EFFECTIVE DOSE ESTIMATION FROM ORGAN DOSE MEASUREMENTS IN FAST-kV SWITCH DUAL ENERGY COMPUTED TOMOGRAPHY

EFFECTIVE DOSE ESTIMATION FROM ORGAN DOSE MEASUREMENTS IN FAST-kV SWITCH DUAL ENERGY COMPUTED... Abstract The purpose of this study was to validate a novel approach to estimating effective dose (E) in ‘fast-kV switch dual energy computed tomography’ using MOSFET detectors. The effective energy of the combined dual energy environment was characterized with the dual energy CT scanner and then MOSFETs were calibrated matching to the effective energy of the dual energy CT beam with a conventional CT beam. The calibration method was then experimentally validated by comparing the dose between MOSFET and an ion chamber (IC) using a standard CTDI body phantom. The measured doses of the MOSFET and IC were 17.1 mGy ± 3.8% and 17.1 mGy ± 0.4%, respectively. To measure organ doses, an adult anthropomorphic phantom loaded with 18 MOSFET detectors was scanned using a standard fast-kV switch dual energy abdomen/pelvis CT protocol. E was calculated by applying ICRP 103 tissue weighting factors as well as partial volume correction factors for organs that were not completely covered by the protocol field-of-view. E from the dual energy abdomen/pelvis CT was calculated to be 17.8 mSv ± 11.6%. This calculation was then compared to E from dose length product method, which yielded 14.62 mSv. INTRODUCTION Dual energy computed tomography (DECT) is a method of computed tomographic (CT) imaging in which two CT scans are performed on the patient using differing X-ray tube potentials which allow for energy-selective reconstruction. Reconstruction in CT relies on measuring the line integral of the linear attenuation coefficient along each projection. The linear attenuation coefficient depends primarily on atomic number and electron density as well as the energy of the X-ray photons. This allows the attenuation coefficient of a material to be modeled as a function of energy using a set of basis functions. These basis functions provide information on how attenuation for a material will change between two energy spectra. In DECT, a second measurement of attenuation at a different energy provides an additional dataset. This additional dataset provides a measurement of the change in attenuation between the two energy projections. This information can then be compared to known changes of attenuation provided by the basis functions for various materials in order to differentiate material composition(1, 2). Acquiring datasets at two separate energies allows for a number of advantages over conventional CT. The two datasets provide attenuation information corresponding to each energy that can then be used for enhanced material discrimination, improved image registration, and a reduction of beam hardening effects(3, 4). There are currently two primary methods used to create a dual energy environment available as commercial products. The first of these methods is referred to as Dual Source Computed Tomography (DSCT), in which two separate X-ray tubes are operated at different energies simultaneously. The two sources are operated at differing tube currents and each have a different level of filtration. The second method of DECT makes use of a single X-ray tube that rapidly switches from low kV to high kV between consecutive projections, alternating at a rate of up to 4.8 kHz. Due to the fast rate of kV switching the single X-ray tube is operated at a constant current, with longer dwell time for the low kV projections to compensate for higher attenuation. For this study we used a fast-kV switching system to create our dual energy environment, with tube potential alternating between 80 and 140 kV. A visualization of this system can be seen in Figure 1. The key advantage of fast-kV switch DECT compared to DSCT is a reduction in motion and misregistration artefacts due to the rapid switching between energy projections; however, it offers limited control of tube current for the different projections. Figure 1. View largeDownload slide Illustration of fast-kV switching DECT. The alternating colors represent consecutive projections alternating between high and low kV. Figure 1. View largeDownload slide Illustration of fast-kV switching DECT. The alternating colors represent consecutive projections alternating between high and low kV. Effective dose (E) is a quantity used to represent the stochastic health risk to the whole body from non-uniform exposure to ionizing radiation(5, 6). The quantity is derived by the summation of organ doses multiplied by a corresponding tissue-specific weighting factor. These tissue weighting factors are described by the International Commission on Radiological Protection (ICRP) to represent the fraction of health risk from a specific tissue or organ. Physical measurements of individual organ dose due to an imaging procedure are generally estimated using small internal dosemeters such as thermoluminescence dosemeters (TLD) or metal oxide semiconductor field-effect transistor (MOSFET) detectors. These dosemeters have an energy dependence which gives rise to difficulty when calibrating the detectors in a dual energy environment. Specifically, between 80 and 140 kV there is a 20% difference in MOSFET response. Currently the few publications that attempt to estimate E from physical organ dose measurements in a dual energy environment pertain specifically to dual source computed tomography(7–9). There is currently a lack of publications pertaining to E from dual energy CT acquired via fast kV switching. The objective of this study was to test a novel method of estimating organ dose by using the effective energy (EE) of the combined dual energy environment to calibrate our internal dosemeters for dose measurements. This process can be viewed as the combination of three primary goals: (1) Validating the novel approach for dose equivalence between the MOSFET method and an ion chamber (IC) which serves as the gold standard; (2) measuring organ dose using MOSFET detectors in an anthropomorphic phantom, and estimating E in clinical dual energy abdomen/pelvis protocol by applying ICRP 103 tissue weighting factors(10); and (3) finally evaluating the applicability of the DLP method for E estimation against clinical data. MATERIALS AND METHODS This study was conducted using a GE Discovery CT 750HD (GE Heathcare, Inc., Waukesha, WI) implementing Gemstone Spectral Imaging (GSI) on an abdomen/pelvis protocol. This GSI protocol is a fast-kV switching dual energy scan, rapidly alternating between 80 and 140 kV. Our experimental method