FOREWORD. Acknowledgements

Transcription

1 ΠΑΝΕΠΙΣΗΜΙΟ ΠΑΣΡΩΝ Διαημημαηικό Πρόγραμμα Μεηαπηστιακών ποσδών ζηην Ιαηρική Φσζική Διπλωμαηική εργαζία «ΔΟΙΜΕΣΡΙΑ ΑΘΕΝΩΝ Ε ΕΞΕΣΑΕΙ ΤΠΟΛΟΓΙΣΙΚΗ ΣΟΜΟΓΡΑΦΙΑ ΠΟΛΛΑΠΛΩΝ ΣΟΜΩΝ» ηέλλα Γ. Θαλαζζινού Α.Μ : 1575 ΕΠΙΒΛΕΠΩΝ: Γεώργιος. Παναγιωηάκης ΚΑΘΗΓΗΣΗ ΙΑΣΡΙΚΗ ΦΤΙΚΗ ΠΑΣΡΑ 2010

2 UNIVERSITY OF PATRAS Ιnterdepartmental Post-graduate Program of Studies in Medical Physics M.Sc. Thesis PATIENT RADIATION DOSIMETRY IN MSCT EXAMINATIONS Stella G.Thalassinou Registration Number: 1575 SUPERVISOR: G. S. Panayiotakis Professor of Medical Physics PATRAS 2010

3 FOREWORD The thesis research was carried out at the University Hospital Attikon in collaboration with the Medical Physics Laboratory of the department of Medicine of the University of Patras. Acknowledgements First of all, I would like to thank my supervisor, Professor of the University of Patras (Faculty of Medicine, Medical Physics Laboratory) G. Panayiotakis for giving me the opportunity to realise this project and for his continuous support and advice throughout this work. I would also like to thank the Assistant Professor of the University of Athens (Faculty of Medicine, Medical Physics Laboratory) E. Efstathopoulos, for all his guidance and help for the completion of this project. It is a pleasure for me to thank Dr I. Tsalafoutas, Medical Physicist in Agios Savvas Hospital for our fruitful cooperation in data processing. I am grateful to Mr I. Antonakos and Dr A. Stefanogiannis, Medical Physicists in University General Hospital of Athens Attikon for their collaboration in the second department of radiology. I would like to extend my appreciation to all the technologists-radiologists staff of the radiation laboratory for the data collection and especially the director of the technologists-radiologists Mr I. Saranteas for giving me his help and experience over subjects that are treated in this thesis. 1

4 ABSTRACT MultiDetector-row Computed Tomography (MDCT) or MultiSlice Computed Tomography (MSCT) has undergone remarkable progress since its first introduction at the end of the 1990s. Given that CT examinations are generally recognized as a relatively high-dose procedure, concern has been expressed at the associated increase in doses. The International Committee on Radiation Protection (ICRP) noted in their report No.87 that absorbed doses in tissues from CT are among the highest observed in diagnostic radiology (i.e mgy). Therefore, the purpose of this thesis is to calculate the dosimetric quantities for brain, chest, and abdomen-pelvis examinations that were carried out using Philips Brilliance 16 and Brilliance 64 CT Scanners of the University General Hospital Attikon, as well as to perform their intercomparison. For brain examinations, axial technique was utilized. However, for chest and abdomen-pelvis examinations, spiral technique was applied. The effect of overranging (or overscanning) is connected with spiral mode and its contribution to patient dose is really important in case of MSCT scanners. Therefore, the contribution of the overrange effect for chest and abdomen-pelvis examinations carried out was calculated. Additionally, dose measurements were carried out in order to estimate the radiation burden to the eye lenses and the thyroid during the typical brain examination, both when eye lenses are inside and outside the irradiation field. 2

5 FIGURES Figure 1: Generations of CT Scanners... 9 Figure 2: Comparison of scanning principle and image reconstruction in third generation and fourth generation CT scanners. The fourth generation uses a stationary detector ring, and the data acquired by one detector are assembled into a projection for the various tube positions Figure 3: Worldwide CT sales Figure 4: Schematic overview of the CT acquisition process Figure 5: Roentgen tube. Within the electrical field electrons are released from the cathode and travel through the vacuum housing of the roentgen tube towards the anode. When the high-energy electrons collide with the anode, photons are emitted. The anode and the location of electron impact rotate to avoid overheating. Low-energy photons, which do not contribute to the image formation,are filtered out. The remaining photons are then collimated to form = a fan- or cone-shaped beam Figure 6: Depending on the traversed material, emitted photons are partially absorbed or scattered. The remaining photons are collected and measured by the detectors on the opposite side. To calculate the attenuation throughout the plane, and reconstruct a CT image, a large number of attenuation profiles from consecutive rotational angles are required Figure 7: Reconstruction image steps Figure 8: Digitization of an (analog) image results in a numerical matrix Figure 9: Tissue densities and window level and width settings.. By setting the window level at a highdensity level, structures such as bone tissue can be evaluated. An intermediate level around 200 HU allows differentiation of the vessels and their relation to the vessel wall and calcifications. A low window level setting allows appreciation of the lungs. Tissues with density values beyond the boundaries of the window width appear as either saturated white (higher density) or black (lower density) Figure 10: Schematics of detector rows and elements Figure 11: Diagram of the detector geometry used in a 4-MDCT from two major manufacturers. The detectors array is 20-mm wide along the longitudinal axis and uses eight rows of varying widths to allow simultaneous scanning of 4 slice up to 5-mm thick Figure 12: Diagram of the detector geometries used in 64-MDCT from four major manufacturers. The Siemens 64-MDCT uses 32 sub-mm detectrors and a moving focal spot to achieve 64 overlapping slice measurements Figure 13: Illustration of the term Computed Tomography Dose Index (CTDI) : CTDI is the equivalent of the dose value inside the irradiated slice (beam) that would result if the absorbed radiation dose profile were entirely concentrated to a rectangular profile of width equal to the nominal width NxT Figure 14: Total dose profile of a scan series with n=9 subsequent rotations. The average level of the total dose profile, which is called Multiple Scan Average Dose (MSAD), is equal to the computed tomography dose index (CTDI) if the table feed (TF) is equal to the nominal beam width NxT (i.e., pitch p = 1) Figure 15: Cylindrical standard computed tomography (CT) dosimetry phantoms (16 cm and 32 cm in diameter) made from Perspex for representative measurements of the computed tomography dose index (CTDI) in regions of the head and the trunk, and a pencil-like detector for measurements Figure 16: The effect of overranging/overscanning increases with the detector width and the pitch used (A, B). The smaller the scan field, the larger the extra radiation exposure due to the overscanning effect Figure 17: Longitudinal dose modulation (LDM) is a refinement of AEC that adapts the mas settings slice-by-slice or rotation by rotation. Those parts of the scan range with reduced attenuation will be less exposed Figure 18: Angular dose modulation (ADM) is another refinement of AEC that adapts the tube current to the varying attenuation at different projection angles. Those projections with reduced attenuation will be less exposed Figure 19: Bandage-shaped hair loss in a 53 yeas old woman with subarachnoid haemorrhage. Temporary hair loss lasted for 51 days was seen on day 37 after the first perfusion study of the head with MDCT. In this patient four perfusion studies of the head with MDCT and two angiographies of the head had been performed within the first 15 days of admission to the hospital. (Reprodused with permission from author, Imanishi et al.2005) Figure 20: ETT Fimel annealing oven Figure 21: FIMEL Reader functional diagram

6 Figure 22: R100B solid state detector Figure 23: Sex distribution of the patients for the brain, chest ad abdomen examinations for the CT Scanners Brilliance 16 and Figure 24: Calibration device of TLD dosimeters using x-ray tube Figure 25: Calibration curve of thermoluminescence dosimeters Figure 26. Planning overview of a cranial tomogram. a, c Normal slice sequence; Figure 27: Mean values of scan length for Brain examinations Figure 28: Mean value of CTDIvol dose for Brain examinations Figure 29: Mean value of DLP dose descriptor for Brain examinations Figure 30: Mean value of scan length for Chest examinations Figure 31: Mean value of CTDIvol dose descriptor for Chest examinations Figure 32: Mean value of DLP dose descriptor for Chest examinations Figure 33: Mean value of scan length for Abdomen-Pelvis examinations Figure 34: Mean value of CTDIvol dose descriptor for Chest examinations Figure 35: Mean value of DLP dose descriptor for Abdomen-Pelvis examinations Figure 36: Mean value of conversion coefficients derived from tables utilizing ICRP 60 and 103 organweighting schemes for Brain examinations Figure 37: Mean value of Effective dose according to ICRP 60 and 103 organ-weighting schemes for Brain examinations Figure 38: Mean value of conversion coefficients derived from tables utilizing ICRP 60 and 103 organweighting schemes for Chest examinations Figure 39: Mean value of Effective dose according to ICRP 60 and 103 organs- weighting schemes for Brain examinations Figure 40: Mean value of conversion coefficients derived from tables utilizing ICRP 60 and 103 organweighting schemes for Chest examinations Figure 41: Mean value of Effective dose according to ICRP 60 and 103 organ-weighting schemes for Abdomen-Pelvis examination Figure 42: Radiation dose in the area of eye lens and thyroid from scatter irradiation Figure 43: Radiation dose in the area of eye lens and thyroid from primary irradiation

7 TABLES Table 1: Conversion Factors for calculating Effective dose Table 2: Characteristics of TLD-100H Table 3: Basic features of Brilliance 16 and Brilliance 64 CT scanners Table 4: Number of patients for Brain, Chest and Abdomen-Pelvis examinations Table 5: CT protocol for brain examination Table 6:CT parameters for chest and abdomen-pelvis examinations with Brilliance 16 and 64 CT scanners Table 7:Mean values and standard deviations (SD) of E and average CC are given for Head,Chest and Abdomen-Pelvis examinations Table 8: Mean values and standard deviations (SD) of E and average CC are given for Brain, Chest and Abdomen-Pelvis examinations

8 CONTENTS 1. INTRODUCTION History Purpose COMPUTED TOMOGRAPHY BASICS Data Acquisition Image Reconstruction Image Display Data Storage MDCT TECHNOLOGY Introduction to MDCT Technology Differences between SDCT and MDCT MDCT DOSIMETRY CT Dose Index (CTDI) Dose Length Product (DLP) Effective dose PARAMETERS AFFECTING DOSE IN MDCT Relevant scanning parameters affecting patient dose Anatomical parameter consideration in dose reduction Technical parameter consideration for dose reduction PERSPECTIVE ON RADIATION RISKS APPROACHES TO THE DOSE PROBLEM OF MDCT The ALARA Principle The Role of the Referrer: Justification The Role of the Operator: Optimization The Role of Guidelines in MDCT The Role of Evidence: Vigilance MEASURING DEVICES-MATERIALS CT scanners Brilliance 16 and 64, Philips Equipment for Thermoluminence Dosimetry TLD-100H FIMEL ETT Annealing oven LTM TLD Reader R100B Solid State Detector Software Tool DATA ACQUISITION AND PROCESSING Patients data Technique and scan data Εxtraction of data examinations Calculation of CT dose quantities Eye lenses and thyroid skin dose measurement Calibration of TLD dosimeters Initialization Batch Uniformity Linearity-Calibration curve TLD Setting and Patient Posture during Brain Examination RESULTS

