Differentiation of Kidney Stones Using Dual-Energy CT With and Without a Tin Filter

Transcription

1 Medical Physics and Informatics Original Research Fung et al. Dual-Energy CT of Kidney Stones Using a Tin Filter Medical Physics and Informatics Original Research JOURNAL CLUB George S. K. Fung 1 Satomi Kawamoto 1 Brian R. Matlaga 2 Katsuyuki Taguchi 1 Xiaodong Zhou 3 Elliot K. Fishman 1 Benjamin M. W. Tsui 1 Fung GSK, Kawamoto S, Matlaga BR, et al. Keywords: differentiation, dual-energy CT, kidney stone, spectrum optimization DOI: /AJR Received May 13, 2011; accepted after revision November 16, This project is funded in part by a Johns Hopkins University Siemens Healthcare research grant. 1 The Russell H. Morgan Department of Radiology and Radiological Science, Johns Hopkins University, 601 N Caroline St, JHOC 4253, Baltimore, MD Address correspondence to G. S. K. Fung (gfung2@jhmi.edu). 2 The James Buchanan Brady Urological Institutes, Johns Hopkins University, Baltimore, MD. 3 Siemens Healthcare, Malvern, PA. AJR 2012; 198: X/12/ American Roentgen Ray Society Differentiation of Kidney Stones Using Dual-Energy CT With and Without a Tin Filter OBJECTIVE. The aim of this in vitro study was to examine the capability of three protocols of dual-energy CT imaging in distinguishing calcium oxalate, calcium phosphate, and uric acid kidney stones. MATERIALS AND METHODS. A total of 48 calcium oxalate, calcium phosphate, and uric acid human kidney stone samples were placed in individual containers inside a cylindric water phantom and imaged with a dual-energy CT scanner using the following three scanning protocols of different combinations of tube voltage, with and without a tin filter: 80 and 140 kvp without a tin filter, 100 and 140 kvp with a tin filter, and 80 and 140 kvp with a tin filter. The mean attenuation value (in Hounsfield units) of each stone was recorded in both low- and high-energy CT images in each protocol. The dual-energy ratio of the mean attenuation values of each stone was computed for each protocol. RESULTS. For all three protocols, the uric acid stones were significantly different (p < 0.001) from the calciferous stones according to their dual-energy ratio values. For differentiating calcium oxalate and calcium phosphate stones, the difference between their dual-energy ratio values was statistically significant, with different degrees of significance (range, p < to p = 0.03) for all three protocols. On the basis of the values of the area under receiver operating characteristic curve (AUC) of calcified stone differentiation, the three protocols were ranked in the following order: the 80- and 140-kVp tin filter protocol (AUC, 0.996), the 100- and 140-kVp tin filter protocol (AUC, 0.918), and the 80- and 140-kVp protocol (AUC, 0.871). CONCLUSION. The tin filter added to the high-energy tube and the use of a wider dual-energy difference are important for improving the stone differentiation capability of dualenergy CT imaging. U nenhanced helical CT is the preferred method for evaluating patients with suspected urinary calculi [1]. At present, CT images provide information about the presence, size, and location of stones, as well as the presence of associated complicating features, such as hydronephrosis or perinephric stranding. Because many patients harboring urinary calculi will ultimately require some form of intervention, there is great importance in further developing the capability of CT to predict stone composition. A reliable determination of the type of stone that is present will allow the clinician to better stratify treatment options for the patient. Indeed, certain stones, such as those composed of uric acid, may be treated medically and may not require a surgical procedure. For calciferous stones, the type of calcium salt present in the stone affects the success of in- terventions such as shock wave lithotripsy [2 5]. As reported in the literature [6 16], CT is reliable for differentiating uric acid stones from calciferous stones. However, it is less certain as to whether CT can be used to differentiate between different calciferous kidney stone types, such as calcium oxalate and calcium phosphate stones. In recent years, clinically viable dualsource CT scanners with dual-energy capability have been made available [17]. Two sets of tube-detector pairs are situated approximately 90 from each other. When the scanner is set to dual-energy mode, the two tubes run at two different voltages and energy spectra (low- and high-energy spectra) to achieve simultaneous dual-energy CT image acquisition. The dual-energy CT images provide an additional dimension of information on the energy-dependent attenuation property of the object of interest. Previously, most 1380 AJR:198, June 2012