includes six primary components: (1) Characterize the EE of the dual energy environment; (2) estimating corresponding f-factor for soft tissue; (3) calibrating our MOSFET detectors using a conventional CT beam with EE as close to that of the dual energy environment; (4) validating our calibration method by comparing dose from MOSFET and IC using a CTDI body phantom in the dual energy environment; (5) estimating E for an abdomen/pelvis DECT scan using an anthropomorphic phantom and applying ICRP 103 tissue weighting factors; and (6) estimating E using AAPM dose length product (DLP) method for comparison. MOSFET detectors provide dose exposure information by measuring change in the voltage necessary to initiate current flow in the detector circuit, known as the device threshold voltage. Ionizing radiation produces electron–hole pairs in the silicon substrate. These pairs move to the oxide-silicon interface which results in a negative threshold voltage shift. The MOSFET detectors have been shown to have a linear energy dependence(11), and have previously been validated against thermoluminescent detectors in 120 kVp CT beams.(12) Characterizing EE of combined dual energy environment The EE of an X-ray beam is defined as the monoenergetic photon energy that would produce the same first half-value layer (HVL) as the beam itself in a given material. By directly measuring the HVL (in mm AL) of the combined dual energy beams we can calculate a linear attenuation coefficient using the simple relation:   μ=ln2HVL (1) Once this value is known we can use attenuation data such as NIST X-ray attenuation databases(13) to find our EE. The GSI protocol was performed in diagnostic mode, which allows the X-ray tube to remain stationary above the patient couch. The combined dual energy beam was then measured using a 6 ml IC (10 × 5–6, Radcal, Monrovia, CA). This measurement was then repeated four times adding different thickness of thin aluminum sheets between the X-ray tube and IC. The IC measurements were plotted vs. thickness of added aluminum filtration in order to perform exponential regression analysis, which was used to calculate a HVL of 5.02 mm Al. Estimating f-factor for organs It is necessary to specify in our calibration that we seek to find dose in soft tissue. This is accomplished by using an f-factor, which translates the exposure we measure in air to dose in a medium of interest, in this case, soft tissue(14). The equation for f-factor can be seen below:   f-factor=0.873(μenρ)ST(μenρ)AIR[cGyR] (2)where (μenρ)ST(μenρ)AIR represents the ratio of mass energy absorption coefficients for soft tissue and air. The f-factor value was later used to formulate calibration factors for each individual MOSFET detector. A separate f-factor of was calculated to determine dose to bone surface by using the same equation with (μenρ)Bone substituted for (μenρ)ST. It is also worth noting that the soft tissue f-factor changes very little within the EE range of the separate low and high kV beam, which is roughly 40–55 kV. This relative stability in f-factor can be seen in Figure 2. Calibration of MOSFET detectors MOSFET detectors were calibrated by matching the calibration beam quality (HVL 5.42 mm Al, EE 44.7 keV) as close as possible to the GSI beam quality (HVL 5.02 mm Al, EE 42.8 keV). HVL was measured by using a Piranha multifunction X-ray meter (RTI Electronics, Mölndal, Sweden). The MOSFET detectors were placed adjacent to the active area of a 6 ml IC (10 × 5–6, Radcal, Monrovia, CA), and then irradiated using the 80 kV calibration beam in order to calculate calibration factors for the 18 individual MOSFET detectors. These calibration factors convert the raw millivolt readings of the MOSFET detectors into absorbed dose in soft tissue (rad). The equation for these calibration factors can be seen in the following equation.   CF=MOSFETmeasurement((ionchamberreading)×(ionchambercorrectionfactor)×(f-factor))[mVcGy] (3) The calibration method was tested by inserting two MOSFET detectors along with a 0.18 ml IC (10 × 5–0.18, Radcal, Monrovia, CA) in the center cavity of a CTDI body phantom (West Physics, Atlanta, GA). The 0.18 ml IC was chosen based on two criteria. The first was its fairly small energy dependence over the energy range produced in our dual energy protocol. The second benefit of the 0.18 ml IC is that its small size allows for the detector to be easily placed into our CTDI phantom. Our ICs (Radcal model 10 × 5–6 ml, 0.18 ml) were calibrated at University of Wisconsin Radiation Dosimetry Laboratory (AAPM accredited dosimetry laboratory) that is NIST traceable. In Figure 3 we can see the scout image from the GSI abdomen/pelvis protocol. The CTDI body phantom was scanned using the dual energy abdomen/pelvis protocol. Detector readings were averaged over three scans and standard deviation was recorded. These results experimentally confirm dose equivalence between our calibrated MOSFET detectors and IC. Figure 3. View largeDownload slide Scout image showing the scanning region of the abdomen/pelvis protocol. Figure 3. View largeDownload slide Scout image showing the scanning region of the abdomen/pelvis protocol. Organ dose measurement After verifying our calibration, the MOSFETs were inserted into an adult anthropomorphic phantom (Model 702-BR190, CIRS) in order to measure organ doses. This phantom is composed of 39 slabs of thickness 2.5 cm, and simulates the radiological behavior of different tissue types including soft tissue, lung, bone, brain, and spinal cord and disks. These slabs have many numbered predrilled holes that correspond to different tissue and organ locations listed in the phantom user manual. The phantom is designed to represent a patient of 73 kg weight and 173 cm height. The phantom was loaded with 18 MOSFET detectors were inserted into the phantom in locations representative or tissues of interest that would be exposed during the procedure. These locations were esophagus, lungs, heart, liver, stomach, pancreas, gall bladder, spleen, small intestine, large intestine, kidney, ovary, uterus, bladder, prostate, testis, bone marrow and bone surface. In addition to these locations, an additional detector was used to record skin entrance dose on the abdomen surface. The parameters of the dual energy protocol were fast kV switch between 80 and 140 kV, 640 mA. The loaded phantom was scanned four times using the dual energy protocol. Uncertainty was evaluated using a combined expanded uncertainty with a coverage factor of 2 (k = 2). Type B uncertainties were included in to account for observed variation in