Computed Tomography Dosimetry and Dose Risks - Research Paper Example

Purpose Dual Energy CT (DECT) examinations yield medical benefits that must be balanced against potential risk from patient radiation exposure. Consequently, clinical tools for measuring internal organ dose are needed for medical risk assessment. To measure both organ dose and effective dose (ED) for adult cardiac CT examination by using a potential clinical protocol for Dual Energy CT imaging (DECT), as well as to estimate the lifetime attributable risk (LAR) of organs cancer incidence on the basis of the ED and organ doses. Materials and Methods Cardiac DECT scan was performed by using potential clinical protocol (GSI 15 imaging protocol, 64 sections at 0.625-mm collimation, alternating energy between 80 kVp and 140 kVp, 640 mA, 0.6 second tube rotation). Absorbed organ doses were measured by using an adult male breast (female breast attached) anthropomorphic phantom and metal oxide semiconductor field effect transistor detectors (MOSFET), and were obtained also by the computational method based on the Monte Carlo simulation (Im-PACT calculator). Results The MOSFET dosimeters were evaluated for reproducibility, linearity, energy, and angular dependence. Our results indicated that the MOSFET dosimeter has excellent linearity within diagnostic dose ranges, but in low dose regions the values are less reliable. The energy dependence was about 7% at tube potentials from 80 kVp to 140 kVp. The results from performing GSI 15 protocol Im-PACT calculator showed that the breast, lung, stomach, and esophagus had the highest recorded absorbed organ doses. For the same organs, the MOSFET dose measurements were consistently lower than the calculated doses by Im-PACT. The estimates of radiation risk in this study are relatively small for any individual patient. Chapter 1 1 Introduction The X-ray Computed Tomography technology has become the method of choice for most diagnostic imaging procedures due to the remarkable advances over the past few decades, contributing to the improvement of diagnostic image quality and the reduction of examination time and cost (AAPM 23, 2008, the International Commission on Radiological Protection Publication (ICRP) 103, 2007, Huda et al, 2008). This has led to a rapid increase in using the CT scanners around the world. In the United States alone, 62 million CT scans were performed in 2006, of which 4 million were for children (Brenner & Hall, 2007). In Japan there are 91 scanners per million people (Coach, 2008). The recent developments in CT technology have made it possible to use Dual Energy Computed (DECT) in routine clinical radiology. The technique utilizes two photon energy spectra to construct CT images and it was first described in the late 1970s (Millner et al, 1979). DECT imaging technology can provide image contrast optimisation, material decomposition, and monochromatic spectral images (Langan, 2008). Because X-ray absorption is energy dependent, changing the kilovoltage of the X-ray tube results in material specific changes in attenuation. This means that CT is able to differentiate various materials (tissues) in one scan. Until now, using CT without the Dual Energy feature would require that two scans be needed separate between two materials. The breakthrough offers potential to enhance cardiovascular and vascular, body perfusion, brain, lung and liver imaging, which enables physicians to diagnose their patients’ conditions faster and more accurately (Langan, 2008). Unfortunately, the benefits of CT examinations in general come at a cost. The radiation dose of CT examinations is much higher compared with other X-ray diagnostic procedures (AAPM 23, 2008). One estimate showed it contributes almost 70% of the total dose (Huda, 2008). The typical dose for a single CT scan is between 15 mSv in the case of adults, and to 30 mSv in the case of neonates (Brenner & Hall, 2007). Despite these risks, the use of CT is growing annually, offering significant improvements in the variety and quality of CT clinical applications, which raise the need to reevaluate these exam protocols and to compare the radiation risk versus medical benefit (ICRP 103, 2007, AAPM 23, 2008). The risk of cancer might be increased slightly with the dose from a CT scan, but also it is possible that a poorer image could lead to miss an important detail necessary for correct diagnosis (Brenner & Hall, 2007). Therefore, the two basic principles of radiation protection of the patient that are recommended by the ICRP (justification of practice and optimisation of protection) and the ALARA principle (As Low As Reasonably Achievable) call for the most detailed, accurate radiation dose information available. The Effective Dose (ED) is designed specifically to help protect from radiation. In fact, it is the only dose quantity that attempts to link with the risk and therefore serves an important role in identifying the exposure during medical imaging (Martin, 2007). The ED is used in situations where the dose distribution is not homogenous as after CT examinations, and is “designed to be proportional to a generic estimate of the overall harm to the patient caused by the radiation exposure” (Brenner & Hall, 2007). The most common CT dosimetry approaches for estimating the ED are Computed tomography dose index (CTDI) values and Dose length product (DLP), Monte Carlo simulation, and the use of dosimeters for point dose measurements. The first approach is the most readily available technique and uses values displayed on the CT console at the time of examination. This method employs the CT dose index volume, or CTDIvol, and the DLP as part of an equation that results in an estimate of patient exposure, the ED to a patient for any given examination. However, this approach provides an easily estimated value