an object is usually scanned with 120-kVp X-rays possessing a polychromatic spectrum (Fig. 3). Fig.: 3: Spectrum of X-ray 120-kVp X-rays possess a polychromatic spectrum, while virtual monochromatic image (VMSI) is a like as if the X-ray source produced only Xray photons at a single energy References: M. Jinzaki; Diagnostic Radiology, Keio University, 8582, JAPAN As the X-ray beam passes through the patient, the lower energy, or softer X-rays are filtered out resulting in beam hardening. Beam hardening can shift the HU value for a material within the scan field of view (SFOV) of a patient or between patients. Current CT systems correct this phenomenon, called beam hardening effective energy shift, by using calibration data measured in specific phantoms and calculated by the specific function during the image reconstruction process. If beam hardening correction is not sufficient, nonlinear artifacts such as shading and dark artifacts occur, which might mimic pathologies and lead to misdiagnosis (Fig. 4). These artifacts are recognized as severe problems in CT imaging. Page 4 of 37

5 Fig.: 4: Beam hardening artifact (BHA) left: BHA between aorta and left ventricle, which may cause misdiagnosis of cardiac ischemia. right: well-known BHA in posterior fossa References: M. Jinzaki; Diagnostic Radiology, Keio University, 8582, JAPAN If projection-based beam hardening correction through two different materials can be performed by using two different kvp data for the same projection angle, the rigorous beam hardening correction will be available, which enables accurate material density images and mass attenuation coefficients to be obtained in the material decomposition calculation process. VMS image is reconstructed from a pair of accurate material density images and mass attenuation coefficients. 2) Three kinds of Dual Energy System Early work in the 1970s and 1980s demonstrated that dual-energy technology improved tissue characterization; however, its utility was limited because of noise in the lowkilovoltage images and the amount of time required for data acquisition, which led to misregistration. Newer CT technologies that allow for more rapid data acquisition have sparked renewed interest in dual-energy applications. There are three kinds of dual energy systems. The first one is rotate-rotate system, in which the CT gantry fully rotates once to acquire data at the first kilovolt peak setting, then rotates again almost immediately to acquire data at the second kilovolt peak setting (Fig. 5a). The nonsimultaneous nature of data acquisition and low speed are disadvantages of this method. Page 5 of 37

6 Fig.: 5a: Rotate-rotate dual energy CT system References: M. Jinzaki; Diagnostic Radiology, Keio University, 8582, JAPAN The second one is dual-source CT system (Definition or Definition flush; Siemens Medical Solutions), in which the two orthogonally mounted detectors and tube arrays can operate simultaneously and can be set to different tube potentials (Fig. 5b). Page 6 of 37

7 Fig.: 5b: Dual-source dual energy CT system References: Kang MJ, et al. (2010) Dual-energy CT: clinical applications in various pulmonary diseases. Radiographics 30: The third one is the fast kv switch technique (Discovery CT 750 HD, GE Healthcare), which enables rapid and essentially simultaneous acquisition of datasets at two different energies for the same projection angle with one tube and a detector (Fig. 5c). Page 7 of 37

8 Fig.: 5c: Fast kv switch dual energy CT system References: Kang MJ, et al. (2010) Dual-energy CT: clinical applications in various pulmonary diseases. Radiographics 30: Among these three, the dual-source CT system is the most widely used technique. Recently, fast kv switch technique has come to be available. 3) How is the fast kv switch dual energy CT system and its difference from dual source CT The fast kv switch technology obtains images with dynamic switching between 80 and 140 kvp of X-rays in less than 0.5 msec with full 50cm field of view (FOV), while dual source technology obtains images with msec time delay between 80 and 140 kvp of X-rays. Because of such an extremely small time difference, the acquisition data set of 80 and 140 kvp are treated as coincident projection data for both temporally and spatially in the fast kv switch technology. This technology was enabled on the latest 64-slice CT by a new garnet crystal scintillator detector with ultrafast speed optical response, and by a high-voltage generator with ultrafast tube voltage switching mechanism. In addition, Page 8 of 37