2 Dual-Energy CT of Kidney Stones Using a Tin Filter researchers focused only on distinguishing uric acid from non uric acid stones [18 24]. Recently, it was reported that, in addition to reliably differentiating uric acid stones from calciferous stones, calcium oxalate and calcium phosphate stones could be marginally differentiated (p = ) in dual-energy CT images acquired utilizing a Somatom Definition CT scanner (Siemens Healthcare) [25]. In the newer generation of the scanner (Somatom Definition Flash CT scanner, Siemens Healthcare), a tin filter was added to the high-energy tube to improve the spectral separation between the low- and high-energy tubes, which, in theory, improved its capability for material differentiation [26]. In a recent in vitro study [27], dual-energy CT with an added tin filter was reported to have achieved 100% diagnostic accuracy in differentiating uric acid from non uric acid stones. However, whether calciferous stones, such as calcium oxalate versus calcium phosphate stones, can be differentiated in images obtained utilizing dual-energy CT with an added tin filter is still to be determined. Recently, researchers using dual-energy CT with a tin filter claimed (without a detailed explanation) that further differentiation of calciferous stones was not possible [24], whereas multiple subtypes of calciferous stones were reported (without details from the abstract) to be differentiable in another report (Qu M et al., presented at the 2009 annual meeting of the Radiologic Society of North America). In this in vitro study, we focused on investigating the capability of dual-energy CT to differentiate the three most common kidney A stone types: calcium oxalate, calcium phosphate, and uric acid stones. First, we examined whether these stone types were statistically different in the three dual-energy CT protocols, which were 80 and 140 kvp, 100 and 140 kvp with an added tin filter, and 80 and 140 kvp with an added tin filter. Then, the capabilities of these three protocols in differentiating calcium oxalate, calcium phosphate, and uric acid were statistically compared. Materials and Methods Kidney Stone Samples A total of 48 human kidney stone samples (16 pure calcium oxalate, 16 pure calcium phosphate, and 16 pure uric acid stones) with accompanying interpretation reports were obtained from a stone analysis laboratory (Beck Analytic Services). The reported stone compositions were determined by microscopic visual inspection, chemical reaction, and Fourier transform infrared microspectroscopy. For the purposes of this study, we define pure uric acid, calcium oxalate, and calcium phosphate stones as those comprising at least 90% uric acid, monohydrate, and hydroxyapatite, respectively. Phantom A phantom with a circular cross-section of 21.6 cm in inside diameter made of acrylic plastic (Plexiglas, Altuglas International) and filled with water was used. A double-layer rack was custom built with acrylic plastic rods and planes and was mounted firmly inside the phantom. Each stone was placed individually inside a small water-filled container (centrifugal tube) and was held in place by alcohol preparation at approximately the mid height of the container. All stones were Calcium Oxalate Fig. 1 Phantom with kidney stones. A, Photograph of cylindric water phantom with all 48 kidney stones embedded inside. B, Schematic diagram of arrangement of different classes of kidney stones in cross-section of one double-layer rack inside phantom. Uric Acid B immersed in the water phantom for over 24 hours before the experiments were performed, to ensure that the pores of the stones were filled with water. The containers were inserted into the double-layer rack inside the phantom. The exterior photo and the schematic diagram of the water phantom are shown in Figure 1. Protocols The phantom with embedded kidney stones was then scanned with a dual-source CT scanner (Somatom Definition Flash, Siemens Healthcare) utilizing the three different dual-energy protocols (80 and 140 kvp, 100 and 140 kvp with tin filter, and 80 and 140 kvp with tin filter). The x-ray energy spectra from the different protocols are shown in Figure 2. For the 80- and 140-kVp protocol, the low-energy tube was set to 80 kvp with effective tube