our calibration, as well as EE discrepancy. In order to accurately reflect dose to specific tissues several partial volume correction factors (PVCF) were necessary to adjust for organs/tissues that are only partially irradiated in an abdomen/pelvis protocol(15, 16). The organs in our study that required this correction were bone marrow, bone surface, skin, lung and esophagus. The dose to bone marrow was corrected using published estimations of the distribution of active bone marrow in adults(17). This allowed us to scale our measured dose in locations corresponding to sites of bone marrow storage. Dose to bone surface was corrected using percentages of skeletal mass for the bone structures measured in our study(18). The dose to skin surface was calculated using ‘the rule of nines’(19). This is traditionally used in medicine to estimate the percentage of a persons skin affected by burn trauma. This rule separates the skin surface into nine anatomical regions and applies a surface area percentage fraction to each. The surface area factors used for this experiment were chest—0.09, abdomen—0.09 and back of torso—0.18. Using this method we calculated 36% of the total skin surface area was exposed during the scan. Volumetric corrections for lung and esophagus were performed using a method previously derived by our lab dealing specifically with our adult anthropomorphic phantom(15). This method describes the percentage of total lung and esophagus volume present in each slice of the phantom. Using this method, we applied correction factors for lung and esophagus dose based on which sections of the phantom were exposed during the abdomen/pelvis protocol. The remainder of the organs in our study were assumed to have unity PVCF because their entire volume was covered completely by the scan region. Effective dose was then calculated using ICRP 103 tissue weighting factors, WT, along with radiation weighting factors, WR. The formula we used can be seen in the following equation:   E=∑TPVCFT×WT×WR×DT (4)where DT represents the dose measurement from MOSFET detectors. In addition to calculating E from our organ dose measurements, we also compare our result to E calculated using AAPM DLP method(20). This method multiplies the DLP output by the CT system and multiplies it by a coefficient dependent on the age of the patient as well as the region of the body scanned during the CT examination. The DLP of the dual energy abdomen/pelvis protocol was 974.62 mGy cm, and the coefficient was 0.015.   EDLP≈k(mSvmGycm)×DLP(mGycm) RESULTS The EE of the combined dual energy beams was found to be 42.8 kV, corresponding to an f-factor of 0.923 for soft tissue, and 4.937 for bone. The EE of the 80 kV beam used in calibration was found to be 44.7 kV. A good agreement between MOSFET and IC was obtained: 1.71 rad ± 3.8% vs. 1.71 rad ± 0.5%. Organ dose measurements recorded by MOSFETs can be seen below in Figure 4. Using these measurements combined with ICRP 103 tissue weighting factors, the effective dose was calculated as 17.8 mSv ± 11.6%. Effective dose estimated using DLP method was 14.6 mSv. The uncertainty budget for the Effective Dose calculation can be seen in Table 1. Figure 4. View largeDownload slide MOSFET organ dose measurements. Figure 4. View largeDownload slide MOSFET organ dose measurements. Table 1. Uncertainty budget.   Type A (%)  Type B (%)  IC NIST calibration  —  1.9  IC monitor controller  —  0.5  HVL measurement  —  3.9  MOSFET calibration  3.8  —  f-factor  —  1.1  Effective dose calculation  0.3  —  Combined standard uncertainty  5.8  Expanded uncertainty (k = 2)  11.6    Type A (%)  Type B (%)  IC NIST calibration  —  1.9  IC monitor controller  —  0.5  HVL measurement  —  3.9  MOSFET calibration  3.8  —  f-factor  —  1.1  Effective dose calculation  0.3  —  Combined standard uncertainty  5.8  Expanded uncertainty (k = 2)  11.6  Table 1. Uncertainty budget.   Type A (%)  Type B (%)  IC NIST calibration  —  1.9  IC monitor controller  —  0.5  HVL measurement  —  3.9  MOSFET calibration  3.8  —  f-factor  —  1.1  Effective dose calculation  0.3  —  Combined standard uncertainty  5.8  Expanded uncertainty (k = 2)  11.6    Type A (%)  Type B (%)  IC NIST calibration  —  1.9  IC monitor controller  —  0.5  HVL measurement  —  3.9  MOSFET calibration  3.8  —  f-factor  —  1.1  Effective dose calculation  0.3  —  Combined standard uncertainty  5.8  Expanded uncertainty (k = 2)  11.6  DISCUSSION Scattered radiation in the phantom will lead to more attenuation at lower energies of the spectrum. This will have the result of slightly increasing the EE of the spectrum reaching the MOSFET detectors. After calibration, the purpose of the EE is to calculate the f-factors used to translate exposure into dose. In Figure 2 we see that this value is very stable across the EE range we are working with. We feel that the effect of scatter in the phantom will be negligible to measured organ dose. Figure 2. View largeDownload slide Soft tissue f-factor vs. energy in the range of our projected effective energy. Values calculated using Equation 2 and NIST X-ray attenuation databases. Figure 2. View largeDownload slide Soft tissue f-factor vs. energy in the range of our projected effective energy. Values calculated using Equation 2 and NIST X-ray attenuation databases. Furthermore, in our experiment we note a discrepancy in EE of the dual energy environment and the calibration beam. In order to account for this in our results we investigate the difference in dose calculation at the two EE values by comparing the uncertainty in f-factor calculation. This value is included in our uncertainty budget. One limitation of our study is the use of only one dosemeter in each organ. Ideally we would like to sample several points throughout the organs of interest to see a more detailed distribution of dose in each. The choice to use one dosemeter per organ was made considering the technical limitation of the MOSFET system supporting a maximum of 20 detectors. For the purpose of this study we felt the information gained from sampling a greater number organs was preferable to measuring several points in only a few. When looking at the air gaps between slices in our phantom, it is important to note that the active area of the detector is placed inside predrilled holes in the slabs, and taped to secure their positioning. This eliminates the possibility of the MOSFET being directly exposed to the primary beam. We feel the small gaps caused by the MOSFET wires will have a negligible contribution to measured dose. It is also worth noting that the phantoms