Optimization of kvp and mas for Pediatric Low-Dose Simulated Abdominal CT: Is It Best to Base Parameter Selection on Object Circumference?

Transcription

1 Pediatric Imaging Original Research Reid et al. Parameter Selection for Pediatric Abdominal CT Downloaded from by on 2/3/18 from IP address Copyright ARRS. For personal use only; all rights reserved Pediatric Imaging Original Research Janet Reid 1 Jessica Gamberoni 2 Frank Dong 1 William Davros 1 Reid J, Gamberoni J, Dong F, Davros W Keywords: abdominal imaging, ALARA, kvp,, pediatric radiology, radiation dose DOI:1.2214/AJR Received October 26, 9; accepted after revision March, 1. 1 Department of Radiology, Section of Pediatric Radiology, Children s Hospital, Cleveland Clinic, 9 Euclid Ave., Hb6, Cleveland, OH Address correspondence to J. Reid (reidj@ccf.org). 2 Department of Biomedical Engineering, Case Western Reserve, Cleveland, OH. AJR 1; 19: X/1/194 1 American Roentgen Ray Society Optimization of kvp and for Pediatric Low-Dose Simulated Abdominal CT: Is It Best to Base Parameter Selection on Object Circumference? OBJECTIVE. The objective of our study was to determine the effect of and kvp reduction on pediatric phantoms based on patient circumference to optimize dose reduction and maintain image quality for abdominal CT. SUBJECTS AND METHODS. Three polymethylmethacrylate right cylindric CT dose index (CTDI) phantoms with diameters of 1, 16, and 32 cm simulated the abdomen of an infant, child, and adolescent, respectively. Using a National Institute of Standards & Technology ion chamber and Victoreen 66 electrometer, doses at centerline were recorded on a 16- MDCT scanner. Measurements were obtained in incremental steps from to 4 and from 8 to 14 kvp. Noise was calibrated to clinical images through a calibration factor. RESULTS. For phantoms of all circumferences, doses increased linearly with an increase in and by the power function of kvp n for increases in kvp. There was an associated decrease in noise for all circumferences and a sharp decrease at lower doses with a plateau at higher doses. Using a noise threshold of HU and a dose threshold of 2. cgy, a range of imaging parameters was established for each circumference from which technique optimization curves were created to determine optimal and kvp pairs. The mean measured dose was 2.43 ±.19 cgy. The mean measured noise was 29.3 ± 1.4 HU. CONCLUSION. For pediatric CT, the most accurate way to strike the balance between image quality and radiation dose is to adjust dose to abdominal circumference, not body weight or age. Our data support the use of technique optimization curves to optimize kvp and. A lthough radiation may be harmful, it serves a very important function in diagnostic imaging [1]. The radiologist is charged with finding a safe dose range for each patient whereby radiation is minimized for safety and maximized for diagnostic quality. Pediatric radiology has been at the forefront of radiation hygiene in CT for promoting reduced based on patient body weight. Image Gently, an alliance of multiple medical organizations with a focus on radiation reduction, proposes reduced tube current and peak kilovoltage for pediatric CT [2], but to date to our knowledge there are no standardized guidelines for combined kvp and reduction. Our hypothesis is that radiation can be reduced by a combined reduction in kvp and while preserving image noise and that technique optimization curves can be created to guide clinical imaging. Furthermore, because of wide variations in pediatric body shape, radiation doses are better calibrated to patient abdominal circumference than to age or weight. Subjects and Methods Our institutional review board waived the need for formal application. In this in vitro prospective study, phantoms of three circumferences were scanned in an MDCT scanner and radiation dose measurements were recorded by three members of the medical physics team. Right cylindric polymethylmethacrylate (PMMA) CT dose index (CTDI) phantoms (Model , Fluke Biomedical) with diameters of 1, 16, and 32 cm (or 3-, -, and 1-cm-equivalent abdominal circumference) were used to approximate the diameters of an infant, child, and adolescent, respectively (Fig. 1). Equivalent abdominal circumference was calculated as follows: 3.14 diameter because each phantom was a true right cylindric shape. The CTDI phantoms have multiple AJR:19, October 1 1