9 the data acquisition system of that CT allows more than 2.5 times of data sampling in one gantry rotation compared with a conventional 64-slice CT, hence, a data acquisition which keeps in-plane spatial resolution and flux balance between 80 and 140 kvp is achieved (asymmetric sampling in Fig. 6). Fig.: 6: Dynamic switching between 80 and 140 kvp in less than 0.5 msec References: M. Jinzaki; Diagnostic Radiology, Keio University, 8582, JAPAN The fast kv switch data acquisition enables mathematically to transform the attenuation measurements into the density (or amount) of two basis materials in projection data space, while dual source data acquisition transform the attenuation measurements in image-data base. This projection-based process corrects for multi-material beam hardening effects, which provides accurate material decomposition with material density unit. The difference between dual-source CT and fast kvp switch dual energy CT is described in Table 1. Table 1 The difference between dual-source CT and fast kvp switch dual energy CT Dual Source CT Fast kv switch CT Tube Dual Single Page 9 of 37

10 Detector Gd Oxysulfide Gemstone Gantry rotation 300ms or 330ms 350ms Time between images 83ms or 75ms ms FOV 26cm, 33cm 50cm Reconstruction Image data base Projection data base Images for this section: Fig. 1: 1: Photoelectric effect and Compton scattering Page 10 of 37

11 Fig. 2: 2: CT Number of Water, Iodine and Calcium on different kvp The CT attenuation of a material is different in 140kV and 80kV. Each material has its own characteristic changing pattern in 140 and 80kV. Thus, the change in measurement of CT attenuation between 80 and 140kV will enable differentiating iodine and calcium. Page 11 of 37

12 Fig. 3: 3: Spectrum of X-ray 120-kVp X-rays possess a polychromatic spectrum, while virtual monochromatic image (VMSI) is a like as if the X-ray source produced only X-ray photons at a single energy Page 12 of 37

13 Fig. 4: 4: Beam hardening artifact (BHA) left: BHA between aorta and left ventricle, which may cause misdiagnosis of cardiac ischemia. right: well-known BHA in posterior fossa Fig. 5: 5a: Rotate-rotate dual energy CT system Page 13 of 37

14 Fig. 6: 5b: Dual-source dual energy CT system Page 14 of 37

15 Fig. 7: 5c: Fast kv switch dual energy CT system Page 15 of 37

16 Fig. 8: 6: Dynamic switching between 80 and 140 kvp in less than 0.5 msec Page 16 of 37

17 Imaging findings OR Procedure details 1) Potential clinical usefulness # Material decomposition Material decomposition has three major applications in medical imaging. Iodine image, water image (virtual non-contrast image), and the evaluation of chemical composition such as urolithiasis. a) Iodine image Iodine image shows high diagnostic accuracy for the detection of peripheral artery stenosis in severely calcified lesions. In 120 kv image, the evaluation of the degree of stenosis is difficult, however, in iodine image, it is clear where is stenotic and where is non-stenotic, which well corresponded to conventional angiography (Fig. 7, 8). Case 1 Fig.: 7: Usefulness of iodine-image in severely calcified lesion of peripheral artery In the virtual monochromatic image (left image, 70keV; virtual 120 kvp CT image), it was difficult to evaluate the degree of stenosis in left superficial femoral artery due to severe calcification. The significant stenotic legions were clearly demonstrated in the iodinedensity image (middle: arrows), which corresponded well to the CAG findings (right image) References: M. Jinzaki; Diagnostic Radiology, Keio University, 8582, JAPAN Case 2 Page 17 of 37