used in our study are not perfectly representative of the reference male used in E formalism in ICRP publication 110. Where the reference male weighs 70 kg with height 170 cm, our phantom is representative of a person with weight 73 kg and height 173 cm. The experimental validation of our EE calibration of MOSFETs shows that it is a viable option for taking organ dose measurements in DECT. This novel approach to internal dosemeter calibration successfully circumnavigates the inherent challenge of taking physical measurements in a dual energy environment. By calibrating in a conventional CT beam with EE equal to that of the DECT it is possible to minimize the effect of energy dependence in the dosemeter. Currently the DLP method is most often used to estimate E from CT examinations. Our method of E estimation yields a value of 17.8 ± 2.1 mSv, which corresponds to a 21.9% difference compared to the DLP method. Furthermore, E estimated using organ dose measurements depends heavily on what organs are included in the calculation. CONCLUSION This study shows that using the EE of the combined dual energy beams is a viable way to calibrate MOSFET detectors in a dual energy environment. The method is useful particularly in ‘fast-kV switching systems’ where the high and low kV beams cannot be separated for individual treatment. The E from the DLP method differed 22% from the MOSFET method; however, the DLP method still provides a viable alternative to the more complex and labor intensive MOSFET method, particularly in busy clinical setting. Moving forward we would like to apply this method of E estimation to other dual energy systems such as dual source CT. FUNDING This study was funded, in part, by the US Nuclear Regulatory Commission (Grant no. NRC-HQ-12-G-38-0022). REFERENCES 1 Alvarez, R. E. and Macovski, A. Energy-selective reconstructions in x-ray computerised tomography. Phys. Med. Biol.  21( 5), 733 ( 1976). Google Scholar CrossRef Search ADS PubMed  2 Macovski, A., Alvarez, R. E., Chan, J. H., Stonestrom, J. P. and Zatz, L. M. Energy dependent reconstruction in X-ray computerized tomography. Comput. Biol. Med.  6( 4), 325– 336 ( 1976). Google Scholar CrossRef Search ADS PubMed  3 Li, B. Dual-energy CT with fast-kVp switching and its applications in orthopedics. OMICS J. Radiol.  2, 1– 8 ( 2013). Google Scholar CrossRef Search ADS   4 Yu, L., Leng, S. and McCollough, C. H. Dual-energy CT–based monochromatic imaging. Am. J. Roentgenol.  199( 5_supplement), S9– S15 ( 2012). Google Scholar CrossRef Search ADS   5 Jacobi, W. The concept of the effective dose a proposal for the combination of organ doses. Radiat. Environ. Biophys.  12( 2), 101– 109 ( 1975). Google Scholar CrossRef Search ADS PubMed  6 ICRP 1977. Recommendations of the ICRP. Ann. ICRP , vol. 1, ( New York: Pergammon). 7 Paul, J., Banckwitz, R., Krauss, B., Vogl, T. J., Maentele, W. and Bauer, R. W. Estimation and comparison of effective dose (E) in standard chest CT by organ dose measurements and dose-length-product methods and assessment of the influence of CT tube potential (energy dependency) on effective dose in a dual-source CT. Eur. J. Radiol.  81( 4), e507– e512 ( 2012). Google Scholar CrossRef Search ADS PubMed  8 Schenzle, J. C., Sommer, W. H., Neumaier, K., Michalski, G., Lechel, U., Nikolaou, K., Becker, C. R., Reiser, M. F. and Johnson, T. R. Dual energy CT of the chest: how about the dose? Invest. Radiol.  45( 6), 347– 353 ( 2010). Google Scholar PubMed  9 Henzler, T., Fink, C., Schoenberg, S. O. and Schoepf, U. J. Dual-energy CT: radiation dose aspects. Am. J. Roentgenol.  199( 5_supplement), S16– S25 ( 2012). Google Scholar CrossRef Search ADS   10 ICRP 2007. The 2007 Recommendations of the Internal Commission on Radiological Protection. ICRP Publication 103, Ann. ICRP. vol. 37, ( New York: Pergammon). 11 Ding, L. ( 2013). The use of a novel radiation detector on quantifying PET/nuclear medicine occupational and non-occupational doses and calibration of MOSFET radiation detectors against effective energy. Doctoral dissertation, Duke University. 12 Yoshizumi, T. T., Goodman, P. C., Frush, D. P., Nguyen, G., Toncheva, G., Sarder, M. and Barnes, L. Validation of metal oxide semiconductor field effect transistor technology for organ dose assessment during CT: comparison with thermoluminescent dosimetry. Am. J. Roentgenol.  188( 5), 1332– 1336 ( 2007). Google Scholar CrossRef Search ADS   13 Hubbell, J. H. and Seltzer, S. M., 1995. Tables of X-ray mass attenuation coefficients and mass energy-absorption coefficients 1 keV to 20 MeV for elements Z = 1 to 92 and 48 additional substances of dosimetric interest (No. PB—95-220539/XAB; NISTIR—5632). National Inst. of Standards and Technology-PL, Gaithersburg, MD. Ionizing Radiation Div. 14 Hendee, W. R. Medical Radiation Physics , 2nd edn. Chicago: Year Book Medical Publishers) 171 ( 1979). 15 Januzis, N., Nguyen, G., Hoang, J. K., Lowry, C. and Yoshizumi, T. T. A novel method of estimating effective dose from the point dose method: a case study—parathyroid CT scans. Phys. Med. Biol.  60( 5), 1763 ( 2015). Google Scholar CrossRef Search ADS PubMed  16 Januzis, N., Nguyen, G., Frush, D. P., Hoang, J. K., Lowry, C. and Yoshizumi, T. T. Feasibility of using the computed tomography dose indices to estimate radiation dose to partially and fully irradiated brains in pediatric neuroradiology examinations. Phys. Med. Biol.  60( 14), 5699 ( 2015). Google Scholar CrossRef Search ADS PubMed  17 Cristy, M. Active bone marrow distribution as a function of age in humans. Phys. Med. Biol.  26( 3), 389 ( 1981). Google Scholar CrossRef Search ADS PubMed  18 Valentin, J. Basic anatomical and physiological data for use in radiological protection: reference values: ICRP Publication 89. Ann. ICRP  32( 3), 1– 277 ( 2002). Google Scholar CrossRef Search ADS   19 Hettiaratchy, S. and Papini, R. Initial management of a major burn: II—assessment and resuscitation. Br. Med. J.  329( 7457), 101– 103 ( 2004). Google Scholar CrossRef Search ADS   20 McCollough, C. et al.  . The measurement, reporting, and management of radiation dose in CT. Rep. AAPM Task Group  23, 1– 28 ( 2008). © The Author(s) 2018. Published by Oxford University Press. All rights reserved. For Permissions, please email: journals.permissions@oup.com This article is published and distributed under the terms of the Oxford University Press, Standard Journals Publication Model (https://academic.oup.com/journals/pages/about_us/legal/notices) http://www.deepdyve.com/assets/images/DeepDyve-Logo-lg.png Radiation Protection Dosimetry Oxford University Press