2 Reid et al. Downloaded from by on 2/3/18 from IP address Copyright ARRS. For personal use only; all rights reserved Fig. 1 To obtain estimates of CT dose index, right cylindric polymethylmethacrylate phantoms (Model , Fluke Biomedical) with diameters of 1, 16, and 32 cm (or 3-, -, and 1-cm-equivalent abdominal circumference, respectively) were used to approximate the diameters of an infant, child, and adolescent, respectively. Photograph shows phantom with 32-cm diameter. holes cut at the center and also at 12-, 3-, 6-, and 9-o clock locations about 1 cm from the phantom edge. These holes were mainly used for placing the dosimetry probe. Using a National Institute of Standards & Technology (NIST)-calibrated ion chamber and electrometer (Victoreen 66, Victoreen Instruments), radiation doses at the center and 12-o clock locations of the CTDI phantoms were recorded. Accuracy and reproducibility were verified according to the vendor s specifications. Measurements were obtained for various settings (, 1,, 3, 4 ) and kvp settings (8, 1, 1, 14 kvp) for each phantom. All measurements were A Measured Power fit acquired on a single 16-MDCT scanner (Sensation-16, Siemens Healthcare). Noise was recorded as the SD of the Hounsfield units at regions of interest (ROIs) within multiple images from each scan, with an ROI size of 1 mm 2. Each ROI was placed at the same location, close to the center of the image, using image-processing software (MATLAB, version 7.9., MathWorks). Noise values were obtained from a random sample of deidentified clinical images with acceptable image quality from pediatric CT scans acquired at the same kvp; effective, which is [tube current (ma) exposure time (s) / pitch factor]; and other scanning parameters. Using these noise values, we calculated a calibration factor (f factor) of 3. and applied it to the phantom data to create equivalent clinical data. Acceptable image quality was determined by a pediatric radiologist as the maximum noise visually tolerated. The reproducibility of the dose measurements was assessed by collecting 1 sets of dose data on the center of the 32-cm-diameter phantom scanned at, 1 kvp, and 12-mm collimation width. The noise measurement error was also assessed by placing the ROI near the center of the phantom images (away from the dose probe) using a MATLABbased image-processing tool and image noise was given as the SD of the pixel values in the ROI. A one-tailed paired Student s t test was used as the statistical method to determine whether the difference between the measurement results for two separate sets of scanning parameters was significant, where p <. was considered significant. A linear regression method was used to curve-fit the measured dose versus. For the measured dose at different kvp settings, a 1-based logarithm was taken for the measured dose and kvp before the linear regression was applied because the relationship between dose and kvp is better described by a power function. Results For all phantom circumferences, doses increased with increases in and kvp (Fig. 2). The relationship was linear for, but was approximately the nth power of kvp with n being between 2.49 and As dose increased, noise decreased for all circumferences with a sharp decrease at lower doses and a plateau effect at higher doses (Fig. 3). For example, at 1 kvp, if dose increased from 2 to 4 cgy, noise decreased by 3. HU, whereas when dose changed from 6 to 8 cgy, noise decreased by only.8 HU. While holding kvp constant and increasing, absorbed dose increased as circumference decreased (Fig. 4). Adding an increase in kvp to Figure 4 resulted in significant increases (p <.2, one-tailed paired Student s t test) in the absorbed dose for the same abdominal circumference (Fig. ). The dose reduction from combined and kvp reduction is shown in Table 1 and is illustrated in Figure 6. Using a noise threshold of HU and a dose threshold of 2. cgy (dose measured with 16-cm-diameter CTDI phantom with -cm-equivalent abdominal circumference), a range of imaging