18 Fig.: 8: Usefulness of iodine-image in severely calcified lesion of peripheral artery In the virtual monochromatic image (left image, 70keV; virtual 120 kvp CT image), it was difficult to evaluate the degree of stenosis in right superficial femoral artery due to severe calcification. The significant stenotic legions were clearly demonstrated in the iodine-density image (middle: arrows), which corresponded well to the CAG findings (right image) References: M. Jinzaki; Diagnostic Radiology, Keio University, 8582, JAPAN This technology has not yet been applicable to clinical cardiac scan so far, we applied this technology to ex-vivo human-heart specimen. In the single-kvp CT image, it was difficult to evaluate the degree of stenosis in the left anterior descending artery, diagonal branch or left circumflex artery, where severely calcified plaques existed (Fig. 9, 10; right image). Iodine-density and calcium-density image were generated by fast kv switch dual energy CT data. The significant stenotic legions were clearly demonstrated in the iodine-density image (Fig. 9, 10; arrows in middle image), which corresponded well to the CAG findings. Thi;s result supports the application of fast kv switch dual energy CT has a possibility for the evaluation of severely calcified coronary arteries and may overcome the limitations of conventional CT. Page 18 of 37

19 Fig.: 9: Usefulness of iodine-image in severely calcified coronary artery of ex-vivo human-heart specimen The significant stenotic legions of severely calcified coronary artery (left image) were clearly demonstrated in the iodine-density image (middle image: arrows), which corresponded well to the CAG findings (right image: arrows). References: Yamada M, et al. (2011) Evaluation of Severely Calcified Coronary Artery Using Fast-Switching Dual-kVp 64-Slice Computed Tomography. Circ J. 75: Fig.: 10: Curved MPR image (the same case in figure 9) References: M. Jinzaki; Diagnostic Radiology, Keio University, 8582, JAPAN Page 19 of 37

20 An iodine map from dual-energy CT can demonstrate the distribution of pulmonary perfusion (Fig. 11). False positive perfusion defects are reported to be found associated with beam-hardening artifacts or motion artifacts on dual-source CT. further studies will be necessary to evaluate the diagnostic accuracy of fast kv switch CT perfusion mapping for the assessment of pulmonary embolism. Fig.: 11: Lobar perfusion defect due to a pulmonary embolism. (right image; axial, left image; coronal) References: M. Jinzaki; Diagnostic Radiology, Keio University, 8582, JAPAN b) Virtual non-contrast images Virtual non-contrast images have a potential to eliminate the unenhanced scan in the protocol of abdominal organ evaluation. Liver metastases evaluation is one of a good application of Virtual non-contrast images. We assessed the relationship between CT value (HU) in true non-contrast image and water-equivalent density value (mg/cc) in virtual non-contrast image derived from spectral imaging with fast kv switch in twelve patients with liver metastases underwent both noncontrast scan at 120kVp and enhanced scan at spectral imaging with fast kv switch (80kVp and 140kVp)(Fig. 12). Page 20 of 37

21 Fig.: 12: Virtual non-contrast image of liver metastases right image; true non-contrast image (120 kvp), left image; virtual non-contrast image (70keV) References: M. Jinzaki; Diagnostic Radiology, Keio University, 8582, JAPAN The R-square of linear regression was excellent (1.000+/-0.01) between all the measured values of virtual non-contrast image and true non-contrast image (Fig.13). Virtual noncontrast images derived from enhanced spectral imaging are equivalent to true noncontrast images, and will be clinically applicable. Page 21 of 37

22 Fig.: 13: The relationship between CT value (HU) in true non-contrast image and water-equivalent density value (mg/cc) in virtual non-contrast image References: M. Jinzaki; Diagnostic Radiology, Keio University, 8582, JAPAN The detection of urolithiasis is another good application of virtual non-contrast image (Fig. 14). However, image quality of virtual non-contrast images is lower than that of true noncontrast image. Page 22 of 37