9 10.1. Calculation of CTDIvol and DLP CT dose descriptors for Brain,Chest and Abdomen-Pelvis examinations Calculation of Effective dose and overscan contribution for Brain,Chest and Abdomen-Pelvis examinations Calculation of eye lenses and thyroid mean skin dose for Brain examinations DISCUSSION-CONCLUSIONS REFERENCES

10 1. INTRODUCTION 1.1. History Computed tomography (CT) has been one of the biggest breakthroughs in diagnostic radiology. The first clinical CT scanner was developed by Godfrey N. Hounsfield for examinations of the head and was installed in 1971 at Atkinson- Morley's Hospital in Wimbledon, England. The first body CT scanner was installed in 1974 and before the end of the 1970s, the basic technical evolution of CT was complete. Technical details were refined during the 1980s, and CT technology remained on a plateau until the early 1990s, when the advent of spiral (helical) CT scanning sparked a further, rapid evolution leading to improved diagnostic capabilities, 3D imaging techniques, and CT angiography. The latest innovation is the introduction of multi-slice CT in This new technology is vastly expanding the performance of CT scanners: it truly transforms CT from a transaxial imaging modality to a 3D technique that yields high quality images in arbitrary planes and forms the basis for an expanding variety of 3D visualization techniques, including virtual endoscopy. In addition, these scanners have the potential to revolutionize cardiac imaging with CT. Computed tomography is an x-ray tomographic technique in which an x-ray beam passes through a thin axial section of the patient from various directions. Parallel collimation is used to shape the x-ray beam to a thin fan, which defines the thickness of the scan plane. Detectors measure the intensity of the attenuated radiation as it emerges from the body. A mathematical image reconstruction (inverse Radon transformation) calculates the local attenuation at each point within the CT section. These local attenuation coefficients are translated into "CT numbers" and are finally converted into shades of gray that are displayed as an image. With conventional CT scanners the volume of interest is scanned in a sequential fashion, usually proceeding one section at a time. The first two generations of CT scanners (Figure 1) were superseded in the late 1970s by third and fourth generation scanners, which are still in use today. In third-generation scanners, tube and detector array rotate synchronously around the patient. The detector array covers the full width of the fan beam (Figure 2). In fourth 8

11 Figure 1: Generations of CT Scanners [3] generation scanners, the detector elements cover a fully cycle around the scanner opening and remain stationary during the scan, while only the x-ray tube rotates around the patient (Figure 2). Third generation scanners, however, offer better scatter suppression and require less detector elements, which is the reason why all multi-slice CT scanners use third generation technology. Figure 2: Comparison of scanning principle and image reconstruction in third generation and fourth generation CT scanners. The fourth generation uses a stationary detector ring, and the data acquired by one detector are assembled into a projection for the various tube positions [3] Attempts to speed up the imaging process led to the development of a multitube CT scanner called the dynamic spatial reconstructor, electron beam CT scanning, spiral CT and recently multi-slice CT have achieved large-scale clinical impact. [3] 9

12 Multi-Detector Computed Tomography (MDCT) has dramatically increased the performance capability of CT. Successive generations of systems capable of acquiring 4, 8, 16 and 64 sections simultaneously exist and greater configurations are now becoming available, with the latest cone beam systems capable of simultaneously acquiring 256 sections and 320 sections [4] However, the expansion of CT scanners leads in concern about the resulting increased radiation burden. The International Committee on Radiation Protection (ICRP) noted in their report No.87 [1] that Absorbed doses in tissues from CT are among the highest observed in diagnostic radiology (i.e mgy). Figure 3 shows the worldwide sales of CT scanners has more than doubled since 1998, and is predicted to continue increasing at the same pace. The trend of increasing use of CT scanning has been and is being documented by international organizations, national bodies and in individual studies. Information is available from the NEXT study in the USA [5], UNSCEAR [6], the EU [7], the NRPB/HPA in the UK [8-12], and from other sources [13-19]. The overwhelming thrust of the data is that the number of installations, the frequency and type of examinations, and the dose per examination are all increasing throughout the world, to the extent that the CT dose now accounts for 60 to 70% of the patient dose in some US tertiary referral centres. Figure 3: Worldwide CT sales [98] 10

13 The United Nations Scientific Committee on the Effects of Atomic Radiation (UNSCEAR) has highlighted that worldwide there about 93 million CT examinations performed annually at a rate of about 57 examinations per 1000 persons. UNSCEAR also estimated that CT constitutes about 5% of all X-ray examinations worldwide while accounting for about 34% of the resultant collective dose. [6] In the countries that were identified as having the highest levels of healthcare, the corresponding figures were 6% and 41% respectively In a frequently cited study performed by the Federal Bureau on Radiation Protection in Germany, it was found that between 1990 and 1992 only 4% of all X-ray examinations were performed on CT scanners, yet CT accounted for 35% of the collective effective dose [20]. In the United Kingdom, in 1991 the National Radiological Protection Board (NRPB) pointed out that CT makes a disproportionately large contribution to dose, at that time representing only 2.5% of examinations but constituting 25% of the collective dose to the population from diagnostic use [21]. Subsequent studies indicate that this proportion has increased; in 1998 Shrimpton and Edyvean [22] suggested that the cumulative radiation dose was closer to 40%. Mettler et al [23] have indicated that in their department CT comprises 11% of examinations and 67% of the collective dose, 11% of these examinations being carried out in children, in whom radiation protection considerations are paramount. The concern regarding radiation dose associated with CT is not really new in the literature. In 1981,the concept of CTDI to estimate and standardize measurements of radiation from CT and to take into account the scattered radiation along the z-axis, which varies with section thickness, was introduced. CTDI was then defined as the integral of the single-scan radiation dose profile along the z-axis normalized to the nominal beam width (NT). To standardize CTDI measurements, in 1984 the Food and Drug Administration (FDA) introduced the integration limits of _7NT. CTDI with these limits was abbreviated CTDI FDA and measured on standard 16- and 32-cm diameter phantoms, which were considered head and body phantoms. This method was suited for thick-section CT performed with single-detector scanners, but it yielded underestimated measurements of the radiation doses delivered at thin-section CT [24]. In 2002, the CTDI FDA measurement method was adapted for multi-detector CT to take into account helical scanning with large detector arrays, as well as thin collimations [24,25]. CTDI FDA was then replaced by CTDI 100, which integrated the 11

14 dose profile along the z-axis with fixed limits of 0.50 cm centered over the dose profile and corresponded to a length of 100 mm. Dose descriptors were also defined to take into account the dose distribution within phantoms. The dose at the periphery (ie, 1 cm below the surface) of the phantom typically is double the dose at the phantom center. A weighted CTDI was therefore introduced to represent an estimate of the average CTDI across the entire field of view. The weighted CTDI was calculated as the sum of one-third the CTDI 100 measured at the center plus two-thirds the CTDI 100 measured at the periphery of the phantom. The volume CTDI was created to include the calculation of table feed and was defined as the weighted CTDI divided by the pitch. To take into account the scanned body region, the DLP, defined as the product of volume CTDI times scanning length, was introduced. Each of these measurements was obtained in homogeneous phantoms and with constant tube current settings [25-29]. 12

15 1.2. Purpose According to the information above, the purpose of this thesis is the calculation of the dosimetric quantities for brain, chest and abdomen-pelvis examinations carried out using the Phillips Brilliance 16 and 64 CT Scanners of the University Hospital Attikon and the comparison between them. More specifically, the patient dose related quantities were calculated are: CTDI vol dose indicator which is the parameter that best estimates the average dose at a point with the scan volume for a particular scan protocol, since, the CTDIw is only indicator of the level of local dose in the irradiated slice. DLP dose parameter, which reflects the total energy absorbed (and thus the potential biological effect) from a specific scan acquisition. Effective dose, which is a dose parameter that reflects the risk of a non-uniform exposure in terms of a whole body exposure. Skin dose, in the area of eyes and thyroid. In this way the estimation of eye lenses and thyroid gland absorbed dose was approximately done. All dosimetric indices assessed in this study (DLP, CTDI vol and E) were calculated with a novel method which utilizes the DICOM headers of the soft copy CT images and conversion coefficient tables specific to the CT scanner and the tube potential used for acquisition. In order to check the software accuracy the calculated values CTDI vol and DLP were compared with the displayed values recorded at the CT scanner console. The effective dose was calculated according to the conversion coefficient (CC) tables from ICRP 60 and 103. For the calculation of patients effective dose value, the percentage of dose due to overrange effect was taken into account and separately calculated as well. Concerning patient skin dose, the direct measurement was performed using the radiosensitive thermoluminescence dosimeters TLD100H, placed on the area of interest (area of eyes and thyroid). 13

16 2. COMPUTED TOMOGRAPHY BASICS The process of CT can be divided into the following steps: data acquisition, image reconstruction, post-processing, evaluation and reporting, and data storage and exchange (Figure 4). Figure 4: Schematic overview of the CT acquisition process [32] 2.1. Data Acquisition Data acquisition refers to the collection of X-ray transmission measurements through the patient. It requires an X-ray source that produces an X-ray beam, which is collimated into the shape of a fan or cone (Figure 5). Figure 5: Roentgen tube. Within the electrical field electrons are released from the cathode and travel through the vacuum housing of the roentgen tube towards the anode. When the high-energy electrons collide with the anode, photons are emitted. The anode and the location of electron impact rotate to avoid overheating. Low-energy photons, which do not contribute to the image formation,are filtered out. The remaining photons are then collimated to form = a fan- or cone-shaped beam [32] 14

17 When an X-ray beam passes through an object, some of the photons are absorbed or scattered. The reduction of X-ray transmission, which is called attenuation, depends on the atomic composition and density of the traversed tissues, as well as on the energy of the photons. After passing through an object the partially attenuated X-rays are collected by X-ray detectors on the opposite side and converted from X-ray photons to electrical signals (Figure 6). These signals are then converted into digital data, after which the attenuation value is calculated. While the X-ray tube and detectors rotate around the patient, a large number of projections are collected from consecutive angular orientations. [32] Figure 6: Depending on the traversed material, emitted photons are partially absorbed or scattered. The remaining photons are collected and measured by the detectors on the opposite side. To calculate the attenuation throughout the plane, and reconstruct a CT image, a large number of attenuation profiles from consecutive rotational angles are required [32] 15