parameters was established for each abdominal circumference (Table 1). From these data, a technique optimization curve was established to determine optimal and kvp pairs for patient abdominal circumference (Fig. 7). Estimated optimal parameters were as follows: for the smallest patients (3- to 6-cm equivalent abdominal circumference), 8 kvp and Peak Kilovoltage (kvp) Fig. 2 Dose versus and kvp. A, Center absorbed dose (cgy) versus for three phantom abdominal circumferences (ACs): = 1 cm, X = cm, and 6 = 3 cm. B, Center absorbed dose (cgy) versus peak kilovoltage (kvp) for -cm-equivalent abdominal circumference at 4. B 116 AJR:19, October 1

3 Parameter Selection for Pediatric Abdominal CT Downloaded from by on 2/3/18 from IP address Copyright ARRS. For personal use only; all rights reserved Noise (HU) ; for 6- to 7-cm-equivalent abdominal circumference, 1 kvp and 6 ; and for 7- to 1-cm-equivalent abdominal circumference, 1 kvp, and 6. Error and Reproducibility The reproducibility of the dose measurements was assessed by collecting 1 sets of dose data on the center hole of the 32-cm-diameter phantom scanned at and 1 kvp with a 12-mm collimation width. The mean and SD of the measured dose was 2.43 ±.19 cgy; the SD was within 1% of the mean. 8 kvp 1 kvp 1 kvp 14 kvp Fig. 3 Center noise (HU) versus center dose (cgy) for phantom with -cmequivalent abdominal circumference at various peak kilovoltage (kvp) settings Center The noise measurement error was assessed by placing the ROI near the center of the phantom images (away from the dose probe). The results showed a mean and SD of noise at 29.3 ± 1.4 HU. The SD was well within 1% of the mean. The center dose versus and kvp was curve-fitted to the following functions: dose = a + b (1) dose = a (kvp) n (2) Equation 1 shows that dose has a linear relationship with. Both a and b are the Fig. 4 Absorbed dose (cgy) versus for three abdominal circumferences at 14 kvp: = 3 cm, X = cm, and 6 = 1 cm. Fig. Absorbed dose (cgy) versus for 3- cm abdominal-equivalent circumference at various peak kilovoltage (kvp) settings. = 8 kvp, = 1 kvp, = 1 kvp, = 14 kvp fitting parameters. Equation 2 indicates that dose varies with kvp by an n th power relationship. The coefficients were generated by a regression routine LINEST from Microsoft Excel (Table 2). The correlation coefficient, R 2, was >.97 for all regressions, indicating that it is reasonable to use the proposed functions in equations 1 and 2. From Table 2, the exponent (n) in the powerfitting curve for the dose versus kvp was 2.49 for the smallest phantom and 3.12 for the largest phantom, indicating the beam-hardening effect was stronger with the large phantom. Discussion Today, medical x-rays contribute to 48% of one s total radiation exposure [3]. CT accounts for approximately % of diagnostic x-ray procedures but up to 67% of medical radiation [4, ]. With the passing of the - year anniversary of the atomic bombs in Hiroshima, study of the survivors who were exposed to a mean dose of 4 msv has shown an increase in lifetime risk of cancers [6 9]. These results can be extrapolated to CT (with an organ dose of 1 3 msv) such that as many as 1. 2% of cancers may be attributable to radiation exposure from CT [8, 9]. The risk increases three- to fivefold for children and those with smaller cross-sectional areas [1]. Over the past two decades this information has been the stimulus for a widespread effort to decrease radiation dose in pediatric diagnostic imaging. A survey of pediatric radiology practice in showed that only 11% of radiologists reduced for chest CT and that none adjusted CT parameters for patient age or weight to reduce radiation exposure AJR:19, October 1 117