23 Fig.: 14: The detection of urolithiasis on virtual non-contrast image. A small sized renal stone masked by contrast material in excretory phase image (right image)is detected on virtual non-contrast image (middle image) as well as true non-contrast image(left image). References: M. Jinzaki; Diagnostic Radiology, Keio University, 8582, JAPAN c) Chemical composition of urolithiasis Determining the composition of a calculus may be clinically useful because the effectiveness of different treatment options is determined by the stone composition: For example, uric acid calculi can be medically dissolved with urine alkalization, and cystine calculi are known to be shock wave resistant and thus should be removed invasively. Dual-energy CT has inherent capability to help differentiate different materials that have similar electron densities but varying photon absorption, and can help differentiate between calcified, uric acid, and cystine calculi. However, the data using fast kvp switch technology is not yet reported. # Virtual monochromatic spectrum image VMS image is expected to provide improved image quality by reducing beam hardening artifacts, but the image quality and radiation dose have not been well estimated yet. We compared the image quality obtained with VMS imaging in phantoms with that obtained with 120-kVp CT for a given radiation dose. We measured CT attenuation in four plastic syringes (volume, 20 ml; diameter, 2 cm) simulating vascular structures filled with water or a diluted contrast medium at three different iodine concentrations (5, 10, and 15 mg of iodine per milliliter) (Fig. 15), placed in the torso phantom. Page 23 of 37

24 Fig.: 15: Reconstructed images of the syringes obtained with 70-keV VMS imaging. The region of interest was placed in the inner lumen of each syringe filled with a diluted contrast medium ( 1 = 15 mg of iodine per milliliter, 2 = 10 mg of iodine per milliliter, and 3 = 5 mg of iodine per milliliter) or water (4) to measure CT attenuation References: M. Jinzaki; Diagnostic Radiology, Keio University, 8582, JAPAN Image noise on VMS images in the range of kev was significantly lower than that on the 120-kVp CT images (Fig. 16), while CNR on the VMS images was approximately 30% higher compared with that on the 120-kVp CT images (Fig. 17). Page 24 of 37

25 Fig.: 16: Graph shows mean standard deviation of background noise on VMS images and 120-kVp CT images. Mean standard deviation on 120-kVp CT images was 16.6 HU. Background noise on VMS images was lowest at 69 kev, and background noise levels on VMS images in the range of kev were significantly lower than were those on 120-kVp CT images. References: M. Jinzaki; Diagnostic Radiology, Keio University, 8582, JAPAN Page 25 of 37

26 Fig.: 17: Graph shows mean CNR and corresponding standard deviation (vertical bars) on VMS and 120-kVp CT images. The CNR for each contrast medium concentration was highest at 68 kev. The CNR on the VMS images was significantly higher in the range of kev (5 mg of iodine per milliliter) and kev (10 and 15 mg of iodine per milliliter) compared with that on 120-kVp CT images. References: M. Jinzaki; Diagnostic Radiology, Keio University, 8582, JAPAN VMS image at approximately 70 kev yielded lower image noise and higher contrastnoise ratio than 120-kVp CT for a given radiation dose. VMS image has the potential to replace 120-kVp CT as the standard CT imaging modality, since optimal VMS image may be expected to provide improved image quality in a standard body habitus patient (Radiology, in press). We also estimated the noise of each in-vivo organ on VMS image of enahcned CT at 1-keV intervals (range, kev). The averaged kev with the lowest image noise of each organ is demonstrated in Table 2. Thus, the optimal kev was considered to be in the range of kev on VMS images as shown in the phantom study. Page 26 of 37

27 Table 2 Optimal kev for each organ (n=30) Organ Optimal kev ## Liver Spleen Aorta Pancreas Fat Muscle Kidney ) Limitation In obese patients or small patients, the absorption of low-energy x-rays will be different from that in standard-sized patients, and the image noise, CNR, and optimal energy for VMS imaging might change. Although higher radiation dose would be required in obese patients, this problem may be solved by the use of iterative reconstruction. Further study is necessary to determine how to obtain dual energy CT images in obese or small patients. Images for this section: Fig. 1: 7: Usefulness of iodine-image in severely calcified lesion of peripheral artery In the virtual monochromatic image (left image, 70keV; virtual 120 kvp CT image), it was difficult to evaluate the degree of stenosis in left superficial femoral artery due to Page 27 of 37