18 2.2. Image Reconstruction The reconstruction of images from the X-ray measurements involves the following steps (Figure 7). First, the measured X-rays are preprocessed, which is necessary to correct for beam hardening and scattered radiation. After pre-processing the raw data are filtered using convolution kernels. Figure 7: Reconstruction image steps [32] The filtering can result in very smooth to very sharp images based on the selected kernel. Depending on the reconstruction algorithm, projections from a 180 o or 360 o rotation are used for image reconstruction. Because spiral CT continuously acquires measurements at slightly varying longitudinal positions, an additional step is required before reconstruction. A complete set of projections at the selected plane position is created by interpolation of adjacent measurements in the longitudinal direction. The final step is to calculate the variation of regional attenuation within the image plane based on the collection of angular projections using back-projection reconstruction technique. The selected field of view is divided into small image elements, called pixels. The density value of each pixel depends on the composition of 16

19 the tissue it represents and is expressed in Hounsfield units (HU). The Hounsfield units are calculated from the attenuation measurements relative to the attenuation of water and range from 1024 to HU Hounsfield unit (x,y) = 1000 Υ μ(x,y) μ water / μ water (1) The result is a two-dimensional matrix of preselected size and detail, with each element representing the average attenuation of that location relative to water [32] 2.3. Image Display Contrary to continuous analog images (conventional X-ray), CT images are digital or numerical Images. The attenuated X-ray that reaches the detector is transformed into an analog electrical signal and then converted into a discrete digital format that can be processed by computer. The CT images on screen are an analog visual representation of binary values that have been digitally processed. The image consists of a matrix of discrete attenuation values that are the result of sampling and computer processing (Figure 8). Theoretically this entire range of attenuation values (-1024 to HU) could be displayed in a gradually sliding scale from black to white. Unfortunately, the human eye is incapable of distinguishing these fine nuances. Therefore, it is important to adjust the display setting in such a way that the range of density values of the structure of interest are displayed with optimal contrast (Figure 9). 17

20 Figure 8: Digitization of an (analog) image results in a numerical matrix [32] The window level indicates the density value at the center of the displayed gray scale, which determines the brightness of the image. The window width indicates which density values around the window level are within the gray scale display. Figure 9: Tissue densities and window level and width settings.. By setting the window level at a high-density level, structures such as bone tissue can be evaluated. An intermediate level around 200 HU allows differentiation of the vessels and their relation to the vessel wall and calcifications. A low window level setting allows appreciation of the lungs. Tissues with density values beyond the boundaries of the window width appear as either saturated white (higher density) or black (lower density) [32] 18

21 Therefore, the width determines the image contrast. All matrix elements with attenuation values beyond the window limits appear as either saturated white or black on the screen. The density value of water is predefined at 0 HU. The density value of soft tissues, such as non-enhanced muscle and blood, varies between approximately 100 and +200 HU, while fat tissue is at the lower end of that scale and bone or other calcified tissues have higher attenuation values. Using routine intravenous contrastenhancement protocols and contrast media, the attenuation value of the arterial blood is increased to a level between +200 and +400 HU. Metal has density values that overlap and exceed bone. Air and lung tissue have very low attenuation values. Taking the attenuation characteristics of the different tissue types into consideration, the window level of a contrast-enhanced CT image is set at around +250 HU and the window width at 400 HU. This allows appreciation of the contrastenhanced blood and its relation to the surrounding tissues, with bone and other high-density structures displayed as saturated white, and fat and air displayed as saturated black. Depending on the users interests and preferences, as well as the scanning conditions, these settings can be altered to allow optimal appreciation of the coronary luminal integrity. [32] 2.4. Data Storage Data can be stored on film but particularly if additional post-processing is anticipated in the future, digital storage is preferable. The size of a single CT image is approximately 500 kb. A complete cardiac study, including several reconstructions containing more than 250 slices each, will require at least 500 MB. Data can be archived on various magnetic and optical media. Alternatively, digital archiving using a picture archiving and communication system (PACS) may be preferred for convenient exchange and retrieval of data. To assure communication of imaging data between various imaging modalities, digital archives, evaluation workstations, printers, etc. from different manufacturers, a standard called DICOM (Digital Imaging and Communication in Medicine) has been developed. [32] 19

22 3. MDCT TECHNOLOGY Computed Tomography (CT) technology and its clinical applications have shown enormous resilience against alternative diagnostic methods and at the moment is stronger than ever. Enabled by technology that provides high power x-ray tubes, magnificent computing power, multi channel detectors to give sub millimetre slices with wider scan coverage, faster rotation times to complete one rotation in one third of a second, all have moved CT to dynamic applications in cardiology and 3-dimensional imaging of vascular and musculoskeletal anatomy [33] Introduction to MDCT Technology MDCT systems are CT scanners with a detector array consisting of more than a single row of detectors. The multi-detector-row nature of MDCT scanners refers to the use of multiple detector arrays (rows) in the longitudinal direction (that is, along the length of the patient lying on the patient table). MDCT scanners utilize third generation CT geometry in which the arc of detectors and the x-ray tube rotate together. All MDCT scanners use a slip-ring gantry, allowing helical acquisition at rotation speeds as fast as 0.33 second for a full rotation of 360 degrees of the X-ray tube around the patient. The primary advantage of these scanners is the ability to scan more than one slice simultaneously and hence more efficiently use the radiation delivered from the X-ray tube (Figure 10). The time required to scan a certain volume could thus be reduced considerably MDCT scanners can also be used to cover a specific anatomic volume with thinner slices. This considerably improves the spatial resolution in the longitudinal direction without the drawback of extended scan times. Improved resolution in the longitudinal direction is of great value in multi-planar reformatting (MPR, perpendicular or oblique to the transaxial plane) and in 3-dimensional (3D) representations. Spiral scanning is the most common scan acquisition mode in MDCT, since the total scan time can be reduced most efficiently by continuous data acquisition and overlapping data sets and this allows improved multi-planar reconstruction (MPR) and 3D image quality to be reconstructed without additional radiation dose to the patient. [33] 20

23 3.2. Differences between SDCT and MDCT One essential difference between SDCT and MDCT is how the thickness represented by an image, or slice, is determined. For a SDCT, slice thickness is determined by a combination of pre-patient and post-patient collimation. Therefore, the dimension of the detector array along the longitudinal axis can extend beyond the anticipated width of the x-ray beam or image slice (Figure 10) (i.e. the detector width is greater than the beam width). For MDCT, the converse is true and the x-ray beam width must be large enough to allow irradiation of all active detector rows (i.e. all those being used for a particular scan acquisition); slice thickness is instead determined by the width of the individual active detector rows. Figure 10: Schematics of detector rows and elements [33] In Figure 10, the single-detector row CT (SDCT) system on the left has one detector element along the longitudinal axis (indicated by z) and many (approx. 900) elements on the arc around the patient. The width of the detector (relative to the center of the gantry) is 20 mm, although the maximum beam width is only 10 mm. Thus the detector is wider than the x-ray beam. The multiple-detector row CT (MDCT) system on the right has 16 detector elements each of 1.25-mm along the longitudinal axis for each of the approximately 900 positions around the patient. The width of the detector 21

24 is also 20 mm at isocentre. The four data channels allow the acquisition of 4 simultaneous slices, of 1.25, 2.5, 3.75 or 5-mm width. Larger slice thicknesses (2.5 mm, 5 mm, 10 mm) can be generated by electronically combining the signal from several of these rows. Therefore the slice thickness used for the purposes of image review often differs from the slice thickness used for data acquisition. It may be larger, but never smaller. In this document, the term slice thickness always refers to that used for data acquisition (slice collimation). Due to the narrow width of the rows and the use of 4th generation geometry, gas ionization detectors are not used for MDCT scanners. In order to generate an image of a 1-mm slice of anatomy, detector rows of not much more than 1 mm in width must be used (detector dimensions are normalized relative to their coverage at the center of the CT gantry). The detector arrays are made from multiple rows, each approximately 1-mm wide (e.g. sixteen 1.25-mm wide detector rows). Another design for 4-MDCT detector arrays is illustrated in Figure 11. When small slices are desired, only the central portion of the array is used. It is therefore not necessary to have narrow rows in the outer portions of the array. The wider detectors at the periphery allow simultaneous acquisition of four slices each of 5 mm thickness. This design is somewhat less expensive and more geometrically efficient. 22

25 Figure 11: Diagram of the detector geometry used in a 4-MDCT from two major manufacturers. The detectors array is 20-mm wide along the longitudinal axis and uses eight rows of varying widths to allow simultaneous scanning of 4 slice up to 5- mm thick [33] Currently, MDCT systems are capable of acquiring up to 64 slices simultaneously in the z-direction (Figure 12). Three of the four manufacturers use 64 rows of either mm or 0.5 mm detectors. The fourth manufacturer uses 32 rows of 0.6 mm detectors and oscillates the focal spot to acquire 64 overlapping slices. This results in the reduction of spiral artifacts and improved spatial resolution along the longitudinal axis. Figure 12: Diagram of the detector geometries used in 64-MDCT from four major manufacturers. The Siemens 64-MDCT uses 32 sub-mm detectors and a moving focal spot to achieve 64 overlapping slice measurements [33] 23

26 For sequential data acquisitions (e.g. the table is stationary during the rotation of the x-ray tube around the patient), each channel collects sufficient data to create one slice or image, so as many as 64 independent images along the z axis could theoretically be reconstructed. For narrow slice widths, geometrical cone-beam considerations may limit the number of allowed images per rotation to less than 64. For example, one manufacturer s 16-detector scanner allows only 12 data channels to be used in sequential scanning because of cone beam considerations [34,35]. The primary attribute of MDCT systems is not the number of physical detectors rows, but the number of slices that are acquired simultaneously. The speed needed to cover a given volume is improved by a factor equivalent to number of slices included in the scan simultaneously. The reason why the number of simultaneous slices was initially limited to four was the amount of data to be acquired and transferred simultaneously. At that time, engineering and cost considerations limited the systems to four simultaneous data collection systems. Additionally, cone beam artifacts were not severe in 4-MDCT, but as the number of simultaneous slices increased, these artifacts become more problematic using conventional fan-beam reconstructions methods. Once 3-D cone-beam reconstruction algorithms (or advanced fan-beam algorithms with cone-beam corrections) and the increased computational power needed for these algorithms became available, 8- and 16-MDCT scanners were introduced. The advent of spiral CT introduced an additional acquisition parameter into the CT vocabulary, pitch. Pitch is defined as the ratio of the table travel per x-ray tube rotation to the x-ray beam width. With MDCT, a significant amount of confusion was introduced regarding the definition of pitch, as some manufacturers used an altered definition of pitch that related the table travel per x-ray tube rotation to the width of an individual data channel. The International Electrotechnical Commission CT Safety Standard specifically addressed the definition of pitch, reestablishing the original definition (table travel normalized to the total beam width) as the only acceptable definition of pitch [36,37]. This definition of pitch conveys the degree of overlap of the radiation beam: a pitch of 1 indicates contiguous radiation beams, a pitch less than 1 indicates overlap of the radiation beams, and a pitch greater than 1 indicates gaps between the radiation beams. Two manufacturers (Siemens and Philips) report the milliampere second (mas) as the average mas per unit length along the longitudinal axis, called either 24