4 Reid et al. Downloaded from by on 2/3/18 from IP address Copyright ARRS. For personal use only; all rights reserved 4 8 kvp, 1-cm AC kvp, 1-cm AC 1 kvp, 1-cm AC 14 kvp, 1-cm AC 8 kvp, -cm AC 1 kvp, -cm AC 1 kvp, -cm AC 14 kvp, -cm AC 8 kvp, 3-cm AC 1 kvp, 3-cm AC 1 kvp, 3-cm AC 14 kvp, 3-cm AC Fig. 6 Composite graph of absorbed dose (cgy) versus for various abdominal circumferences (ACs) at various peak kilovoltage (kvp) settings. [11]. This served as the pilot data for the work by Frush et al. [12] in developing color-coded tables for dose reduction based on weightmatched tube current reduction. Weightbased reduction is now used in 98% of pediatric imaging facilities across North America and kvp reduction has increased from 4% to 48% in the past decade [13]. Most recently Image Gently, an alliance for radiation safety for pediatric imaging, was established to educate health care professionals and the lay public about the potential hazards of medical radiation and how to minimize exposure [2]. In children, decreasing the kvp from 14 to 1 reduces organ dose for abdominal CT by 4%, whereas decreasing it from 1 to 8 kvp reduces the organ dose by 6% [9, 11]. Adjustments in peak kilovoltage (kvp) have been shown primarily in adult clinical studies to provide superior contrast resolution, especially for CT angiography, whereby the kvp more closely approximates the k- edge of iodine [14, ]. By lowering the kvp, improved vascular enhancement is achieved [16, 17] with a reduction in IV contrast dose of up to 6% [14] and a 2% reduction in radiation dose [18]. Similar results have been shown for cerebral CT angiography with an increase in noise but a beneficial increase in the contrast-to-noise ratio [19]. The combination of automated tube current modulation, which adjusts to the contour of the patient s body, and low kvp has been shown to significantly reduce radiation exposure for pediatric patients undergoing cardiovascular imaging on 64-MDCT; however, choosing the correct combination of parameters on the basis of body weight remains challenging []. To date there are no standard clinical protocols for combined kvp and reduction in pediatric abdominal CT. The challenge remains to determine an effective organ dose range that balances image quality and image noise and then to manipulate tube current and peak kilovoltage while conforming to the range. In vitro evaluation of radiation doses in pediatric abdominal CT is limited by the lack of TABLE 1: Absorbed and Noise (HU) Values at Various Effective a and Peak Kilovoltage (kvp) Settings in Phantoms of Different Abdominal Circumferences Abdominal Circumference of Phantom 8 kvp 1 kvp 1 kvp 14 kvp CTDI (V) Noise (v) Liver (n) CTDI (V) Noise (v) Liver (n) CTDI (V) Noise (v) Liver (n) CTDI (V) Noise (v) Liver (n) 1 cm cm cm Note Values shown in boldface are optimal values, which were chosen on the basis of noise (n) < (HU) and dose (V) 2 cgy or cgy. Values shown in italics represent the borderline noise and dose levels. Unacceptable noise and dose values are set in Roman with neither boldface nor italics. CTDI = CT dose index. a Effective = tube current (ma) exposure time (s) / pitch factor. 118 AJR:19, October 1

5 Parameter Selection for Pediatric Abdominal CT Downloaded from by on 2/3/18 from IP address Copyright ARRS. For personal use only; all rights reserved kvp (6 ) 8 kvp (1 ) 14 kvp ( ) 1 kvp (6 ) Abdominal Circumference (cm) a variety of phantoms of different sizes that can represent the wide variation of sizes of patients who range from to 18 years. The standard 16-cm phantom is inadequate for neonates and large adolescents. In addition, current phantom design does not accurately approximate the shape of a pediatric patient s body or the tissue characteristics [1]. These obstacles have prevented stringent testing to determine kvp pairs that can be applied to clinical imaging. Finally, the work that has gone into pediatric dose reduction has adjusted dose to patient weight or, in some cases, to patient age [21 23]. The wide heterogeneity in patient weight for age argues for improved accuracy in matching dose to patient abdominal circumference rather than to weight. Boone et al. [24] performed an in vitro study on acrylic phantoms of different sizes in which they measured noise, dose, and contrast values generated by changing the kvp from 8 to 14 and the from 1 to 3. That study was one of the first to address the importance of abdominal circumference rather than body weight using a 16-MDCT scanner. The main weakness of that study was that the noise and Fig. 7 Technique optimization curve shows optimal peak kilovoltage (kvp) setting and dose pairs for absorbed dose and patient abdominal circumference. TABLE 2: Linear