EFFECTIVE DOSE ESTIMATION FROM ORGAN DOSE MEASUREMENTS IN FAST-kV SWITCH DUAL ENERGY COMPUTED TOMOGRAPHY

Loading next page...
 
/lp/ou_press/effective-dose-estimation-from-organ-dose-measurements-in-fast-kv-LKgjGw30jX
Publisher
Oxford University Press
Copyright
© The Author(s) 2018. Published by Oxford University Press. All rights reserved. For Permissions, please email: journals.permissions@oup.com
ISSN
0144-8420
eISSN
1742-3406
D.O.I.
10.1093/rpd/ncy072
Publisher site
See Article on Publisher Site

Abstract

Abstract The purpose of this study was to validate a novel approach to estimating effective dose (E) in ‘fast-kV switch dual energy computed tomography’ using MOSFET detectors. The effective energy of the combined dual energy environment was characterized with the dual energy CT scanner and then MOSFETs were calibrated matching to the effective energy of the dual energy CT beam with a conventional CT beam. The calibration method was then experimentally validated by comparing the dose between MOSFET and an ion chamber (IC) using a standard CTDI body phantom. The measured doses of the MOSFET and IC were 17.1 mGy ± 3.8% and 17.1 mGy ± 0.4%, respectively. To measure organ doses, an adult anthropomorphic phantom loaded with 18 MOSFET detectors was scanned using a standard fast-kV switch dual energy abdomen/pelvis CT protocol. E was calculated by applying ICRP 103 tissue weighting factors as well as partial volume correction factors for organs that were not completely covered by the protocol field-of-view. E from the dual energy abdomen/pelvis CT was calculated to be 17.8 mSv ± 11.6%. This calculation was then compared to E from dose length product method, which yielded 14.62 mSv. INTRODUCTION Dual energy computed tomography (DECT) is a method of computed tomographic (CT) imaging in which two CT scans are performed on the patient using differing X-ray tube potentials which allow for energy-selective reconstruction. Reconstruction in CT relies on measuring the line integral of the linear attenuation coefficient along each projection. The linear attenuation coefficient depends primarily on atomic number and electron density as well as the energy of the X-ray photons. This allows the attenuation coefficient of a material to be modeled as a function of energy using a set of basis functions. These basis functions provide information on how attenuation for a material will change between two energy spectra. In DECT, a second measurement of attenuation at a different energy provides an additional dataset. This additional dataset provides a measurement of the change in attenuation between the two energy projections. This information can then be compared to known changes of attenuation provided by the basis functions for various materials in order to differentiate material composition(1, 2). Acquiring datasets at two separate energies allows for a number of advantages over conventional CT. The two datasets provide attenuation information corresponding to each energy that can then be used for enhanced material discrimination, improved image registration, and a reduction of beam hardening effects(3, 4). There are currently two primary methods used to create a dual energy environment available as commercial products. The first of these methods is referred to as Dual Source Computed Tomography (DSCT), in which two separate X-ray tubes are operated at different energies simultaneously. The two sources are operated at differing tube currents and each have a different level of filtration. The second method of DECT makes use of a single X-ray tube that rapidly switches from low kV to high kV between consecutive projections, alternating at a rate of up to 4.8 kHz. Due to the fast rate of kV switching the single X-ray tube is operated at a constant current, with longer dwell time for the low kV projections to compensate for higher attenuation. For this study we used a fast-kV switching system to create our dual energy environment, with tube potential alternating between 80 and 140 kV. A visualization of this system can be seen in Figure 1. The key advantage of fast-kV switch DECT compared to DSCT is a reduction in motion and misregistration artefacts due to the rapid switching between energy projections; however, it offers limited control of tube current for the different projections. Figure 1. View largeDownload slide Illustration of fast-kV switching DECT. The alternating colors represent consecutive projections alternating between high and low kV. Figure 1. View largeDownload slide Illustration of fast-kV switching DECT. The alternating colors represent consecutive projections alternating between high and low kV. Effective dose (E) is a quantity used to represent the stochastic health risk to the whole body from non-uniform exposure to ionizing radiation(5, 6). The quantity is derived by the summation of organ doses multiplied by a corresponding tissue-specific weighting factor. These tissue weighting factors are described by the International Commission on Radiological Protection (ICRP) to represent the fraction of health risk from a specific tissue or organ. Physical measurements of individual organ dose due to an imaging procedure are generally estimated using small internal dosemeters such as thermoluminescence dosemeters (TLD) or metal oxide semiconductor field-effect transistor (MOSFET) detectors. These dosemeters have an energy dependence which gives rise to difficulty when calibrating the detectors in a dual energy environment. Specifically, between 80 and 140 kV there is a 20% difference in MOSFET response. Currently the few publications that attempt to estimate E from physical organ dose measurements in a dual energy environment pertain specifically to dual source computed tomography(7–9). There is currently a lack of publications pertaining to E from dual energy CT acquired via fast kV switching. The objective of this study was to test a novel method of estimating organ dose by using the effective energy (EE) of the combined dual energy environment to calibrate our internal dosemeters for dose measurements. This process can be viewed as the combination of three primary goals: (1) Validating the novel approach for dose equivalence between the MOSFET method and an ion chamber (IC) which serves as the gold standard; (2) measuring organ dose using MOSFET detectors in an anthropomorphic phantom, and estimating E in clinical dual energy abdomen/pelvis protocol by applying ICRP 103 tissue weighting factors(10); and (3) finally evaluating the applicability of the DLP method for E estimation against clinical data. MATERIALS AND METHODS This study was conducted using a GE Discovery CT 750HD (GE Heathcare, Inc., Waukesha, WI) implementing Gemstone Spectral Imaging (GSI) on an abdomen/pelvis protocol. This GSI protocol is a fast-kV switching dual energy scan, rapidly alternating between 80 and 140 kV. Our experimental method includes six primary components: (1) Characterize the EE of the dual energy environment; (2) estimating corresponding f-factor for soft tissue; (3) calibrating our MOSFET detectors using a conventional CT beam with EE as close to that of the dual energy environment; (4) validating our calibration method by comparing dose from MOSFET and IC using a CTDI body phantom in the dual energy environment; (5) estimating E for an abdomen/pelvis DECT scan using an anthropomorphic phantom and applying ICRP 103 tissue weighting factors; and (6) estimating E using AAPM dose length product (DLP) method for comparison. MOSFET detectors provide dose exposure information by measuring change in the voltage necessary to initiate current flow in the detector circuit, known as the device threshold voltage. Ionizing radiation produces electron–hole pairs in the silicon substrate. These pairs move to the oxide-silicon interface which results in a negative threshold voltage shift. The MOSFET detectors have been shown to have a linear energy dependence(11), and have previously been validated against thermoluminescent detectors in 120 kVp CT beams.