of effective dose, but its uncertainty can be as much as ± 40% and the associated risk could be higher or lower by a factor of 3 depending on the individual characteristics (Martin, 2007). The second approach is a computational method employs Monte Carlo (MC) techniques and CTDI values to simulate the interaction of X-ray photons within a mathematical phantom and calculate the ED of organs. The third approach is a more direct technique that involves the use of anthropomorphic phantoms with internally placed dosimeters to generate more accurate measurements of organ doses and EDs. The risk is also present for low levels of ionising radiation (Kinahan et al, 2003). However, the risk from diagnostic imaging involving exposure to radiation is not so clear. The lifetime attributable risk (LAR) is an approximation of the risk of exposure-induced death and is used to estimate lifetime risks. It “describes excess deaths (or disease cases) over a follow-up period with population background rates determined by the experience of unexpected individuals” (ICRP 103, 2007). To the above list can also be added the size of the patient. The actual risk during a CT scan is linked to carcinogenesis, i.e. “the production and development of cancer” (Chambers, 1992) as well as genetic effects. 1.1 Motivations Discovery CT750 HD scanner is a new arrival to clinical centers. The scanner is state of the art in DECT imaging techniques manufactured by GE Healthcare (Mil- waukke, IL, USA). This machine can perform clinical DECT protocols using rapid kV switching between two energy levels (80-140 kVp) to acquire the dual energy within 0.5 msec (GE Healthcare, 2008). Although CT is clearly providing many clinical benefits, the doses from this technology should be reported and justified. To our knowledge, presently in the UK there are three machines, but there are no studies that have yet been reported about the radiation doses and the associated risks. Our motivation in this study is to report the absorbed doses that eventuate during a cardiac DECT scan to be able to balance the benefits with possible risks as recommended by ICRP and AAPM, and legislated by the European law (AAPM 23, 2008, ICRP 103, 2007). 1.2 Aims The primary aim of this study was to measure both organ dose and ED for adult cardiac CT examinations by using clinical protocol for DECT, and to estimate the lifetime attributable risk (LAR) of organs cancer incidence on the basis of the ED and organ doses. Gemstone Spectral imaging protocol No. 15 (GSI 15) was the investigated clinical protocol. The two energies of DECT scanned a length of 160 mm across the heart with 40 mm slice thickness. The imaging parameters of this protocol is shown in table *. This clinical protocol was recommended by radiologists at the hospital since it has potential to be used in calcium subtraction, plaque characterisation, and in perfusion imaging of the myocardium. Therefore, an objective assessment was needed to justify the use. There are two methods that were utilised to estimate the doses. First was Im-PACT calculator, which is a computational method based on the Monte Carlo simulation. The second method used was an anthropomorphic phantom (ATOM phantom, CIRS) with MOSFET detectors (Best Medical). A comparison between these two methods is also discussed in this study. 1.3 Organisation of the Dissertation This dissertation consists of seven chapters. Chapter 1 introduces the new technology DECT and the issues that are related to justifying its use clinically and the issues related to dose measurements. Chapters 2, 3, and 4 are theoretical backgrounds for the study. Chapter 2 discusses the latest technologies in CT, the basic principle of CT, CT dose distribution, Dual energy X-ray CT, and Discovery CT750 HD scanner. Chapter 3 deals with the theoretical review of CT dosimetry, especially the concepts that were specifically developed to describe the radiation dose from CT, CTDI measures, and DLP, as well as the methods of estimating the organ specific doses that were used in this study (Im-PACT calculator and anthropomorphic phantoms with MOSFET dosimeters). Chapter 4 discusses the risk from diagnostic imaging involving exposure to radiation and the lifetime attributable risk (LAR). In Chapter 5, the materials and the methodology of conducting the study are explained in detail. Then, Chapter 6 presents the results. Finally, Chapter 7 is the discussion and the conclusion of the study as well as the analyses of errors of the calculations. Chapter 7 Discussion and Conclusion 7.1 MOSFET performance assessment The MOSFET characteristics that were tested were the reproducibility, angular dependence, linearity, energy dependence, and plug effect. Most of these parameters were tested with the CT machine and dual energy source. MOSFET dosimeter showed excellent linearity over the full clinical range. No corrections were made for changes in MOSFET sensitivity. The result of MOSFET angular dependence agreed with the manufacturer of ±2%. This was also confirmed by Roshau and Hintenlang (2003). The reproducibility uncertainty of the MOSFETs was within 15%, and it was decided that any phantom measurements should be performed at least three times to obtain accurate results. The energy dependence of the high sensitivity MOSFET dosimeter system was higher than 10%, which was the value of the energy dependence that was mentioned by the manufacturer (Operators Manual mobile MOSFET). 7.2 Organ Doses Figure 22 shows the calculated organ doses by Im-PACT from performing the GSI 15 protocol. Organs that were directly in the field of view (breast, lung, esophagus, heart) had the highest absorbed doses. According to the Im-PACT result, it was decided that the phantom measurements will be carried out for four organs (lung, breast, stomach and esophagus) due to time and equipment restrictions. These were chosen because, according to Im-PACT calculation, they contributed about 80% of the total effect that was inherited by the mathematical phantom. 