Regression Coefficients (a, b, and R 2 ) for Dose Versus Effective a and (a, n, and R 2 ) Dose Versus Peak Kilovoltage (kvp) Abdominal Circumference of Phantom Dose = a + b (equation 1) Dose = a kvp n (equation 2) a b R 2 a n R 2 1 cm E cm E cm E Note Equation 1 shows that dose has a linear relationship with. Both a and b are the fitting parameters. Equation 2 indicates that dose varies with kvp by an nth power relationship (E-4 = xxx, E-6 = 1 4 ). a Effective = tube current (ma) exposure time (s) / pitch factor. dose tables generated cannot be applied easily to the clinical arena and, as recognized by the authors, that more work is necessary to prove the results on vendor platforms other than GE Healthcare [24]. Two studies evaluated the effects of varying kvp and on 4-MDCT scanners. Cody et al. [2] used anthropomorphic phantoms and generated dose reduction tables recommended for clinical use. The main weaknesses of the tables were the -mm slice thickness and axial mode used for scanning and the need to extrapolate what the values would be when scanning in the helical mode, which is more commonly used clinically. Siegel and colleagues [26] were the first to address the smallest of pediatric patients in studying an array of acrylic phantoms from 8 to 32 cm using a limited range of on a 4-MDCT scanner manufactured by Siemens Healthcare. Unfortunately the information displayed graphically is not easily transferred to the clinical arena. Our results suggest that diagnostic-quality images can be obtained within a dose range of cgy (measured with 16-cm-diameter CTDI phantom) for pediatric abdominal CT. Within this range, a significant reduction in both tube current and peak kilovoltage can be achieved with the lower limits determined by noise. In addition, for every patient, the upper limits for tube current and peak kilovoltage are determined by dose. Although, in general, these upper and lower limits show a gradual increase with increasing patient age, idiosyncrasies are determined more by a patient s body shape. For abdominal CT, because radiation attenuation is greatly affected by patient thickness, we measured doses against body shape as dictated by abdominal circumference rather than weight. Taking into account four parameters abdominal circumference, CTDI volume,, and kvp and determining a range of doses with the lower limit governed by noise and the upper limit governed by dose, we constructed a simple technique optimization curve that can be posted at the scanner or online, with the ultimate goal to guide clinical practice. The major limitation of this study is the need to extrapolate optimal scanning parameters for actual pediatric patients from phantom work. Further work with human patients is necessary to validate this work. Another limitation relates to the right cylindric geometry of the phantoms and inherent properties leading to higher attenuation than human abdominal tissues. Our study made no allowance for tissue-mimicking materials, but instead created an f factor to extrapolate to human subjects based on a sample of pediatric patients who underwent scanning with identical parameters. This study was conducted on a single vendor s platform, which makes it difficult to transfer our results to other platforms. Although not limited to this study, to date to our knowledge there are no established standards in the scientific community for low contrast conspicuity. For this reason, optimal noise values were selected on the basis of personal preference. Further study would involve the construction of pediatriclike oval phantoms and a detailed evaluation of the contrastto-noise ratio as well as the dose. Radiation dose reduction for pediatric abdominal CT remains an important but surmountable challenge. The most accurate way to strike the fine balance between image quality and radiation dose is through consideration of individual habitus; for abdominal CT the most important consideration is the patient s abdominal circumference. The two technical parameters most effective in reducing dose are kvp and with a dramatic decrease in dose related to their combined reduction. Our data support the use of AJR:19, October 1 119