28 severe calcification. The significant stenotic legions were clearly demonstrated in the iodine-density image (middle: arrows), which corresponded well to the CAG findings (right image) Fig. 2: 8: Usefulness of iodine-image in severely calcified lesion of peripheral artery In the virtual monochromatic image (left image, 70keV; virtual 120 kvp CT image), it was difficult to evaluate the degree of stenosis in right superficial femoral artery due to severe calcification. The significant stenotic legions were clearly demonstrated in the iodine-density image (middle: arrows), which corresponded well to the CAG findings (right image) Fig. 3: 9: Usefulness of iodine-image in severely calcified coronary artery of ex-vivo human-heart specimen The significant stenotic legions of severely calcified coronary artery (left image) were clearly demonstrated in the iodine-density image (middle image: arrows), which corresponded well to the CAG findings (right image: arrows). Page 28 of 37

29 Fig. 4: 10: Curved MPR image (the same case in figure 9) Fig. 5: 11: Lobar perfusion defect due to a pulmonary embolism. (right image; axial, left image; coronal) Page 29 of 37

30 Fig. 6: 12: Virtual non-contrast image of liver metastases right image; true non-contrast image (120 kvp), left image; virtual non-contrast image (70keV) Page 30 of 37

31 Fig. 7: 13: The relationship between CT value (HU) in true non-contrast image and waterequivalent density value (mg/cc) in virtual non-contrast image Fig. 8: 14: The detection of urolithiasis on virtual non-contrast image. A small sized renal stone masked by contrast material in excretory phase image (right image)is detected on virtual non-contrast image (middle image) as well as true non-contrast image(left image). Page 31 of 37

32 Fig. 9: 15: Reconstructed images of the syringes obtained with 70-keV VMS imaging. The region of interest was placed in the inner lumen of each syringe filled with a diluted contrast medium ( 1 = 15 mg of iodine per milliliter, 2 = 10 mg of iodine per milliliter, and 3 = 5 mg of iodine per milliliter) or water (4) to measure CT attenuation Page 32 of 37

33 Fig. 10: 16: Graph shows mean standard deviation of background noise on VMS images and 120-kVp CT images. Mean standard deviation on 120-kVp CT images was 16.6 HU. Background noise on VMS images was lowest at 69 kev, and background noise levels on VMS images in the range of kev were significantly lower than were those on 120-kVp CT images. Page 33 of 37

34 Fig. 11: 17: Graph shows mean CNR and corresponding standard deviation (vertical bars) on VMS and 120-kVp CT images. The CNR for each contrast medium concentration was highest at 68 kev. The CNR on the VMS images was significantly higher in the range of kev (5 mg of iodine per milliliter) and kev (10 and 15 mg of iodine per milliliter) compared with that on 120-kVp CT images. Page 34 of 37

35 Conclusion 1) The fast kv switch technology is a new dual energy CT system, which enables rapid and essentially simultaneous acquisition of datasets at two different energies for the same projection angle with one tube and a detector. The same projection angle acquisition enables the reconstruction images in projection data space, which results in the rigorous beam hardening correction and high quality VMS image. 2) VMS image has the potential to replace 120-kVp CT as the standard CT imaging modality in a standard body habitus patient, since optimal VMSI provide improved image quality in the fast kv switch technology. 3) The clinical usefulness of material decomposition image is almost same with those reported in dual-source dual energy CT system. Iodine image will enables to evaluate severely calcified vessels, which has been the limitation of conventional 64-slice CT. Virtual non-contrast image has the potential to reduce radiation dose in a proper diagnostic task by eliminating true non-contrast image acquisition. 4) Further study is necessary to determine whether these results are also applicable to obese or small patients. Personal Information M. Jinzaki, Y. Tanami, H. Sugiura, M. inoue, S. Koga, Y. Yamada, S. Kuribayashi Department of Diagnostic Radiology, Keio University School of Medicine, Tokyo, Japan jinzaki@rad.med.keio.ac.jp References 1) Genant HK, Boyd D. Quantitative bone mineral analysis using dual energy computed tomography. Invest Radiol Nov-Dec;12(6): Page 35 of 37