27 effective mas or mas/slice, and calculated as actual mas/pitch. This distinction between mas and mas per unit length is important, because as the pitch is increased, scanner software may automatically increase the ma such that the image noise (and patient dose) remains constant with increasing pitch values [38-40]. When the effective mas or mas/slice is displayed, the user may be unaware that the actual ma is increased. On General Electric MDCT systems, the ma value is adjusted automatically to the value that will keep image noise constant as pitch or slice width is changed, and the selection box is turned orange to alert the user of the change in the prescribed ma value. [33] 25

28 4. MDCT DOSIMETRY The dose quantities used in projection radiography are not applicable to CT for three reasons: First, the dose distribution inside the patient is completely different from that for a conventional radiogram, where the dose decreases continuously from the entrance of the X-ray beam to its exit, with a ratio of between 100 and 1000 to 1. In the case of CT, as a consequence of the scanning procedure that equally irradiates the patient from all directions, the dose is almost equally distributed in the scanning plane. A dose comparison of CT with conventional projection radiography in terms of skin dose therefore does not make any sense. Second, the scanning procedure using narrow beams along the longitudinal z- axis of the patient implies that a significant portion of the radiation energy is deposited outside the nominal beam width. This is mainly due to penumbra effects and scattered radiation produced inside the beam. Third, the situation with CT unlike with conventional projection radiography is further complicated by the circumstances in which the volume to be imaged is not irradiated simultaneously. This often leads to confusion about what the dose from a complete series of, for example, 15 slices might be compared with the dose from a single slice. Consequently, dedicated dose quantities that account for these peculiarities are needed: the computed tomography dose index (CTDI), which is a measure of the local dose, and the dose length product (DLP), representing the integral radiation exposure associated with a CT examination. Fortunately, a bridge exists that enables comparison of CT with radiation exposure from other modalities and sources; this can be achieved by the effective dose (E). Therefore, there are three dose descriptors in all, which everyone dealing with CT should be familiar with. [2] 26

29 4.1. CT Dose Index (CTDI) The CTDI is the fundamental CT dose descriptor. By making use of this quantity, the first two peculiarities of CT scanning are taken into account: The CTDI [unit: milli-gray (mgy)] is derived from the dose distribution along a line that is parallel to the axis of rotation for the scanner (= z-axis) and is recorded for a single rotation of the X-ray source. Figure 13 illustrates the meaning of this term: CTDI is the equivalent of the dose value inside the irradiated slice (beam) that would result if the absorbed radiation dose profile were entirely concentrated to a rectangular profile of width equal to the nominal beam width NxT, with N being the number of independent (i.e., non-overlapping) slices that are acquired simultaneously. Accordingly, all dose contributions from outside the nominal beam width, i.e., the areas under the tails of the dose profile, are added to the area inside the slice Figure 13: Illustration of the term Computed Tomography Dose Index (CTDI) : CTDI is the equivalent of the dose value inside the irradiated slice (beam) that would result if the absorbed radiation dose profile were entirely concentrated to a rectangular profile of width equal to the nominal width NxT [2] The corresponding mathematical definition of CTDI therefore describes the summation of all dose contributions along the z-axis: 1 (2) CTDI D( z) dz NT where D(z) is the value of the dose at a given location, z, and NxT is the nominal value of the total collimation (beam width) that is used for data acquisition. 27

30 CTDI is therefore equal to the area of the dose profile (the dose profile integral ) divided by the nominal beam width. In practice, the dose profile is accumulated in a range of -50 mm to +50 mm relative to the center of the beam, i.e., over a distance of 100 mm. Figure 14: Total dose profile of a scan series with n=9 subsequent rotations. The average level of the total dose profile, which is called Multiple Scan Average Dose (MSAD), is equal to the computed tomography dose index (CTDI) if the table feed (TF) is equal to the nominal beam width NxT (i.e., pitch p = 1) [2] The relevance of CTDI becomes obvious from the total dose profile of a scan series with, for example, n = 9 subsequent rotations (Figure 14). The average level of the total dose profile, which is called multiple scan average dose (MSAD) [41], is higher than the peak value of each single dose profile. This increase results from the tails of the single dose profiles for a scan series. Obviously, MSAD and CTDI are exactly equal if the table travel per rotation (I) is equal to the nominal beam width NxT, i.e., if the pitch factor p I N T (3) is equal to 1. In general (i.e., if the pitch is not equal to 1,Figure 14), the relationship between CTDI and MSAD is given by MSAD 1 CTDI (4) p 28

31 The equivalence of the MSAD and the CTDI requires that all contributions from the tails of the radiation dose profile be included in the CTDI dose measurement. The exact integration limits required to meet this criterion depend upon the width of the total beam width and the length of the scattering medium. To standardize CTDI measurements, the FDA introduced the integration limits of ± 7T, where T represented the nominal slice width [41]. Interestingly, the original CT scanner, the EMI Mark I, was a dual-detector-row system. Hence, the nominal radiation beam width was equal to twice the nominal slice width (i.e., N x T mm). To account for this, the CTDI value, while integrated over the limits ± 7T, was normalized to 1/NT: CTDI FDA NT 7T 1 (5) 7T D( z) dz where D(z) represents the radiation dose profile along the z axis. However, the FDA definition neglected to account for the need to integrate over a longer limit (±7NT). The scattering media for CTDI measurements were also standardized by the FDA [42]. These consist of two polymethylmethacrylate (PMMA, e.g., acrylic or lucite,(figure 15) cylinders of 14-cm length. Figure 15: Cylindrical standard computed tomography (CT) dosimetry phantoms (16 cm and 32 cm in diameter) made from Perspex for representative measurements of the computed tomography dose index (CTDI) in regions of the head and the trunk, and a pencil-like detector for measurements of the dose-profile integral [2] 29

32 To estimate dose values for head examinations, a diameter of 16 cm is used, and to estimate dose values for body examination, a diameter of 32 cm is used. These are typically referred to, respectively, as the head and body CTDI phantoms. A CTDI obtained using a 100-mm chamber (the most commonly used type) is referred to as a CTDI 100. The CTDI 100, like the CTDI FDA, requires integration of the radiation dose profile from a single axial scan over specific integration limits. In the case of CTDI 100, the integration limits are ± 50 mm, which corresponds to the 100 mm length of the commercially available pencil ionization chamber [43-44,49] CTDI 50mm D( z) NT 50mm dz (6) CTDI 100 is acquired using a 100-mm long, 3 cm 3 active volume CT pencil ionization chamber and the two standard CTDI acrylic phantoms. The measurement must be performed with a stationary patient table. The CTDI can vary across the field-of-view. For body imaging, the CTDI is typically a factor or two higher at the surface than at the centre of rotation. The average CTDI across the field-of-view is given by the Weighted CTDI (CTDI w ) where: 1 3 CTDI w CTDI 100, CENTER CTDI 100, EDGE 2 3 (7) The values of 1/3 and 2/3 approximate the relative areas represented by the centre and edge values [45-49]. CTDI w is a useful indicator of scanner radiation output for a specific kvp and mas and is used as one of two dose descriptors for dose recommendations that have been introduced by the European Commission [49] To represent dose for a specific scan protocol, which almost always involves a series of scans, it is essential to take into account any gaps or overlaps between the radiation dose profiles from consecutive rotations of the x-ray source. This is accomplished with use of a dose descriptor known as the Volume CTDI w (CTDI vol ) [49], where: 30

33 NT CTDI VOL CTDIW I (8) In helical CT, the ratio of the table travel per rotation (I) to the total beam width (N T) is referred to as pitch; hence, CTDI VOL CTDI p W (9) So, whereas CTDI w represents the average absorbed radiation dose over the x and y directions, CTDI vol represents the average absorbed radiation dose over the x, y and z directions. It is conceptually similar to the MSAD, but is standardized with respect to the integration limits (±50 mm) and the f-factor used to convert the exposure or air kerma measurement into dose to air. CTDI vol is the parameter that best represents the average dose at a point with the scan volume for a particular scan protocol for a standardized phantom [47]. The SI units are milli-gray (mgy). It is a useful indicator of the dose for a specific exam protocol, because it takes into account protocol-specific information such as pitch. Its value is required to be displayed prospectively on the console of newer CT scanners. While CTDI vol estimates the average radiation dose within the irradiated volume of a CT acquisition for an object of similar attenuation to the CTDI phantom, it does not well represent the average dose for objects of substantially different size, shape, or attenuation. Additionally, it does not indicate the total energy deposited into the scan volume because is independent of the length of the scan Dose Length Product (DLP) To better represent the overall energy delivered by a given scan protocol, the CTDIvol can be integrated over the scan length to compute the Dose-Length Product (DLP) [49], where: DLP( mgy cm) CTDIvol ( mgy) scan length ( cm) (10) 31

34 The DLP reflects the total energy absorbed (and thus the potential biological effect) from a specific scan acquisition. Thus, while an abdominal CT might have the same CTDIvol as an abdominal and pelvic CT, the latter exam would have a greater DLP, proportional to the greater anatomic coverage of the scan Effective dose The radiation risk to the patients can be estimated from the effective dose (in msv), which can be calculated by summing the absorbed doses to individual organs weighted for their radiation sensitivity [30-31], hence E=w T X H T, where (11) H T is the equivalent dose in tissue T and W T is the tissue-weighting factor for tissue T. However, because we cannot obtain accurate measurements of all pertinent organ doses and the risk coefficients specific to age, gender and organ being irradiated, the estimated dose is calculated for an idealized 70-Kg, 30-year-old patient. In spite of these limitations, effective dose is the most widely used quantity for comparison between radiological procedures. There are various computer programs that can calculate dose for individual organs using the volume CT dose index and organ weighting factors from International Commission on Radiological Protection Publication 60 (ICRP 60) and International Commission on Radiological Protection Publication 103 (ICRP 103) [30-31]. From the effective dose, the risk estimates for stochastic effects can be determined with linear extrapolation of radiation exposure data from Japanese atomic bomb survivors. Calculation of the effective dose is usually made with mathematic anthropomorphic phantom or computer-simulated irradiation Monte Carlo techniques (statistical calculations of photon interactions). However, a reasonable approximation of the effective dose can be obtained with a conversion factor k (msv/mgy.cm), as presented in Table 1, that varies depending on the body region being imaged [48] 32