6 Reid et al. technique optimization curves to allow more accurate selection of kvp and based on abdominal circumference for pediatric abdominal CT. 9. Nickoloff EL, Alderson PO. Radiation exposures to patients from CT: reality, public perception, and policy. AJR 1; 177: Strauss KJ, Goske MJ, Frush DP, Butler PF, Mor- tion doses on 16-MDCT aortography. AJR 6; 187:1266; [web]:w49 W Bahner ML, Bengel A, Brix G, et al. Improved vascular opacification in cerebral computed to- Downloaded from by on 2/3/18 from IP address Copyright ARRS. For personal use only; all rights reserved References 1. Amis ES Jr, Butler PF, Applegate KE, et al.; American College of Radiology. American College of Radiology white paper on radiation dose in medicine. J Am Coll Radiol 7; 4: Goske MJ, Applegate KE, Boylan J, et al. The Image Gently campaign: increasing CT radiation dose awareness through a national education and awareness program. Pediatr Radiol 8; 38: Schauer DA, Linton OW. National Council on Radiation Protection and Measurements report shows substantial medical exposure increase. Radiology 9; 23: Frush DP, Donnelly LF, Rosen NS. Computed tomography and radiation risks: what pediatric health care providers should know. Pediatrics 3; 112: Mettler FA Jr, Wiest PW, Locken JA, et al. CT scanning: patterns of use and dose. J Radiol Prot ; : Brenner D, Elliston C, Hall E, Berdon W. Estimated risks of radiation-induced fatal cancer from pediatric CT. AJR 1; 176: Brenner DJ, Doll R, Goodhead DT, et al. Cancer risks attributable to low doses of ionizing radiation: assessing what we really know. Proc Natl Acad Sci U S A 3; 1:13,761 13, Brenner DJ, Hall EJ. Computed tomography: an increasing source of radiation exposure. N Engl J Med 7; 37: rison G. Image Gently Vendor Summit: working together for better estimates of pediatric radiation dose from CT. AJR 9; 192: Paterson A, Frush DP, Donnelly LF. Helical CT of the body: are settings adjusted for pediatric patients? AJR 1; 176: Frush DP, Soden B, Frush KS, Lowry C. Improved pediatric multidetector body CT using a size-based color-coded format. AJR 2; 178: Arch ME, Frush DP. Pediatric body MDCT: a -year follow-up survey of scanning parameters used by pediatric radiologists. AJR 8; 191: Kalva SP, Sahani DV, Hahn PF, Saini S. Using the K-edge to improve contrast conspicuity and to lower radiation dose with a 16-MDCT: a phantom and human study. J Comput Assist Tomogr 6; 3: Huda W, Vance A. Patient radiation doses from adult and pediatric CT. AJR 7; 188: Szucs-Farkas Z, Kurmann L, Strautz T, et al. Patient exposure and image quality of low-dose pulmonary computed tomography angiography: comparison of 1- and 8-kVp protocols. Invest Radiol 8; 43: Wintersperger B, Jakobs T, Herzog P, et al. Aorto-iliac multidetector-row CT angiography with low kv settings: improved vessel enhancement and simultaneous reduction of radiation dose. Eur Radiol ; : Nakayama Y, Awai K, Funama Y, et al. Lower tube voltage reduces contrast material and radia- mography angiography with 8 kvp. Invest Radiol ; 4: Herzog C, Mulvihill DM, Nguyen SA, et al. Pediatric cardiovascular CT angiography: radiation dose reduction using automatic anatomic tube current modulation. AJR 8; 19: Donnelly LF, Emery KH, Brody AS, et al. Minimizing radiation dose for pediatric body applications of single-detector helical CT: strategies at a large Children s Hospital. AJR 1; 176: Huda W, Scalzetti EM, Levin G. Technique factors and image quality as functions of patient weight at abdominal CT. Radiology ; 217: Thomas KE, Wang B. Age-specific effective doses for pediatric MSCT examinations at a large children s hospital using DLP conversion coefficients: a simple estimation method. Pediatr Radiol 8; 38: Boone JM, Geraghty EM, Seibert JA, Wootton- Gorges SL. Dose reduction in pediatric CT: a rational approach. Radiology 3; 228: Cody DD, Moxley DM, Krugh KT, O Daniel JC, Wagner LK, Eftekhari F. Strategies for formulating appropriate MDCT techniques when imaging the chest, abdomen, and pelvis in pediatric patients. AJR 4; 182: Siegel MJ, Schmidt B, Bradley D, Suess C, Hildebolt C. Radiating dose and image quality in pediatric CT: effect of technical factors and phantom size and shape. Radiology 4; 233: 22 1 AJR:19, October 1