36 2) Avrin DE, Macovski A, Zatz LE. Clinical application of Compton and photo-electric reconstruction in computed tomography: preliminary results. Invest Radiol MayJun;13(3): ) Chiro GD, Brooks RA, Kessler RM, Johnston GS, Jones AE, Herdt JR, Sheridan WT. Tissue signatures with dual-energy computed tomography. Radiology May;131(2): ) Millner MR, McDavid WD, Waggener RG, Dennis MJ, Payne WH, Sank VJ. Extraction of information from CT scans at different energies. Med Phys Jan-Feb;6(1): ) Kalender WA, Perman WH, Vetter JR, Klotz E. Evaluation of a prototype dualenergy computed tomographic apparatus. I. Phantom studies. Med Phys MayJun;13(3): ) Sun H, Qiu S, Lou S, Liu J, Li C, Jiang G. A correction method for nonlinear artifacts in CT imaging. Conf Proc IEEE Eng Med Biol Soc. 2004; 2: ) Wu X, Langan DA, Xu D, Benson TM, Pack JD, Schmitz AM, J. Tkaczyk JE. Monochromatic CT Image Representation via Fast Switching Dual kvp. Conf Proc IEEE Eng Med Biol Soc. 2004; 2: ) Menvielle N, Goussard Y, Orban D, Soulez G. Reduction of beam hardening artifact in X-ray CT, Conf Proc IEEE Eng Med Biol Soc. 2005; 2: ) Johnson TR, Krauss B, Sedlmair M, Grasruck M, Bruder H, Morhard D, Fink C, Weckbach S, Lenhard M, Schmidt B, Flohr T, Reiser MF, Becker CR. Material differentiation by dual energy CT: initial experience. Eur Radiol 2007 Jun;17(6): ) Graser A, Johnson TR, Bader M, Staehler M, Haseke N, Nikolaou K, Reiser MF, Stief CG, Becker CR. Dual energy CT characterization of urinary calculi: initial in vitro and clinical experience. Invest Radiol 2008 ; 43 ( 2 ): ) Szucs-Farkas Z, Verdun FR, von Allmen G, Mini RL, Vock P. Effect of X-ray tube parameters, iodine concentration, and patient size on image quality in pulmonary computed tomography angiography: a chest-phantomstudy. Invest Radiol 2008 ; 43 ( 6 ): ) Coursey CA, Nelson RC, Boll DT, Paulson EK, Ho LM, Neville AM, Marin D, Gupta RT, Schindera ST. Dual-energy multidetector CT: how does it work, what can it tell us, and when can we use it in abdominopelvic imaging? Radiographics JulAug;30(4): ) Rodríguez-Granillo GA, Rosales MA, Degrossi E, Rodriguez AE. Signal density of left ventricular myocardial segments and impact of beam hardening artifact: implications Page 36 of 37

37 for myocardial perfusion assessment by multidetector CT coronary angiography. Int J Cardiovasc Imaging Mar;26(3): ) Kang MJ, Park CM, Lee CH, Goo JM, Lee HJ. Dual-energy CT: clinical applications in various pulmonary diseases. Radiographics May;30(3): ) Yamada M, Jinzaki M, Imai Y, Yamazaki S, Imanishi N, Tanami Y, Yamazaki A, Aiso S, Kuribayashi S. Evaluation of Severely Calcified Coronary Artery Using Fast-Switching Dual-kVp 64-Slice Computed Tomography. Circ J Jan 25;75(2): ) Matsumoto K, Jinzaki M, Tanami Y, Ueno A, Yamada M, Kuribayashi S. Virtual Monochromatic Spectral Imaging with Fast Kilovoltage Switching: Improved Image Quality as Compared with that Obtained with Conventional 120-kVp CT. Radiology 2011, in press. Page 37 of 37