current of 248 ma, and the high-energy tube was set to 140 kvp with effective tube current of 45 ma. For the 100- and 140-kVp tin filter protocol, the low-energy tube was set to 100 kvp with effective tube current of 171 ma, and the highenergy tube was set to 140 kvp with effective tube current of 132 ma and an added tin filter. For the 80- and 140-kVp tin filter protocol, the low-energy tube was set to 80 kvp with an effective tube current of 244 ma, and the high-energy tube was set to 140 kvp with effective tube current of 132 ma and an added tin filter. The ratios of effective tube currents were chosen by the manufacturer. The volume CT dose index was automatically set to approximately 6 mgy for each protocol, which is lower than that used in a clinical protocol, because the phantom was smaller than a standard human torso size. A detector collimation of mm, a pitch of 0.8, and a rotation time of 0.5 second were set. The phantom was positioned at the center of the CT scanner with the axis of the cylindric phantom aligned with the z-axis of the scanner. In each dual-energy CT protocol setting, the phantom was scanned and two separate sets of CT images were reconstructed for the low- and high-energy tube settings. The images were reconstructed with D30f kernel for dual-energy imaging, with a slice width of 0.6 mm and an increment of 0.4 mm. The diameter of the reconstructed FOV was 275 mm with a pixel matrix. Image Segmentation Thresholding was applied to the pixel values to segment the 48 kidney stones from the background in the acquired high-energy CT images. The threshold was determined by adding the average attenuation (measured in Hounsfield units) of the water in the CT images to three times the SD of the attenuation at the center of the phantom with the water background. Because the containers for AJR:198, June

3 Fung et al. Photon Flux ( 10 3 ) Kiloelectron Voltage the stone samples had an attenuation value similar to that of water, which is much lower than the attenuation values of kidney stones, the proximity of the containers to the stones did not affect the segmentation. To avoid partial volume effects on the measured attenuation value of the kidney stones near the boundaries of the stones, a 3D morphologic erosion operation with a mask was applied to the segmentation results to remove the boundary regions of the stones. The average attenuation value of each stone in the high-energy CT image was computed. The segmentation mask of each stone, determined in the high-energy CT image, was then applied to the low-energy CT image, and the average low-energy attenuation value was also computed. The image analysis program was written in Matlab (version 7.6, MathWorks). Dual-Energy Measurements The dual-energy ratio of the stone was defined as the mean attenuation value in the low-energy CT image divided by that in the high-energy CT image of the stone, as follows: dual-energy ratio (protocol 1, stone 1) = low-energy attenuation value (protocol 1, stone 1) / high-energy attenuation value (protocol 1, stone 1). The dual-energy ratio value should remain constant for stones composed of the same material but with different degrees of porosity, given that the water is the background filling material, or if the density of the stones is the only factor that is different. Statistical Analysis To examine whether the dual-energy ratios of calcium oxalate, calcium phosphate, and uric acid stones were statistically different in the three different dual-energy CT protocols, the repeated measures analysis of variance was used to test the equivalency of the mean dual-energy ratios from the three different 80 kvp 100 kvp 140 kvp 140 kvp tin filter 130 Fig. 2 X-ray spectra of CT scanner (Definition Flash, Siemens Healthcare) at 80, 100, and 140 kvp with and without tin filter added. protocols. The Tukey honestly significant difference pairwise comparisons were performed in conjunction with the analysis of variance to adjust for multiple comparisons. A p value of less than 0.05 was considered statistically significant. To compare the capabilities of the three protocols in differentiating calcium oxalate and calcium phosphate stones, a receiver operating characteristic (ROC) curve analysis was used, and the area under ROC curve (AUC) value was used to rank the differentiation capabilities of the three protocols. Data analysis was performed using Stata software (version 11.1, StataCorp). Fig. 3 Dual-energy CT images of phantom. A, Image was obtained at 80 kvp (low