(12) Characterizing EE of combined dual energy environment The EE of an X-ray beam is defined as the monoenergetic photon energy that would produce the same first half-value layer (HVL) as the beam itself in a given material. By directly measuring the HVL (in mm AL) of the combined dual energy beams we can calculate a linear attenuation coefficient using the simple relation:   μ=ln2HVL (1) Once this value is known we can use attenuation data such as NIST X-ray attenuation databases(13) to find our EE. The GSI protocol was performed in diagnostic mode, which allows the X-ray tube to remain stationary above the patient couch. The combined dual energy beam was then measured using a 6 ml IC (10 × 5–6, Radcal, Monrovia, CA). This measurement was then repeated four times adding different thickness of thin aluminum sheets between the X-ray tube and IC. The IC measurements were plotted vs. thickness of added aluminum filtration in order to perform exponential regression analysis, which was used to calculate a HVL of 5.02 mm Al. Estimating f-factor for organs It is necessary to specify in our calibration that we seek to find dose in soft tissue. This is accomplished by using an f-factor, which translates the exposure we measure in air to dose in a medium of interest, in this case, soft tissue(14). The equation for f-factor can be seen below:   f-factor=0.873(μenρ)ST(μenρ)AIR[cGyR] (2)where (μenρ)ST(μenρ)AIR represents the ratio of mass energy absorption coefficients for soft tissue and air. The f-factor value was later used to formulate calibration factors for each individual MOSFET detector. A separate f-factor of was calculated to determine dose to bone surface by using the same equation with (μenρ)Bone substituted for (μenρ)ST. It is also worth noting that the soft tissue f-factor changes very little within the EE range of the separate low and high kV beam, which is roughly 40–55 kV. This relative stability in f-factor can be seen in Figure 2. Calibration of MOSFET detectors MOSFET detectors were calibrated by matching the calibration beam quality (HVL 5.42 mm Al, EE 44.7 keV) as close as possible to the GSI beam quality (HVL 5.02 mm Al, EE 42.8 keV). HVL was measured by using a Piranha multifunction X-ray meter (RTI Electronics, Mölndal, Sweden). The MOSFET detectors were placed adjacent to the active area of a 6 ml IC (10 × 5–6, Radcal, Monrovia, CA), and then irradiated using the 80 kV calibration beam in order to calculate calibration factors for the 18 individual MOSFET detectors. These calibration factors convert the raw millivolt readings of the MOSFET detectors into absorbed dose in soft tissue (rad). The equation for these calibration factors can be seen in the following equation.   CF=MOSFETmeasurement((ionchamberreading)×(ionchambercorrectionfactor)×(f-factor))[mVcGy] (3) The calibration method was tested by inserting two MOSFET detectors along with a 0.18 ml IC (10 × 5–0.18, Radcal, Monrovia, CA) in the center cavity of a CTDI body phantom (West Physics, Atlanta, GA). The 0.18 ml IC was chosen based on two criteria. The first was its fairly small energy dependence over the energy range produced in our dual energy protocol. The second benefit of the 0.18 ml IC is that its small size allows for the detector to be easily placed into our CTDI phantom. Our ICs (Radcal model 10 × 5–6 ml, 0.18 ml) were calibrated at University of Wisconsin Radiation Dosimetry Laboratory (AAPM accredited dosimetry laboratory) that is NIST traceable. In Figure 3 we can see the scout image from the GSI abdomen/pelvis protocol. The CTDI body phantom was scanned using the dual energy abdomen/pelvis protocol. Detector readings were averaged over three scans and standard deviation was recorded. These results experimentally confirm dose equivalence between our calibrated MOSFET detectors and IC. Figure 3. View largeDownload slide Scout image showing the scanning region of the abdomen/pelvis protocol. Figure 3. View largeDownload slide Scout image showing the scanning region of the abdomen/pelvis protocol. Organ dose measurement After verifying our calibration, the MOSFETs were inserted into an adult anthropomorphic phantom (Model 702-BR190, CIRS) in order to measure organ doses. This phantom is composed of 39 slabs of thickness 2.5 cm, and simulates the radiological behavior of different tissue types including soft tissue, lung, bone, brain, and spinal cord and disks. These slabs have many numbered predrilled holes that correspond to different tissue and organ locations listed in the phantom user manual. The phantom is designed to represent a patient of 73 kg weight and 173 cm height. The phantom was loaded with 18 MOSFET detectors were inserted into the phantom in locations representative or tissues of interest that would be exposed during the procedure. These locations were esophagus, lungs, heart, liver, stomach, pancreas, gall bladder, spleen, small intestine, large intestine, kidney, ovary, uterus, bladder, prostate, testis, bone marrow and bone surface. In addition to these locations, an additional detector was used to record skin entrance dose on the abdomen surface. The parameters of the dual energy protocol were fast kV switch between 80 and 140 kV, 640 mA. The loaded phantom was scanned four times using the dual energy protocol. Uncertainty was evaluated using a combined expanded uncertainty with a coverage factor of 2 (k = 2). Type B uncertainties were included in to account for observed variation in our calibration, as well as EE discrepancy. In order to accurately reflect dose to specific tissues several partial volume correction factors (PVCF) were necessary to adjust for organs/tissues that are only partially irradiated in an abdomen/pelvis protocol(15, 16). The organs in our study that required this correction were bone marrow, bone surface, skin, lung and esophagus. The dose to bone marrow was corrected using published estimations of the distribution of active bone