35 Table 1: Conversion Factors for calculating Effective dose Anatomic Region Conversion factor (msv/mgy.cm) Head Neck Chest Abdomen PARAMETERS AFFECTING DOSE IN MDCT 5.1. Relevant scanning parameters affecting patient dose The scan geometry, the tube current, the applied high tension, the scanning mode, the length of the scan, the speed of the couch, the speed of rotation of the gantry and shielding are some major parameters that influence patient dose in CT procedures. The operator can monitor most of these parameters and modify them to achieve the desired image quality with minimum possible patient dose. Scanning Geometry: The distance between the focal spot of the tube and the isocenter depends on the geometry. A single- or multiple-detector row helical CT can have a long or short geometric configuration. The intensity of the radiation beam varies between the source and the patient according to the inverse square law of distance. Therefore with all the other scanning parameters fixed, a short geometry scanner will produce more interactions in, therefore more dose to the patient and will have lower image noise than a long one. Tube Current: Reducing the tube current or the beam intensity is a way of reducing patient dose. A 50% reduction in tube current reduces dose by half. The beam energy and the photon fluence vary with the kvp and the tube current in a given procedure. The current-time settings (mas) are proportional to the photon fluence. Although some authors have claimed that it is possible to reduce the current without adverse effect on image quality [51-53], such reduction should be made with caution 33

36 because it is accompanied by increase in image noise, which degrades image quality. This is particularly so in abdominal scans where low contrast regions are greatly affected by noise. Tube Potential (kvp): This determines the radiation quality and its variation causes variation in patient dose. The relationship between kvp and image quality is complex because it affects both the image noise and the tissue contrast. Decrease in kvp causes increase in noise. This is particularly so when the patient size is large and the current is not appropriately increased to compensate for the low kvp. Dose is proportional to the square of the change in kvp while the latter is inversely proportional to the noise change. The choice of high tension is therefore crucial. An optimal kvp for abdominal scan for an averagely sized patient may be 120 kvp instead of 140 kvp, this will lead to 20 to 40% reduction in patient dose [54]. This value has to increase for a large size patient for adequate beam penetration [55] have published the results of a study showing that skull CT in children at substantially reduced tube potential with increased tube current produced the lowest possible dose without compromising the contrast to- noise ratio and image quality. Scanning Mode: All reconstruction algorithms require projection data for reconstructing the first and the last image in the spiral acquisition. At least half an extra rotation is therefore required at each end of the scan run to ensure reconstruction of the required target anatomy range; this is known as overranging/overscanning and substantially contributes to the overall radiation exposure. Its relative contribution is especially high when a small volume is scanned using a high pitch factor and wide detectors (Figure 16). In case of MDCT only the first detector row contributes to imaging [56]. As the acquisition proceeds, additional detector rows enter the imaging region until all the rows contribute. As a result, it is generally more dose-efficient to use a single helical scan rather than multiple helical scans if there are no overriding clinical considerations such as breath holding of the patient. The need to prescribe multiple contiguous helical scans should be infrequent with modern high-speed multi-detector row scanners. 34

37 Figure16: The effect of overranging/overscanning increases with the detector width and the pitch used (A, B). The smaller the scan field, the larger the extra radiation exposure due to the overscanning effect [96] Scanning length: With the increasing availability of helical CT scanners today, there is a tendency to extend the area of coverage to include regions beyond the actual area of interest in the chest, abdomen, or pelvis, which will further increase patient dose. Therefore, it is essential to draw the attention of referring physicians and radiologists to consider the consequence on the patient dose and to establish scanning protocols that restrict the examination to what is absolutely essential. Collimation, Couch Speed, and Pitch: In helical scanners, pitch is defined as the ratio of couch feed per 360 o gantry rotation to the normal collimator width of the x-ray beam. An increase in the pitch decreases the duration of exposure of the patient being scanned, hence the patient dose. Beam collimation, couch speed, and pitch are interlinked parameters that affect the image quality. Faster couch speed for a given collimation resulting in higher pitch will reduce patient dose, especially if other scanning parameters, including the tube current, are kept constant. This is because of the shorter exposure time, whereas narrow collimation with slow couch speed results in a longer exposure time, and hence higher patient dose. This is not true for scanners that use effective milliampere- second (mas) setting and maintain a constant mas value. In such scanners, the effective mas value is held constant irrespective of pitch value, so that the dose does not vary when the pitch changes. For a given collimation, an increase in couch speed increases the pitch and reduces the radiation dose [18,57]. Although scanning at a higher pitch is generally more dose efficient, it tends to cause helical artifacts, degradation of the section-sensitivity profile or section broadening, and consequently, decrease in spatial resolution. Alternations in pitch can have varying effects on image quality and in different situations. For instance, in CT colonoscopy [58], image quality and reconstruction artifacts are less affected by pitch 35

38 than by beam collimation, so that a higher pitch with narrow beam collimation are preferable for reducing dose. However, in some situations such as metastatic liver, which generally require thin collimation, an increased pitch may affect the detectability as the lesion may be missed owing to degradation of the sectionsensitivity profile. Due to overbeaming in multi-detector row CT, some amount of the x-ray beam is incident beyond the edges of the detector rows [54,59]. Generally, thicker beam collimation in multidetector row CT results in more dose-efficient examination, because overbeaming constitutes a smaller proportion of the detected x-ray beam. Depending on the scanner type, thick collimation limits the width of the thinnest sections that can be reconstructed. On the other hand, although thin collimation increases the proportion of overbeaming, it allows reconstruction of thinner sections. Hence, beam collimation and pitch must be carefully selected to address specific clinical requirements. For instance, a thicker collimation and a pitch greater than 1:1 is usually sufficient for screening in colonography and the urinary tract calculus. However scanning in certain clinical situations such as liver resection or transplantation, work-up is often performed with thin collimation and a pitch of less than 1:1. Gantry Rotation Time: There has been a dramatic decrease in the tube rotation times with recent technological innovations, most notably with development of 4-, 8-, 16-, 64-, 256- and 320 -detector row CT scanners. A 4-row scanner with 0.8 s rotation time requires 16 s breath hold to scan the entire abdomen, while an 8-row scanner will cover this length in 8 s. If the tube rotation time is decreased the exposure time will decrease and the tube current may have to be increased to maintain constant image quality. Modern 16-, 64-, 256- and 320 row scanners are capable of high scanning speeds and sub-millimeter section thickness. Thin collimation can lead to higher dose, especially if tube current is increased to maintain image noise at a level similar to that of thicker sections. The contrast resolution of small lesions improves because of reduced partial volume effect; hence greater noise on thinner sections may often be acceptable [60]. In addition, submillimeter collimation scans can normally be reconstructed as thicker sections, which reduce inherent noise. Thus it is important to optimize beam collimation for different multi-detector row scanners. 36

39 Shielding: Protection of radiosensitive organs such as the breast, eye lenses and gonads, is particularly relevant in pediatric patients and young adults, because these structures frequently lie in the beam pathways [61] have reported that with lead shield, thyroid and breast doses were reduced by 45 and 76% respectively in 110 procedures. Therefore shielding of the tissues not included in the examination is helpful in reducing patient dose. If the gonads are included in the field but are not the organs examined, some form of shielding could be used [62] have again reported reduction in dose to the testes up to 95%, using testis capsule during abdominal CT procedures, whereas lead apron is not appropriate for the ovaries due to their nonconstant position. Hein et al. [63] as well reported the use of shield for protection of the eye lens in paranasal sinus CT as a suitable and effective means of reducing patient dose by 40% Anatomical parameter consideration in dose reduction Most patient dose optimization methods involve modulation of the scanning parameters, especially tube current, based on patient weight and cross-sectional abdominal size. That is to say that weight and patient size also influence patient dose. Weight: Several investigators have suggested that mas value can be substantially reduced for CT of the chest in both adult and children [64,65]. Image quality identical to that in adult can be obtained in paediatric patients using significantly reduced exposures. For abdominal CT, Donnelly et al. [53] described modulation of scanning parameters in children on the basis of weight. They reported that paediatric patient weight can be used to select appropriate mas that are much lower than for adult in abdominal CT. They also suggested the use of substantially reduced mas for children weighing 4.5 to 68.0 kg. For abdominal CT in adults, tube current can be reduced on the basis of patient weight [52]. Selection of CT parameters on the basis of a patient s weight can lead to large variation in image quality between, for instance, two persons with the same weight but different heights. Cross-sectional Dimension: Attenuation of the incident x-ray beam in CT depends on the size of the body portion being studied; and greater exposure is required in corpulent patients to attain image quality equal to that in slimmer patients. 37

40 For the same exposure needed to compensate for a large size patient, the image quality is better with a slimmer patient because more photons reach the detector and the image noise is reduced. However, the dose to the slimmer patient is higher than necessary to produce good diagnostic image. Scanning parameters can, therefore, be modified on the basis of cross-sectional sizes to optimize patient dose. Haaga et al. [66] have reported that image noise was related to patient cross-sectional area and advocated the use of cross sectional measurements for optimizing scanning parameters and dose. A new method recently reported is patient dose variation in order to achieve similar levels of image noise for patients with different abdominal diameters [67]. Modulation of scanning parameters using anatomical diameter has yielded a dose reduction in slim patients and a significant correlation has been reported [52] between patient dose reduction, image quality and abdominal crosssectional parameters such as abdominal circumference, cross-sectional area, and anteroposterior and transverse diameters. At 50% reduced tube current (i.e. about 50% of the patient dose), image quality was acceptable in patients with a cross-sectional area of less than 800 cm 2, a circumference of less than 105 cm, a root mean square diameter of less than 44 cm, an anteroposterior diameter of less than 28 cm and a transverse diameter of less than 34.5 cm. Conversely, image quality with reduced tube current was unacceptable in patients with larger abdominal dimensions (i.e., exceeding the aforementioned values). These dimensions can be estimated before examination with a caliper. Alternatively, the technologist or the Medical Physicist can directly measure them on the CT console monitor. McCollough et al. [68] evaluated the use of size-based CT charts for reducing dose to paediatric and small patients and for improving image quality in large patients. They reported that modification of tube current in proportion to patient width is feasible and that it results in a 2- to 4-fold dose reduction in small patients Technical parameter consideration for dose reduction A variety of technical strategies that aim at decreasing patient dose in CT procedures have been developed, and many others are still in experimental stage. The majority of the technical innovations address patient dose optimization by improving scanning efficiency and image quality thus aiding image acquisition with reduced exposure. These innovations include collimation of X ray beams, use of better filters 38