energy). B, Image was obtained at 140 kvp (high energy) with tin filter added. A Results A sample pair of low- and high-energy CT images acquired at the same slice position of the phantom in the 80- and 140-kVp tin filter protocol are depicted in Figure 3. Not all of the stones showed up in all of the containers in a single CT image, because the stones were located at approximately the mid height of the tube but not at the exact same transaxial slice position. The sizes of the 16 calcium oxalate, 16 calcium phosphate, and 16 uric acid stones in the CT images were analyzed by measuring their largest diameters. From the segmentation result of the 140-kVp tin filter CT image, the mean (± SD) largest diameters of the stones after erosion operation was 4.24 ± 1.80 mm for calcium oxalate, 4.40 ± 2.42 mm for calcium phosphate, and 4.34 ± 1.79 mm for uric acid. There was no significant difference between the sizes of the three stone types (p = 0.977) using the analysis of variance test. The dual-energy plots of the attenuation values of the kidney stone samples in the dual-energy CT images from the dual-energy acquisition protocols are depicted in Figure 4. They represent the scatterplots for all the kidney stone samples, with the average attenuation values of the high-energy CT image on the horizontal axis and the average attenuation values of low-energy CT on the vertical axis. Each symbol represents a stone sample. In Figure 4A, the dual-energy plots of the stones acquired in the 80- and 140-kVp protocol and the 80- and 140-kVp tin filter protocol are shown. In Figure 4B, the dual-energy plots of the stones acquired in the 100- and 140-kVp tin filter protocol and the 80- and 140-kVp tin filter protocols are shown. Linear regression was applied to show the slope of the best-fit line for each stone type for each protocol. Figure 5 shows the scatterplot of the dualenergy ratio of all of the stone samples for B 1382 AJR:198, June 2012

4 Dual-Energy CT of Kidney Stones Using a Tin Filter Attenuation at 80 kvp (HU) Calcium Oxalate 80 and 140 kvp Tin Filter 80 and 140 kvp Tin Filter Uric Acid 80 and 140 kvp Tin Filter Calcium Oxalate 80 and 140 kvp 80 and 140 kvp Uric Acid 80 and 140 kvp Attenuation at 140 kvp With or Without Tin Filter (HU) Attenuation at 80 or 100 kvp (HU) each of the three protocols. Table 1 presents the dual-energy ratios for each stone type obtained with each protocol, as well as the multiple comparison results of the Tukey honestly significant difference pairwise comparisons. For all three protocols, the dual-energy ratio values of the uric acid stones were significantly different from the calcium oxalate and calcium phosphate stones (p < for all calcium oxalate vs uric acid stones and calcium phosphate vs uric acid stones). For calcium oxalate versus calcium phosphate stones, the dual-energy ratio values were significantly different with the 80- and 140-kVp protocol (p = 0.030), the 100- and 140-kVp tin filter protocol (p = 0.001), and the 80- and 140-kVp tin filter protocol (p < 0.001). It is important to note that the p value for the 80- and 140-kVp tin filter protocol was lower than that for the 100- and 140-kVp tin filter protocol and much lower than that for the 80- and 140- kvp protocol. Therefore, the 80- and 140-kVp tin filter protocol should be preferred over the other two protocols. The ROC curves of the differentiation of the calcium oxalate and calcium phosphate stones in the three protocols are shown in Figure 6. The ROC study for the calcium oxalate versus uric acid and the calcium phosphate versus uric acid stones was not shown because the dual-energy ratio values were completely capable of separating uric acid from calciferous stones in all protocols, and thus, those AUC values were all equal to 1. From Figure 6, the AUC values of the curves range from to for the three protocols. The 80- and 140-kVp tin filter protocol showed a very high AUC value of 0.996, which is about 0.08 higher than the AUC value of for the 100- and 140-kVp tin filter protocol, and 0.13 higher than the AUC value of for the 80- and 140-kVp protocol. Discussion From the results in Table 1, we show that uric acid stones can be differentiated from the non uric acid (calcium oxalate and calcium phosphate) stone samples in all three protocols. This finding is consistent with results of previous studies [18 24]. The addition of a tin filter at a high-energy source might not be necessary for the relatively simple differentiation of Calcium Oxalate 80 and 140 kvp