marrow in adults(17). This allowed us to scale our measured dose in locations corresponding to sites of bone marrow storage. Dose to bone surface was corrected using percentages of skeletal mass for the bone structures measured in our study(18). The dose to skin surface was calculated using ‘the rule of nines’(19). This is traditionally used in medicine to estimate the percentage of a persons skin affected by burn trauma. This rule separates the skin surface into nine anatomical regions and applies a surface area percentage fraction to each. The surface area factors used for this experiment were chest—0.09, abdomen—0.09 and back of torso—0.18. Using this method we calculated 36% of the total skin surface area was exposed during the scan. Volumetric corrections for lung and esophagus were performed using a method previously derived by our lab dealing specifically with our adult anthropomorphic phantom(15). This method describes the percentage of total lung and esophagus volume present in each slice of the phantom. Using this method, we applied correction factors for lung and esophagus dose based on which sections of the phantom were exposed during the abdomen/pelvis protocol. The remainder of the organs in our study were assumed to have unity PVCF because their entire volume was covered completely by the scan region. Effective dose was then calculated using ICRP 103 tissue weighting factors, WT, along with radiation weighting factors, WR. The formula we used can be seen in the following equation:   E=∑TPVCFT×WT×WR×DT (4)where DT represents the dose measurement from MOSFET detectors. In addition to calculating E from our organ dose measurements, we also compare our result to E calculated using AAPM DLP method(20). This method multiplies the DLP output by the CT system and multiplies it by a coefficient dependent on the age of the patient as well as the region of the body scanned during the CT examination. The DLP of the dual energy abdomen/pelvis protocol was 974.62 mGy cm, and the coefficient was 0.015.   EDLP≈k(mSvmGycm)×DLP(mGycm) RESULTS The EE of the combined dual energy beams was found to be 42.8 kV, corresponding to an f-factor of 0.923 for soft tissue, and 4.937 for bone. The EE of the 80 kV beam used in calibration was found to be 44.7 kV. A good agreement between MOSFET and IC was obtained: 1.71 rad ± 3.8% vs. 1.71 rad ± 0.5%. Organ dose measurements recorded by MOSFETs can be seen below in Figure 4. Using these measurements combined with ICRP 103 tissue weighting factors, the effective dose was calculated as 17.8 mSv ± 11.6%. Effective dose estimated using DLP method was 14.6 mSv. The uncertainty budget for the Effective Dose calculation can be seen in Table 1. Figure 4. View largeDownload slide MOSFET organ dose measurements. Figure 4. View largeDownload slide MOSFET organ dose measurements. Table 1. Uncertainty budget.   Type A (%)  Type B (%)  IC NIST calibration  —  1.9  IC monitor controller  —  0.5  HVL measurement  —  3.9  MOSFET calibration  3.8  —  f-factor  —  1.1  Effective dose calculation  0.3  —  Combined standard uncertainty  5.8  Expanded uncertainty (k = 2)  11.6    Type A (%)  Type B (%)  IC NIST calibration  —  1.9  IC monitor controller  —  0.5  HVL measurement  —  3.9  MOSFET calibration  3.8  —  f-factor  —  1.1  Effective dose calculation  0.3  —  Combined standard uncertainty  5.8  Expanded uncertainty (k = 2)  11.6  Table 1. Uncertainty budget.   Type A (%)  Type B (%)  IC NIST calibration  —  1.9  IC monitor controller  —  0.5  HVL measurement  —  3.9  MOSFET calibration  3.8  —  f-factor  —  1.1  Effective dose calculation  0.3  —  Combined standard uncertainty  5.8  Expanded uncertainty (k = 2)  11.6    Type A (%)  Type B (%)  IC NIST calibration  —  1.9  IC monitor controller  —  0.5  HVL measurement  —  3.9  MOSFET calibration  3.8  —  f-factor  —  1.1  Effective dose calculation  0.3  —  Combined standard uncertainty  5.8  Expanded uncertainty (k = 2)  11.6  DISCUSSION Scattered radiation in the phantom will lead to more attenuation at lower energies of the spectrum. This will have the result of slightly increasing the EE of the spectrum reaching the MOSFET detectors. After calibration, the purpose of the EE is to calculate the f-factors used to translate exposure into dose. In Figure 2 we see that this value is very stable across the EE range we are working with. We feel that the effect of scatter in the phantom will be negligible to measured organ dose. Figure 2. View largeDownload slide Soft tissue f-factor vs. energy in the range of our projected effective energy. Values calculated using Equation 2 and NIST X-ray attenuation databases. Figure 2. View largeDownload slide Soft tissue f-factor vs. energy in the range of our projected effective energy. Values calculated using Equation 2 and NIST X-ray attenuation databases. Furthermore, in our experiment we note a discrepancy in EE of the dual energy environment and the calibration beam. In order to account for this in our results we investigate the difference in dose calculation at the two EE values by comparing the uncertainty in f-factor calculation. This value is included in our uncertainty budget. One limitation of our study is the use of only one dosemeter in each organ. Ideally we would like to sample several points throughout the organs of interest to see a more detailed distribution of dose in each. The choice to use one dosemeter per organ was made considering the technical limitation of the MOSFET system supporting a maximum of 20 detectors. For the purpose of this study we felt the information gained from sampling a greater number organs was preferable to measuring several points in only a few. When looking at the air gaps between slices in our phantom, it is important to note that the active area of the detector is placed inside predrilled holes in the slabs, and taped to secure their positioning. This eliminates the possibility of the MOSFET being directly exposed to the primary beam. We feel the small gaps caused by the MOSFET wires will have a negligible contribution to measured dose. It is also worth noting that the phantoms used in our study are not perfectly representative of the reference male used in E formalism in ICRP publication 110. Where the reference male weighs 70 kg with height 170 cm, our phantom is representative of a person with weight 73 kg and height 173 cm. The experimental validation of our EE calibration of MOSFETs shows that it is a viable option for taking organ dose measurements in DECT. This novel approach to internal dosemeter calibration successfully circumnavigates the inherent challenge of taking physical measurements in a dual energy environment. By calibrating in a conventional CT beam with EE equal to that of the DECT it is possible to minimize the effect of energy dependence in the dosemeter. Currently the DLP method is most often used to estimate E from CT examinations. Our method of E estimation yields a value of 17.8 ± 2.1 mSv, which corresponds to a 