41 and image processing algorithms, automatic tube current modulation, and efficient detector configuration and shielding. Beam Collimation: Focal spot tracking, control of x-ray tube focal spot motion, and beam collimation enhance scanning efficiency. Overbeaming is reduced by measuring the beam position every few milliseconds and continual repositioning of the source aperture so that a narrow beam reaches the detector. The beam is thus stabilized on the detectors, with exposure profile narrower than the detected x-ray profile, and the patient dose associated with multi-detector row is reduced in comparison to that of systems with no focal spot tracking. Beam Filtration: X-ray filters absorb the soft x-rays that constitute superfluous radiation which do not reach the detectors and thus do not contribute to image formation, but contribute to patient dose. Efficient filters selectively remove soft x-rays to reduce patient dose. Itoh et al. [69] compared doses with a 5.8 mmal with conventional filter in a phantom and patient study. They noted a 17% reduction in dose and a 9% decrease in image noise with the new filter. Bow-tie filters and beam shaping filters reduce the skin dose by 50% compared with flat filters [56]. Bow-tie and beam shaping filters minimize dose in the thinner proportion of patient, thereby providing better noise consistency within the image while saving substantial amount of radiation exposure. Automated Tube Current Modulation: In the CARE Dose system, during each rotation of the tube and detector assembly around the patient, a small number of the central detector channels provide attenuation information, which is dependent upon the patient cross section and scan angle, to the X-ray generating system [70]. The information provided by these detector channels is used to determine to what extent the ma can be modulated, with respect to an initial tube current setting, without adversely affecting the image quality. As a result the tube current is modulated dynamically with a delay of one rotation relative to the attenuation measurement. The first patient based assessment by Greess et al. [46] showed that, when CARE Dose is used, a dose reduction of approximately 25% (in terms of total mas reduction) is possible in pelvic scanning "with no significant decrease" in subjective assessments of image quality. Similar percentage dose reductions have 39

42 been demonstrated in other clinical work [71] and these showed good agreement with phantom based data [70]. Most of the published work has used image noise and/or subjective image assessment to quantify image quality. A small number of papers, Mastora et al [72] and Jacobs et al [73] have used standard deviations from regions of interest (ROIs) to yield a more objective assessment of image noise. Tube current modulation is a new technical innovation that can substantially reduce dose [74]. The concept of automatic tube current modulation is based on the premise that pixel noise is attributable to quantum noise in the projections. By adjusting the tube current to follow the changing patient anatomy, quantum noise can be adjusted to maintain the desired noise level. There are two current modulation methods used in CT scanners today: the longitudinal (z-axis) and angular (x and y-axis) modulation. In z-axis modulation (Figure 17), tube current is adjusted to maintain a user-selected quantum noise level. Noise is regulated on the final image to a level desired by the user. Z-axis modulation is the CT equivalence of the automatic exposure control systems used for many years with the conventional x-ray systems. It is an attempt to make all images have similar noise, independent of patient size and anatomy. The dose savings in z-axis modulation are expected to be greater than those with fixed-tube current methods since the tube current will be automatically reduced for smaller patients and anatomical regions. Z- axis modulation has been recently introduced for multi-detector row CT scanners such as Autom A by GE Medical Systems. Tube current modulation is determined from the attenuation and shape of scout scan projections in the patient just prior to the CT examination. Clinical results of these techniques have not yet been published in the literature. 40

43 Figure 17: Longitudinal dose modulation (LDM) is a refinement of AEC that adapts the mas settings slice-by-slice or rotation by rotation. Those parts of the scan range with reduced attenuation will be less exposed [2] In angular modulation (Figure 18), the tube current is adjusted to minimize x- rays projections that are of less importance for the reduction of the overall image noise. In anatomical parts that are highly asymmetric such as the shoulders, x-rays are much less attenuated in the anteroposterior direction than in the lateral direction [46]. Thus, the overwhelming abundance of anteroposterior x-rays can often be reduced greatly without a marked effect on the image noise. Angular modulation was first introduced on single-detector row scanners in 1994 [70]. Dose reduction of up to 25% was reported at that time, with virtually no change in image noise. On these early systems, both lateral and anteroposterior scout scans were required to determine angular modulation. More recently, angular tube current modulation has been introduced on multi-detector row scanners (CARE Dose by Siemens, Erlangen, Germany). In this system, the modulation is determined in real time by using projection data that lag by 180 from the x-ray generation angle. A recent investigation of 100 helical CT imaging studies in children in which angular modulations were used showed a 10 to 60% decrease in dose, with a mean reduction of 22.3% (neck, 20%; thorax, 23%; abdomen, 22 %) without loss of image quality [71]. 41

44 Figure 18: Angular dose modulation (ADM) is another refinement of AEC that adapts the tube current to the varying attenuation at different projection angles. Those projections with reduced attenuation will be less exposed [2] The ideal CT scanner will employ both z-axis and angular modulation techniques. When available in all commercial CT scanners, use of manual techniques, in which a tube current value is selected on the basis of some simple measurements on the patient (e.g. weight or cross-sectional dimensions), will be replaced with this computerized objective approach. With these developments, tube current modulation in CT scanners will be similar to photographic timing or automatic brightness controls like those currently used in conventional radiography. Indeed, automatic tube current modulation promises to be an important development in the optimization of scanning parameters that will help eliminate the guesswork involved in exposure parameters selection. Projection-Adaptive Reconstruction Filters: A marked decrease in signal is common in regions such as the shoulders due to beam attenuation in a particular projection. This leads to increased image noise and reduction in image quality that result from photon contamination by the electronic noise of the data-acquisition system. Projection space filters increase the filtration of signal-dependent noise in the reconstruction data and thus minimize the loss of resolution. Although there is some loss of image resolution accompanying the use of these filters, this is less than 5%, and the use of projection-adaptive reconstruction filters prevents an otherwise diagnostically compromised image. Kachelriess et al [75] investigated the use of multi-dimensional generalized adaptive filters for reducing image noise and patient dose. They recorded 30 to 60% reduction in image noise, typically along the direction of the highest attenuation in the noncylindrical body regions such as shoulders and metallic implants, without an increase in radiation dose. 42

45 Noise Filters: As discussed earlier, patient dose reduction is limited by increased image noise that can obscure lesions otherwise visible with standard parameters. Noise-reduction filters have been designed to decrease image noise and patient dose. Alvarez and Stonestrom [76] reported that two-dimensional linear filtering of the image may alter the spatial resolution and noise of CT images. They developed filters that minimized the variation in noise subject to a constraint on spatial resolution, with a 17% reduction in noise variance in comparison with that of conventional filters. Use of nonlinear image-processing techniques for improved quality CT images obtained with lower doses has also been reported [77]. Recently, Yu et al [78] reported the use of a new algorithm for reconstruction of CT images with noise properties superior to those of image reconstruction with the conventional fan-beam filtered-back projection (FFBP) algorithm currently used in commercial CT systems, including multi-detector row scanners. This algorithm converts the fan-beam data to non-uniformly sampled parallel-beam data using the Fourier shift theorem in the angular direction. The approach performs ramp filtration on non-uniform sampling grids along the radial direction before back projecting the filtered data to form the image. The decrease in noise with this algorithm may be translated into reduced patient dose and enhanced detection of subtle lesions, compared with reconstruction based on the current widely used FFBP algorithm. Noise reducing filters have also been designed on the basis of the principle that a group of structural pixels representative of structures of interest and a group of non-structural pixels representative of non-structural regions are both present in any image [92,93]. The structural pixels can be identified by determining gradient values for each pixel and by identifying pixels with a desired relationship to the gradient threshold value. The noise reducing filter technique involves isotropic filtering of non-structural regions with a low-pass filter and directional filtering of structural regions with a smoothing filter operating parallel to edges and an enhancing filter operating perpendicular to the edges. A blending parameter regulates the recombination of the structural and nonstructural segments. Noise-reducing filters decrease noise on low-dose CT images but adversely affect contrast and sharpness and may therefore decrease lesion contrast [52,92]. Further improvement in the technique is needed to maintain image contrast while decreasing image noise. 43

46 6. PERSPECTIVE ON RADIATION RISKS Deterministic risk. Although CT contributes a large part of the collective dose, in some countries it amounts to 70% of the dose from medical procedures, the individual patient skin dose in a single procedure is far below that which should cause concern for deterministic injury. This is unlike interventional procedures where peak skin doses in patients have been reported to cross threshold dose for skin injuries and a number of severe skin injuries have been reported [81]. Still the deterministic effects cannot be ruled out as a patient may undergo more than one radiological procedure. In a recent paper, Imanishi et al [82] reported three cases of temporary bandage-shaped hair loss which occurred in patients who had combination of perfusion studies with MDCT and cerebral digital subtraction angiography (DSA) (Figure 19). In all these patients two cerebral angiographies had been performed in the same period as the serial CT examinations. The possibility of such deterministic effects cannot be excluded if multiple radiological procedures are performed on the same patient. Figure 19: Bandage-shaped hair loss in a 53 yeas old woman with subarachnoid haemorrhage. Temporary hair loss lasted for 51 days was seen on day 37 after the first perfusion study of the head with MDCT. In this patient four perfusion studies of the head with MDCT and two angiographies of the head had been performed within the first 15 days of admission to the hospital. (Reprodused with permission from author, Imanishi et al.2005) [33] 44

47 Stochastic risk. It is not possible to prove that a particular cancer in a patient was caused by the few tens of mgy organ doses from a few CT examinations performed earlier in the life of an individual. However, on statistical bases, the exposures encountered in CT examination may increase the risk of certain cancers, especially in children [83]. The lifetime cancer mortality risks per unit dose vary with age. The BEIR VII report states that for the same radiation in the first year of life for boys, produces three to four times the cancer risk as exposure between the ages of 20 and 50 [84]. Further, female infants have almost double the risk as male infants. It is important that society protects those most at risk. CT examinations in children of up to 15 years of age, in many centres, account for nearly 15 to 20% of all CT examinations and the repeat rates of CT are increasing. Since the revelation in 2001 that exposure factors in CT of children are sometimes kept the same as for adults [85,86], there has been a definite increase in awareness about the need to tailor exposure factors for children, with new tools from manufacturers assisting users in this [91] and accreditation and regulatory emphasis on the absolute necessity of adjusting CT doses to patient size. 45