Tin Filter 80 and 140 kvp Tin Filter Uric Acid 80 and 140 kvp Tin Filter Calcium Oxalate 100 and 140 kvp Tin Filter 100 and 140 kvp Tin Filter Uric Acid 100 and 140 kvp Tin Filter Attenuation at 140 kvp With Tin Filter (HU) A Fig. 4 Dual-energy plots for all stone samples grouped into calcium oxalate, calcium phosphate, and uric acid stone types. A, Graph shows 80- and 140-kVp protocols with and without tin filter. B, Graph shows 100- and 140-kVp tin filter protocol versus 80- and 140-kVp tin filter protocol. Linear regression lines and corresponding equations are shown. Fig. 5 Dual-energy ratio scatterplot of stone samples of different stone types for three protocols. Dual-Energy Ratio and 140 kvp Tin Filter uric acid and non uric acid stones. However, if dual-energy CT is to be used to address the more difficult problem of differentiating calcium oxalate and calcium phosphate stones, the added tin filter plays a critical role. Our results based on dual-energy ratio values show that the calcium oxalate stones were significantly different from the calcium phosphate stones in all three protocols. In the protocols with the tin filter added (80- and 140-kVp tin filter protocol and 100- and 140-kVp tin filter protocol), their p values (p < and p = 0.001, respectively) were much lower than that (p = 0.030) for the protocol without the tin filter added (80- and 140-kVp protocol). This outcome is con- 100 and 140 kvp Tin Filter Protocol Calcium Oxalate Uric Acid 80 and 140 kvp B AJR:198, June

5 Fung et al. TABLE 1: Dual-Energy Ratio and Statistical Differences for Calcium Oxalate,, and Uric Acid Stones for Each Protocol Dual-Energy Ratio (Mean ± SD) p Protocol Calcium Oxalate Uric Acid vs Calcium Oxalate Calcium Oxalate vs Uric Acid vs Uric Acid 80 and 140 kvp ± ± ± < < and 140 kvp with tin filter ± ± ± < < and 140 kvp with tin filter ± ± ± < < < Note p values were calculated with Tukey honestly significant difference pairwise comparisons. Sensitivity Specificity 80 and 140 kvp Tin Filter ROC area: and 140 kvp Tin Filter ROC area: and 140 kvp ROC area: Fig. 6 Receiver operating characteristic (ROC) curves and area under the curve values of differentiation of calcium oxalate and calcium phosphate stones for three protocols. sistent with a previous report [25], in which calcium oxalate and calcium phosphate stones were significantly different, with high p values ( ), in the 80- and 140-kVp protocol, and with an another report [28], in which hydroxyapatite (calcium phosphate stones used in this study) were significantly different from calcium oxalate stones according to the 80- and 140-kVp tin filter protocol. However, in a previous study [24], it was claimed that further differentiation of calcified calculi was not possible without experimental support. The possible explanations are, in that study, the sample size was small (six samples per stone class), four calcified stone subclasses were studied, and stones with low purity or with multiple composition were used. However, our results agreed with those of the previous study [24] in that the stone differentiation capability was improved by adding a tin filter to the high-energy source, resulting from the reduction of the overlap between the low- and high-energy x- ray energy spectra, as shown in Figure 2. On the basis of the AUC values, all three protocols were equally capable of distinguishing uric acid from non uric acid stones. In differentiating calcium oxalate from calcium phosphate stones, the 80- and 140-kVp tin filter protocol was better than the 100- and 140- kvp tin filter protocol, which, in turn, was better than the 80- and 140-kVp protocol. However, a statistical analysis of the AUC values could not be performed because the sample size in this study was not sufficiently large to test the equality of the AUC values, which requires a much larger sample size [29]. A limitation of this study is that a phantom with relatively small size was used and it is not readily applicable to reproducing the scenario for normal-sized human torso CT scans. A reduced radiation dose was used to partially compensate for the situation. Small, medium, and large phantoms, which mimic pediatric, adult, and obese patients, at nominal clinical CT dose index radiation doses should be used in future studies. Another limitation is that only a few protocols were available to be tested in this study. Optimization of the dual-energy protocol is needed for different scenarios. Changes to the peak kilovoltage setting and to the