21.9% difference compared to the DLP method. Furthermore, E estimated using organ dose measurements depends heavily on what organs are included in the calculation. CONCLUSION This study shows that using the EE of the combined dual energy beams is a viable way to calibrate MOSFET detectors in a dual energy environment. The method is useful particularly in ‘fast-kV switching systems’ where the high and low kV beams cannot be separated for individual treatment. The E from the DLP method differed 22% from the MOSFET method; however, the DLP method still provides a viable alternative to the more complex and labor intensive MOSFET method, particularly in busy clinical setting. Moving forward we would like to apply this method of E estimation to other dual energy systems such as dual source CT. FUNDING This study was funded, in part, by the US Nuclear Regulatory Commission (Grant no. NRC-HQ-12-G-38-0022). REFERENCES 1 Alvarez, R. E. and Macovski, A. Energy-selective reconstructions in x-ray computerised tomography. Phys. Med. Biol.  21( 5), 733 ( 1976). Google Scholar CrossRef Search ADS PubMed  2 Macovski, A., Alvarez, R. E., Chan, J. H., Stonestrom, J. P. and Zatz, L. M. Energy dependent reconstruction in X-ray computerized tomography. Comput. Biol. Med.  6( 4), 325– 336 ( 1976). Google Scholar CrossRef Search ADS PubMed  3 Li, B. Dual-energy CT with fast-kVp switching and its applications in orthopedics. OMICS J. Radiol.  2, 1– 8 ( 2013). Google Scholar CrossRef Search ADS   4 Yu, L., Leng, S. and McCollough, C. H. Dual-energy CT–based monochromatic imaging. Am. J. Roentgenol.  199( 5_supplement), S9– S15 ( 2012). Google Scholar CrossRef Search ADS   5 Jacobi, W. The concept of the effective dose a proposal for the combination of organ doses. Radiat. Environ. Biophys.  12( 2), 101– 109 ( 1975). Google Scholar CrossRef Search ADS PubMed  6 ICRP 1977. Recommendations of the ICRP. Ann. ICRP , vol. 1, ( New York: Pergammon). 7 Paul, J., Banckwitz, R., Krauss, B., Vogl, T. J., Maentele, W. and Bauer, R. W. Estimation and comparison of effective dose (E) in standard chest CT by organ dose measurements and dose-length-product methods and assessment of the influence of CT tube potential (energy dependency) on effective dose in a dual-source CT. Eur. J. Radiol.  81( 4), e507– e512 ( 2012). Google Scholar CrossRef Search ADS PubMed  8 Schenzle, J. C., Sommer, W. H., Neumaier, K., Michalski, G., Lechel, U., Nikolaou, K., Becker, C. R., Reiser, M. F. and Johnson, T. R. Dual energy CT of the chest: how about the dose? Invest. Radiol.  45( 6), 347– 353 ( 2010). Google Scholar PubMed  9 Henzler, T., Fink, C., Schoenberg, S. O. and Schoepf, U. J. Dual-energy CT: radiation dose aspects. Am. J. Roentgenol.  199( 5_supplement), S16– S25 ( 2012). Google Scholar CrossRef Search ADS   10 ICRP 2007. The 2007 Recommendations of the Internal Commission on Radiological Protection. ICRP Publication 103, Ann. ICRP. vol. 37, ( New York: Pergammon). 11 Ding, L. ( 2013). The use of a novel radiation detector on quantifying PET/nuclear medicine occupational and non-occupational doses and calibration of MOSFET radiation detectors against effective energy. Doctoral dissertation, Duke University. 12 Yoshizumi, T. T., Goodman, P. C., Frush, D. P., Nguyen, G., Toncheva, G., Sarder, M. and Barnes, L. Validation of metal oxide semiconductor field effect transistor technology for organ dose assessment during CT: comparison with thermoluminescent dosimetry. Am. J. Roentgenol.  188( 5), 1332– 1336 ( 2007). Google Scholar CrossRef Search ADS   13 Hubbell, J. H. and Seltzer, S. M., 1995. Tables of X-ray mass attenuation coefficients and mass energy-absorption coefficients 1 keV to 20 MeV for elements Z = 1 to 92 and 48 additional substances of dosimetric interest (No. PB—95-220539/XAB; NISTIR—5632). National Inst. of Standards and Technology-PL, Gaithersburg, MD. Ionizing Radiation Div. 14 Hendee, W. R. Medical Radiation Physics , 2nd edn. Chicago: Year Book Medical Publishers) 171 ( 1979). 15 Januzis, N., Nguyen, G., Hoang, J. K., Lowry, C. and Yoshizumi, T. T. A novel method of estimating effective dose from the point dose method: a case study—parathyroid CT scans. Phys. Med. Biol.  60( 5), 1763 ( 2015). Google Scholar CrossRef Search ADS PubMed  16 Januzis, N., Nguyen, G., Frush, D. P., Hoang, J. K., Lowry, C. and Yoshizumi, T. T. Feasibility of using the computed tomography dose indices to estimate radiation dose to partially and fully irradiated brains in pediatric neuroradiology examinations. Phys. Med. Biol.  60( 14), 5699 ( 2015). Google Scholar CrossRef Search ADS PubMed  17 Cristy, M. Active bone marrow distribution as a function of age in humans. Phys. Med. Biol.  26( 3), 389 ( 1981). Google Scholar CrossRef Search ADS PubMed  18 Valentin, J. Basic anatomical and physiological data for use in radiological protection: reference values: ICRP Publication 89. Ann. ICRP  32( 3), 1– 277 ( 2002). Google Scholar CrossRef Search ADS   19 Hettiaratchy, S. and Papini, R. Initial management of a major burn: II—assessment and resuscitation. Br. Med. J.  329( 7457), 101– 103 ( 2004). Google Scholar CrossRef Search ADS   20 McCollough, C. et al.  . The measurement, reporting, and management of radiation dose in CT. Rep. AAPM Task Group  23, 1– 28 ( 2008). © The Author(s) 2018. Published by Oxford University Press. All rights reserved. For Permissions, please email: journals.permissions@oup.com This article is published and distributed under the terms of the Oxford University Press, Standard Journals Publication Model (https://academic.oup.com/journals/pages/about_us/legal/notices)

Journal

Radiation Protection DosimetryOxford University Press

Published: May 4, 2018

There are no references for this article.

You’re reading a free preview. Subscribe to read the entire article.


DeepDyve is your
personal research library

It’s your single place to instantly
discover and read the research
that matters to you.

Enjoy affordable access to
over 18 million articles from more than
15,000 peer-reviewed journals.

All for just $49/month

Explore the DeepDyve Library

Search

Query the DeepDyve database, plus search all of PubMed and Google Scholar seamlessly

Organize

Save any article or search result from DeepDyve, PubMed, and Google Scholar... all in one place.

Access

Get unlimited, online access to over 18 million full-text articles from more than 15,000 scientific journals.

Your journals are on DeepDyve

Read from thousands of the leading scholarly journals from SpringerNature, Elsevier, Wiley-Blackwell, Oxford University Press and more.

All the latest content is available, no embargo periods.

See the journals in your area

DeepDyve

Freelancer

DeepDyve

Pro

Price

FREE

$49/month
$360/year

Save searches from
Google Scholar,
PubMed

Create lists to
organize your research

Export lists, citations

Read DeepDyve articles

Abstract access only

Unlimited access to over
18 million full-text articles

Print

20 pages / month

PDF Discount

20% off