48 7. APPROACHES TO THE DOSE PROBLEM OF MDCT The answers to the challenges facing the use of MDCT must come both from technological development and from the clinical practice. On the industrial side the significant developments that have already been achieved in dose-constraint technology must continue and must impact on the way that MDCT operates in practice, as described in the following chapters. These advances in practice must be based upon a clear perception of the factors important in protecting the patient in MDCT, as outlined below The ALARA Principle The ALARA principle states that all medicinal exposure for diagnostic purposes shall be kept as low as reasonably achievable. It is based on the radiation assurance recommendations of various international expert committees and organizations and forms the cornerstone of radiation protection. Based on the assumption that there is no lower threshold for Carcinogenesis (i.e. that there is no dose that can be considered completely safe or harmless), the reduction of radiation exposure to ALARA remains an ongoing challenge. [2] 7.2. The Role of the Referrer: Justification It is a sine qua non of investigational medicine that the risk of the procedure is outweighed by the putative benefit to the patient. Although simple in essence, this principle may be difficult to put into practice. In many areas of established use of CT the potential benefit to the patient is clear and its application therefore well justified. However, patients are all individuals and in other areas it may be difficult to quantify accurately the potential benefit to the patient; in many instances, it is accepted, clinicians may tend to refer patients for examination in order to give themselves reassurance concerning their intended management regime; in such cases benefit is difficult to demonstrate. Encouraging referring clinicians to adopt a critical appraisal of their own referral practice may best meet the aims of radiation protection and of effective 46

49 justification and the ALARA Principle.The clinician needs to ask, before referring a patient for MDCT, Do I really need this investigation? Will it change what I do? If the answer to these questions is positive, the next critical question is to ask whether the information that is needed could be obtained without the use of ionizing radiation. In many abdominal and pelvic applications ultrasound and MRI provide acceptable alternatives to MDCT, and MRI is also an effective competitor elsewhere in the body. Even where these two techniques may not be as sensitive as MDCT, there may be a case for employing them first, especially in young patients, on the basis that if they yield the required information then exposure of the patient to radiation may not be required. In our own practice, the investigation of some cases of orbital fracture an application usually regarded as exclusively a requirement for CT has been successfully achieved using MRI. In such clinical decisions referral guidelines such as those issued by the Royal College of Radiologists (2006) in the UK (reference) have an established value. [2] 7.3. The Role of the Operator: Optimization It should be a given principle that all MDCT equipment is operated at optimum technical performance and subject to regular quality assurance. However, the objectives of optimization of the examination go beyond this. As indicated above, there are current technological advances which may be used to constrain exposure and, in appropriate circumstances, image quality can be manipulated to reduce exposure, provided that the resulting examination does not fall below an acceptable threshold of image quality and therefore of sensitivity appropriate to the clinical application. All departments should have in place local guidelines, based on the best evidence to date, to ensure that these objectives are met [2] 47

50 7.4. The Role of Guidelines in MDCT As indicated above, the evidence base for dose constraint in CT is not strong and in these circumstances practice guidelines may be important. In 1994 the European Commission set up a working group on image quality and dose in CT, resulting in publication in 2000 of the European Guidelines on Quality Criteria for Computed Tomography [48]. This group has continued and is currently producing a second edition of the guidelines (ref), which concentrates on MDCT. The second edition of the guidelines surveys technical and clinical principles in MSCT and make recommendations on good technique in 26 common areas of application, together with the guidelines on dose measurement and audit. Particular attention is paid to pediatrics. The group has also been active in promoting research studies to generate an evidence basis, principally a European field survey. One problem that the group has had to face is the variation in the performance of individual CT scanners. Whereas in the first edition it was possible to make specific recommendations on slice thickness and pitch, only ranges can now be specified. As in the first edition, the guidelines recommend quality criteria that enable examinations to be assessed. However, the key issue of diagnostic effectiveness and exposure still needs to be addressed by robust research studies. [2] 7.5. The Role of Evidence: Vigilance Overall, experience indicates that the dramatic rise in applications of CT has not yet reached a plateau. This is despite the fact that both technically and clinically, MSCT may be used in a way to aid dose constraint [87-89]. A number of factors actually offer the potential of dose reduction if taken into consideration by clinicians. For example, repeat scans which were frequently required if the patient moved significantly or breathed between single scans have been practically eliminated by MDCT. Overlapping scans, which were often selected for good multi-planar or 3D displays and led to corresponding increases in dose, are no longer a necessity because overlapping images are routinely available in helical CT with no additional exposure. Also, the selection of pitch factors greater than 1 results in a reduction in dose corresponding to the pitch factor [90]. Significant reduction of dose can also be obtained through attenuation-dependent tube current modulation which allows 48

51 constant image quality to be maintained regardless of patient attenuation characteristics and is now widely available on most MSCT systems [89]. It is important that all practitioners in CT continue to review emerging evidence and adapt their practice accordingly. For the present dose audit remains mandatory and further surveys of practice are required. Departments must ensure that their justification criteria are soundly applied, and that examinations are carefully targeted to clinical applications and do not exceed the clinical requirements. Where evidence supports the approach, exposure should be adjusted to the lowest threshold that delivers the required clinical sensitivity. It is necessary to follow published guidelines and observe all updates in these. Overall, the challenge of patient exposure in MDCT will best be served by continuing vigilance; from the manufacturers towards new dose-soaring developments and advice to their uses, from clinical referrers to ensure that over-demand is avoided, and from radiology department staff to ensure that the principles of best practice are always applied. This is, therefore, a field in which understanding of the balance between risks and benefit is most likely to be served by effective inter-disciplinary communication and education [2] 49

52 8. MEASURING DEVICES-MATERIALS All the dosimetric calculations and measurements analyzed in this project were realized through computed tomography systems (Philips, Brilliance16 and Brilliance 64), in the University Hospital Attikon in Athens. The in vivo measurements of the patients tested in the computed tomography were made with the help of calibrated TLD-100 H. For the reading of the registered signal, we used an LTM Reader from FIMEL Company. We also used an annealing oven ETT of the same company that has an appropriate heating program for the used dosimeters. A solid-state detector R100B was used for the calibration of the TLD-100H dosimeters. The direct reading of the signal of the detector was done by an electrometer (Barracuda). The detectors and the electrometer used for that purpose are from RTI electronics. Finally, the calculation of the dosimetric quantities, CTDIvol, DLP and Effective dose, was done with the help of a dedicated software tool. 50

53 8.1.CT scanners Brilliance 16 and 64, Philips Brilliance 16 and 64 CT scanners (Philips) are advanced systems of 3 rd generation computed tomography using the technology of slip-ring which covers clinical applications of wide range. The table below (Table 3) presents the basic characteristics of the two multi-slice CT scanners Table 2: Basic features of Brilliance 16 and Brilliance 64 CT scanners Features Brilliance 16 Brilliance 64 No of slices Detector type solid state solid state Focus-icocentre distance(mm) Tilt range(degrees) ±30 0 (0.5 0 increment) ±30 0 (0.5 0 increment) kv settings 90, 120, , 120, 140 ma range (and step size) (1mA steps) (1mA steps) Rotation times(s) 0.4, 0.5, 0.75, 1.0, 1.5, , 0.5, 0.75, 1.0, 1.5, 2.0 Nominal Z-axis X-ray beam collimations(mm) 1.2, 12, 18, , 1.25, 10, 20, 40 Pitches Detector width 0.6, 0.75, 1.5, , 0.625, 1.5 Scan FOV 25, 50 25, 50 Reconstruction matrices 512 2, 768 2, (optional) 512 2, 768 2, (optional) Maximum amount of computed RAM(GB)

54 8.2.Equipment for Thermoluminence Dosimetry TLD-100H TLD-100H is an advanced, relatively new dosimetric material with near tissue-equivalence (z eff =8.2), flat energy response and the ability to measure beta, photon, and more importantly, neutrons all from the same base material. The simple glow curve structure provides insignificant fade over extended dosimetric periods. Some basic features of TLD100-H presented in Table 2. Table 3: Characteristics of TLD-100H Type TLD-100H Formation chips Material LiF: Mg, Cu, P Dimensions (mm 3 ) 3.2 x 3.2 x 0.89 Useful range (Gy) Gy Fading Negligible Light sensitivity Weak Main Peak ( 0 C) 230 Annealing Temperature (C o ) 240 for 10 min TL emission spectra (nm) 400 Because of the hygroscopic attribute, TLD should not be touched with the hands. For this reason appropriate tweezers were used all the time. TLD-100H annealing never gets over 240 o because at annealing temperatures higher than this, differences of only 1C o lead to significant losses of sensitivity. 52

55 FIMEL ETT Annealing oven The oven ETT is a semiautomatic device for the annealing of thermoluminescence dosimeters (Figure 20). It has: The ability of temperature and heating/preheating time choice The ability of heating/preheating temperature augmentation slope choice Automatic ejection of TLD container after heating process Vacuum tweezers for handling TLD dosimeters (rods, chips) Figure 20: ETT Fimel annealing oven The annealing cycle was done with the ETT Fimel oven and the manufacturer recommendation for Harshaw TLD-100H was followed: 240 C o for 10 minutes LTM TLD Reader General principle of the FIMEL Reader is shown in Figure 21. The reader is composed of following parts: A dark room, in which TL material is heated. Emitted light is detected by a photomultiplier (PM), equipped with suitable optical filters. 53

56 A heating system with servo controlled temperature, to allow heating the TL material according to the chosen thermal cinetimatic chain. An electronic device which determines dose by integrating P.M. current between 2 predetermined temperatures (or 2 moments) It also ensures servomechanism operation, P.M. temperature regulation, Reader parameter monitoring, and information exchange with the computer. Figure 21: FIMEL Reader functional diagram 54

57 8.3.R100B Solid State Detector The solid-state detector that used for the calibration of TLD-100H dosimeters is R100B dose detector from RTI electronics (Figure 22). The weight of the detector is approximately 85 g, with cable length of 2.0 m and dimensions of 19.8 x 45.0 x 7.4 mm 3 The detector s main part is a PIN semiconductor photo diode. The diode detects the X-rays and produces a current proportional to the intensity of the X-rays. As the diode responds differently when exposed to X-rays with different spectral content, an energy filter is placed in front of the detector to compensate for this energy dependence. To make the detector more immune to EMI (ElectroMagnetic Interference) one more component is used. As the triaxial cable is not 100 % shielded, HF components might be induced on the signal lines. To prevent these from being rectified by the diode, and thus affecting the measurements, a ferrite core which filters these components, is placed on the end of the cable close to the connector. The currents, and charge, produced by the photo diode can then be measured with an electrometer (Barracuda) and converted to dose or dose rate. The measuring dose of R100B ranges from 0.10 ngy to up 100 KGy and the measuring dose rate range from 1.0 ngy/s to 76 mgy/s with inaccuracy of ± 5%. Figure 22: R100B solid state detector and Barracuda electrometer 55