material and thickness of the added metal filter would also create the potential to further improve kidney stone differentiation. Realistic simulation, by integrating a digital anthropomorphic phantom and an x-ray projection simulator with accurate scanner settings, such as the XCAT/DRASIM simulation package described elsewhere (Fung GSK et al., presented at the 2011 SPIE Medical Imaging Conference), would provide an alternate means for thorough optimization by dramatically reducing the number of experiments needed. In conclusion, on the basis of the results of this in vitro study, all three protocols were capable of distinguishing uric acid, calcium oxalate, and calcium phosphate stones. For differentiating calcium oxalate and calcium phosphate stones, the 80- and 140-kVp tin filter protocol had better capability than did the 100- and 140-kVp tin filter protocol, which in turn had better capability than did the 80- and 140-kVp protocol. With the wider energy spectral difference between lowand high-energy tubes and using a tin filter added in the high-energy tube (the 80- and 140-kVp tin filter protocol), dual-energy CT has achieved significant improvement in differentiating calcium oxalate, calcium phosphate, and uric acid stones. Acknowledgments We thank Jiangxia Wang (Johns Hopkins Biostatistics Center) for her assistance with the statistical analysis, Saul Friedman (Tel Aviv University) for his assistance with the image analysis, and Vince Blasko and Beatrice Mudge (Johns Hopkins Medical Institutions) for their assistance with the CT scans. We also thank Thomas Flohr and Karl Stierstorfer (Siemens Healthcare) for their helpful discussions. References 1. Heidenreich A, Desgrandschamps F, Terrier F. Modern approach of diagnosis and management of acute flank pain: review of all imaging modalities. Eur Urol 2002; 41: AJR:198, June 2012

6 Dual-Energy CT of Kidney Stones Using a Tin Filter 2. Mostafavi MR, Ernst RD, Saltzman B. Accurate Urol 1983; 130: tial in vitro and clinical experience. Invest Radiol determination of chemical composition of urinary 12. Newhouse JH, Prien EL, Amis ES Jr, et al. Com- 2008; 43: calculi by spiral computerized tomography. J puted tomographic analysis of urinary calculi. 21. Stolzmann P, Scheffel H, Rentsch K, et al. Dual- Urol 1998; 159: AJR 1984; 142: energy computed tomography for the differentia- 3. Deveci S, Coskun M, Tekin MI, et al. Spiral com- 13. Saw KC, McAteer JA, Monga AG, et al. Helical tion of uric acid stones: ex vivo performance puted tomography: role in determination of chemical CT of urinary calculi: effect of stone composition, evaluation. Urol Res 2008; 36: compositions of pure and mixed urinary stones an in vitro study. Urology 2004; 64: Hurtado F, Gutierrez J, Castano-Tostado E, et al. Invivo relation between CT attenuation value and shockwave fragmentation. J Endourol 2007; 21: Williams JC Jr, Saw KC, Paterson RF, Hatt EK, McAteer JA, Lingeman JE. Variability of renal stone fragility in shock wave lithotripsy. Urology 2003; 61: Demirel A, Suma S. The efficacy of non-contrast helical computed tomography in the prediction of urinary stone composition in vivo. J Int Med Res 2003; 31: Kuwahara M, Kageyama S, Kurosu S, et al. Computed tomography and composition of renal calculi. Urol Res 1984; 12: Pareek G, Armenakas NA, Fracchia JA. Hounsfield units on computerized tomography predict stone-free rates after extracorporeal shock wave lithotripsy. J Urol 2003; 169: Motley G, Dalrymple N, Keesling C, et al. Hounsfield unit density in the determination of urinary stone composition. Urology 2001; 58: Nakada SY, Hoff DG, Attai S, et al. Determination of stone composition by noncontrast spiral computed tomography in the clinical setting. Urology 2000; 55: Mitcheson HD, Zamenhof RG, Bankoff MS, et al. Determination of the chemical composition of urinary calculi by computerized tomography. J stone size, and scan collimation. AJR 2000; 175: Zarse CA, McAteer JA, Tann M, et al. Helical computed tomography accurately reports urinary stone composition using attenuation values: in vitro verification using high-resolution micro-computed tomography calibrated to Fourier transform infrared microspectroscopy. Urology 2004; 63: Bellin MF, Renard-Penna R, Conort P, et al. Helical CT evaluation of the chemical composition of urinary tract calculi with a discriminant analysis of CT-attenuation values and density. Eur Radiol 2004; 14: Sheir KZ, Mansour O, Madbouly K, et al. Determination of the chemical composition of urinary