7.3 ED measurements The study involved estimates of organ doses by Im-PACT calculator within the computational model, and by MOSFET measurements within the irradiated physical phantom. The imaging parameters of the DECT cardiac scan protocol (GSI 15) were applied on both methods with a scan length of 160 mm over the heart. The result in Table 14 showed comparisons of lung, breast, stomach, and esophagus effective doses that were estimated by both methods. The experimental values were lower than those calculated by Im-PACT calculator. The suggested discrepancy has been attributed mainly to design differences between the ATOM and the mathematical phantoms, positioning dosimeters within the phantom, MOSFET dosimeters uncertainty, and Monte Carlo derived normalized data. Perisinakis et al. (2008) investigated whether or not Monte Carlo derived normalized data can provide accurate estimations of patient dose from CT exposures using the MC methodology, mathematical anthropomorphic phantoms, and a multislice scanner. Different body sizes and both axial and spiral examinations were conducted. Discrepancies were noted in the measurements but “normalized effective dose values for the standard contiguous axial CT examinations derived by Monte Carlo simulation were found to considerably decrease with increasing body size of the mathematical phantom used”. Overall, however, it was found that the CTDI to ED conversion coefficients using MC simulation in axial scans “may provide a good approximation of corresponding coefficients applicable in helical scans”. The Im-PACT calculator employs the mathematical MIRD phantom, which approximates the organs by simplified geometric shapes, while the ATOM phantom is more anatomically realistic. Furthermore, the ATOM phantom is a different size compared with the mathematical phantom. Also, the estimated location of the organs differs significantly in comparison with the distribution of organs in the mathematical phantom. The lack of any defined organ boundaries within the ATOM phantom made accurate determination of organ volumes difficult, and there is scope for misplacement of MOSFETs if there is poor knowledge of anatomy. In this regard, our methodology utilised the organs map that was reported by Scalzetti et al for adult male RANDO Phantom due to the lack of anatomical information of the ATOM (2008). Although, the physical dimensions between the ATOM and RANDO are very similar, we believe that an error in the measurements may arise from this. A further investigation is needed in this area. The MOSFET dose estimations have limitations and statistical uncertainties, which can adversely affect the dose calculations. These factors were discussed previously in the chapter. Between experimental and calculated doses, it was determined that the breast has the highest percentage difference compared to the other organs. This reduced response by MOSFET may have resulted from the dosimeter orientation during dose acquisition, where the exposure was parallel to dosimeter wire-base (see the previous Figure 11). The angular dependence of this orientation (normal-to-axial) was studied by Roshau and Hintenlang, and the result was a standard deviation of 7.9% from the mean measurements for tissue equivalent phantom, where the axial orientation was only 2.7% (2003). Pomije et al reported similar finds and he attributed it to the partial shielding around the active area of the MOSFET in the normal-to-axial orientation (2001). Finally, when comparing the doses measured and calculated doses for fully irradiated organs, the differences were less than those obtained by other authors. For example, a similar study used Im-PACT calculator and RANDO anthropomorphic phantom with thermoluminescent dosimeters (TLDs) for estimating the dose during a whole body examination on a 16-detector unit. The reported relative difference between the calculated and the measured doses of lung, stomach, breast, and esophagus were 35%, 26%, 75% and -7% respectively. Other studies also reported similar or higher discrepancy between the TLD and Monte Carlo simulated dose measurements (Calzado, 1995, Shrimpton, 1991, Nishizawa, 1991, Geleijns, 1994, and Hashemi-Malayeri , 2003 ). 7.5 LAR of organs cancer incidence Although estimates of radiation risk in this study are relatively small for any individual patient, the use of this technology to perform a cardiac scan should be balanced against the risk of this test and the risks of other diagnostic tests. Also, estimates of radiation risk should be interpreted with caution for several factors. The BEIR VII report estimates of LAR for breast and lung cancer were derived primarily from Japanese populations, in whom the baseline risks for lung and breast cancer are different from the European countries’ population. The effectiveness of the diagnostic X-rays used in CT in inducing cancer may be different from that of the higher energy gamma rays that were the source of exposure for the atomic bomb survivors. The estimates of organ doses and LAR are subjected to uncertainties in the parameters. 7.6 Limitations Our study had limitations. First, a limited number of organs and point measurements within these organs were investigated, which might result in inaccurate dose estimation. The organ map that used to obtain the mean organ doses need validation. Only one DECT protocol was studied. The MOSFET dosimeters that were used in this study have limitations in low energy application, which might result in inaccurate dose estimation. Also, these MOSFETs have high energy dependence in low energy range, which might raise the uncertainty in the measurements. Read More