calculi by noncontrast spiral computerized tomography. Urol Res 2005; 33: Flohr TG, McCollough CH, Bruder H, et al. First performance evaluation of a dual-source CT (DSCT) system. Eur Radiol 2006; 16: Primak AN, Fletcher JG, Vrtiska TJ, et al. Noninvasive differentiation of uric acid versus nonuric acid kidney stones using dual-energy CT. Acad Radiol 2007; 14: Grosjean R, Sauer B, Guerra RM, et al. Characterization of human renal stones with MDCT: advantage of dual energy and limitations due to respiratory motion. AJR 2008; 190: Graser A, Johnson TR, Bader M, et al. Dual energy CT characterization of urinary calculi: ini- 22. Thomas C, Patschan O, Ketelsen D, et al. Dual-energy CT for the characterization of urinary calculi: in vitro and in vivo evaluation of a low-dose scanning protocol. Eur Radiol 2009; 19: Boll DT, Patil NA, Paulson EK, et al. Renal stone assessment with dual-energy multidetector CT and advanced postprocessing techniques: improved characterization of renal stone composition pilot study. Radiology 2009; 250: Thomas C, Krauss B, Ketelsen D, et al. Differentiation of urinary calculi with dual energy CT: effect of spectral shaping by high energy tin filtration. Invest Radiol 2010; 45: Matlaga BR, Kawamoto S, Fishman EK. Dual source computed tomography: a novel technique to determine stone composition. Urology 2008; 72: Primak AN, Ramirez Giraldo JC, Liu X, et al. Improved dual-energy material discrimination for dual-source CT by means of additional spectral filtration. Med Phys 2009; 36: Stolzmann P, Leschka S, Scheffel H, et al. Characterization of urinary stones with dual-energy CT. Invest Radiol 2010; 45: Vrtiska TJ, Takahashi N, Fletcher JG, Hartman RP, Yu L, Kawashima A. Genitourinary applications of dual-energy CT. AJR 2010; 194: Hanley JA, McNeil BJ. The meaning and use of the area under a receiver operating characteristic (ROC) curve. Radiology 1982; 143:29 36 AJR:198, June

7 Fung et al. APPENDIX 1: AJR Journal Club Study Guide Differentiation of Kidney Stones Using Dual-Energy CT With and Without a Tin Filter Margaret Mulligan*, Alan Mautz*, Joseph J. Budovec* Medical College of Wisconsin, Milwaukee, WI mmulliga@mcw.edu, amautz@mcw.edu, jbudovec@mcw.edu Introduction 1. What is the clinical question being asked? 2. Is the clinical question timely and relevant? 3. Is dual-energy CT available at your institution? Do you routinely differentiate urinary calculi at your institution? How are urinary stones characterized at your institution? 4. Were specific null and alternate hypotheses formulated? How would you write these hypotheses? Methods 5. Pure stones were used in the study. What are the potential implications to study outcomes from the use of pure stones? 6. Small stones were used for the in vivo study (~3 5 mm or smaller). What implications to practice can one make from a study using smaller stones versus the benefit of an ex vivo study that could have used larger stones? 7. What potential biases exist in this study? 8. What changes could have been made to the study design to improve the clinical relevance of this study? Results 9. The study documented that 80 and 140 kvp with a tin filter protocol showed significantly better differentiation than the two alternate protocols. Are these results applicable in your practice? 10. Given that uric acid can be differentiated from non uric acid in stone samples from all three protocols used in the study, how will you integrate the study findings into your practice? Physics 11. Briefly explain the principle underlying dual-energy CT. How can dual-energy CT be used to identify a substance? Discussion 12. What are the study limitations? 13. Does your institution differentiate uric acid from non uric acid stones? If so, what imaging protocol is used? 14. How would you design an ex vivo study of larger stones to determine the ability of dual-energy CT to differentiate uric acid from non uric acid stones? Background Reading 1. Thomas C, Krauss B, Ketelsen D, et al. Differentiation of urinary calculi with dual-energy CT: effect of spectral shaping by high energy tin filtration. Invest Radiol 2010; 45: Vrtiska TJ, Takahashi N, Fletcher JG, Hartman RP, Yu L, Kawashima A. Genitourinary applications of dual-energy CT. AJR 2010; 194: FOR YOUR INFORMATION For more information on Journal Clubs, see Evidence-Based Radiology: A Primer in Reading Scientific Articles in the July 2010 AJR at *Please note that the authors of the Study Guide are distinct from those of the companion article AJR:198, June 2012