58 8.4.Software Tool Specialized software tool was used to automatically extract the DICOM header information of CT images. Then DICOM data were extracted in a Microsoft Excel spreadsheet format for the calculation of CT dose quantities (CTDIvol, DLP, Effective dose). This spreadsheet use CC tables that have been derived using the Impact CT Patient Dosimetry Calculator [95] for the calculation of Effective dose. 9. DATA ACQUISITION AND PROCESSING 9.1. Patients data Brain, chest and abdomen-pelvis examinations were chosen for the calculation of the dosimetric quantities CTDIvol, DLP and Effective dose as well as the measurement of skin dose. Thus, for the calculation of the above quantities we chose 50 patients per each CT scanner for brain examinations, 51 and 58 patients for chest examinations, as well as 50 and 52 for abdomen-pelvis examinations, performed with Brilliance 16 and 64 CT scanners respectively. In the case of chest and abdomen-pelvis examinations, patients further divided in 3 groups according to the weight protocol that were used. Furthermore, for chest examinations with Brilliance 64, we have also two categories of patients. Patients who underwent this examination with and without the utilization of AEC system (Z DOM). From the total number of patients who underwent brain CT scanning, 15 and 18 patients were chosen for the measurement of eye lenses and thyroid skin dose, with Brilliance 16 and 64 respectively. As far as concern the eye lenses dose, patients further divided in two groups. The group of patients who received scatter irradiation in the lenses of the eyes and the group of patients who received primary irradiation in the lenses of the eyes. The categorization of the patients per each examination is shown in the Table 4 56

59 Table 4: Number of patients for A. Brain, B. Chest and Abdomen-Pelvis examinations A. Examinations Number of patients Calculation of CT dose In vivo measurement of skin dose descriptors Εye lens Thyroid 16 slices 64 slices 16 slices 64 slices 16 slices 64 slices Βrain Scatter Primary Scatter Primary B. Examinations Number of patients Chest AEC Without AEC kg kg kg Abdomen-Pelvis kg kg kg The total number of patients concerns adults from 18 to 75 years old. The average age is 50 years and the weight ranges from 41 to 91 kilos. It was not possible calculate the median weight because the weight of the patients wasn t registered during the examination. The estimation of the weight results from the protocol used during the examination. The sample of patients comes from both sexes and in Figure 23 it can be noticed the men s and women s number per each examination (brain, chest and abdomen-pelvis). 57

60 Number of patients Brain Brilliance 16 Male Female Brilliance 64 Chest Male Female Abdomen- Pelvis Figure 23: Sex distribution of the patients for the Brain, Chest and Abdomen-Pelvis examinations for the CT Scanners Brilliance 16 and Technique and scan data Both technique and scan parameters are significantly different concerning the type of the examination and the type of the CT scanner used. Thus, for the brain examination is normally chosen the axial technique (no involvement of pitch factor) based on a fixed scan protocol. The brain scan is performed with the maximum gantry tilt (30 0 degrees), which reduces by this way the dose to the eye lenses. In all cases we use standard kvp, mas as well as the same collimation. Table 5 provides the detailed scan information applying to Phillips Brilliance 16 and 64 Scanners. Table 5: CT protocol for brain examination CT parameters Brilliance 16 Brilliance 64 kvp mas Collimation(mm) Slice thickness(mm) Rot time(s)

61 The brain area is scanned only once, without the intravenous contrast medium, or twice before and after the contrast medium. In the majority of the cases we can see one scan of the brain area and so the brain examinations chosen for the dose calculation of the patients are only once scanned. Concerning chest and abdomen-pelvis examinations, the choice of the scan information is more complicated. The scan data of the patient depends on the selection of the appropriate weight protocol (Protocol A: kg, protocol B: kg, protocol C: kg), the scan area and the diagnostic particularity of the case. The scan parameters for chest and abdomen examinations in multi detector CT scanners (Brilliance 16 and 64) are analytically given in Table 6. Table 6:CT parameters for chest and abdomen-pelvis examinations with Brilliance 16 and 64 CT scanners CT parameters CT examinations Chest Abdomen-Pelvis Brilliance 16 Brilliance 64 Brilliance 16 Brilliance 64 Tube Voltage (kvp) Current-time product Protocol A (41-60 kg) Protocol B (61-80 kg) Protocol C (81-90 kg) Current-time product (Z-DOM) Protocol A (41-60 kg) Protocol B (61-80 kg) Protocol C (81-90 kg) Rotation time(s) , 0.75 Collimated slice width (mm) 16 x x x x Slice thickness (mm) Pitch Protocol A (41-60 kg) Protocol B (61-80 kg) Protocol C (81-90 kg) The scan technique applied in chest and abdomen-pelvis examinations is spiral. In all cases that were chosen and subjected to these examinations, we have only one scan without contrast. 59

62 Both CT scanners were equipped with automatic exposure control (AEC) systems with which the X-ray tube current is modulated during scanning in spiral mode. The ma values are determined utilizing the information concerning the attenuation characteristics of the anatomy scanned, which is obtained from the scout scans. The trade name of the AEC system of Philips CT scanners is Dose right DOM (D DOM, Z DOM). In case of abdomen pelvis examinations with both CT scanners and for a group of patients who underwent in chest examinations with Brilliance 64, Z DOM reduction system was used Εxtraction of data examinations CT images are stored in DICOM format and thus almost all the technical parameters with which each image was acquired and reconstructed are stored in DICOM headers. Specialized software (Dicom Info Extractor) was used to automatically extract the DICOM header information of CT images stored on DVDrom into a Microsoft Excel format file [99] Calculation of CT dose quantities The information of CT images pasted into a specialized Microsoft Excel spreadsheet containing embedded functions for the calculation of the dosimetric quantities. The calculation method is based on the assumption that the DLP can be considered as the sum of the contributions of the DLP fractions that correspond to each one of the reconstructed images plus the amount of DLP that is due to the overscan: The n CTDI vol of any type of scan can be derived from the n CTDI w using the following equation: n CTDIw n CTDIvol (12) pitch The contribution of each reconstructed image to the DLP, normalized per 100 mas, was termed as n DLP img and can be calculated using the following equation: 60

63 reconstruc tion interval n DLP img n CTDI vol (13) 10 The reconstruction interval values in DICOM headers are given in mm and are divided by 10 in order to calculate n DLP img values in units of Gy. cm. The absolute value was used in the formula because the reconstruction interval in the DICOM header may be a negative number. For each scan series the DLP can be calculated from the sum of the contributions of all images plus the contribution of the overscan, that is: k masi average mas( first and last image in each scan series) DLPc [( n DLPimg ) i ] n DLP(0) (14) i 1 In the above equation DLP c stands for the calculated DLP, n DLP img stands for the contribution (normalized per 100 mas) of each one of the k images of the scan series to the DLP c and mas i stands for the mas value in the DICOM header of i th image. In the last term of the equation it can be seen that the contribution of overscan (DLP o ) is calculated by multiplying the contribution of overscan (normalized per 100 mas) that was termed as n DLP (0) by the average mas of the first and last slices of each helical scan series, since when AEC is used these will be usually different. The average CTDI VOL of each scan series is calculated by the following equation: k masi 10 av CTDIvol Mean[( n DLPimg ) i ] (15) 100 reconstruction interval i 1 It is straightforward that if more than one scan series is performed during an examination, the total DLP c will be the sum of the DLP c values of all scan series. Total Effective dose, calculated as the sum of Eimg of all CT images and the contribution of overscan. The respective E of each image (Eimg) is obtained by multiplying the DLPimg by a conversion coefficient. This conversion coefficient selected from a table of conversion coefficient (ICRP 60 and one the ICRP 103 organ weighting schemes) [99]. 61

64 9.5. Eye lenses and thyroid skin dose measurement During brain CT scanning, sensitive dosimeters TLD- 100H were used on the skin for the approximate estimation of the dose that eye lenses and thyroid receive. Calibration is required before the use of the thermoluminescence dosimeters as it determines the relationship between the intensity of the signal of the dosimeter and the absorbed dose Calibration of TLD dosimeters Calibration includes the following stages: Initialization Batch Uniformity Linearity - Calibration curve Initialization Initialization is realized when the thermoluminescence dosimeters are used for the first time. For the use thus, of the TLD 100H dosimeters four heating cycles are required in order to stabilize their sensitivity. The heating program is on 240 C o for 10 minutes and then follows a sudden drop of temperature on 20 C o (fast cooling) Batch Uniformity This is a selection procedure from an initial group of crystals that have the same sensitivity and give the same signal when they are irradiated under the same conditions. Uniformity of each group is the variation of the values of the dosimeters concerning the average value of the group. It is expressed as a percentage of the standard deviation. The 80 TLD were irradiated in the X-ray tube with 120 kvp, 100mAs. All the dosimeters were irradiated together under the same conditions and at a distance of 1.5 meter from the focus, which is placed on patient table, as much as closer to its centre 62

65 in order to save the uniformity of the radiation and all the dosimeters to receive the same dose. As we need to use all the available dosimeters, we consider that they are part of only one group. First of all, we measure the signal of each dosimeter Ci and then we calculate the average value C mean of the entire group, which includes 80 dosimeters. For each dosimeter we calculate the personal response coefficient e cci, based on the relationship: e cci = C mean (16) The procedure is repeated tree times (N=3) in order to test the reproducibility of the dosimeters. The reproducibility of each dosimeter results from the relevant standard deviation of the average value of the signal of each dosimeter from the three irradiations according to the relationship: Reproducibility N N i 1 Ci 2 N N i 1 N 2 Ci N i 1 Ci 2 (17) From the first group of dosimeters we can conclude that the reproducibility value is smaller than 5 %. From the three measurements, we found the average value for each dosimeter of the response coefficient ecci (mean). For all the dosimeters, the average value ecci (mean) ranges from During the use of the dosimeters at the clinical practice we took into consideration the response coefficient for each dosimeter Linearity-Calibration curve The TLD-100H calibration was done by using the x-ray tube Super Rotalix of GE (General Electric). The TLD were situated at the center of the irradiation field on 63

66 Plexiglas plates (d=15 cm). Thus, the measurement included the backscattering radiation, too. Next to the dosimeters and on the Plexiglas plates there was an ionization chamber for giving a direct value of the dose. The field size was at a distance of 80 cm. The calibration device of the TLD dosimeters using x-ray tube is shown at Figure 24. X-ray tube Solid-state detector 80 cm TLD Plexiglas as Figure 24: Calibration device of TLD dosimeters using x-ray tube At each radiation we also used 5 dosimeters for the dose calculation of the radiation substrate. The radiation was done at 120 kvp for consecutive current values. From Figure 25, we defined the linear relationship, which was used for the dose identification arising from the TLD signal. 64

67 Dose (mgy) Calibration curve y = 6,0639x + 0,5083 R 2 = 0, TLD signal Figure 25: Calibration curve of thermoluminescence dosimeters TLD Setting and Patient Posture during Brain Examination In order to measure the scatter irradiation that thyroid and eye lenses received during brain CT scanning, sensitive dosimeters TLD100H were used. They were placed at the patient s skin upwards of the instrument of interest and before carrying out the examinations. In this procedure the total number of patients participated were 33, 15 and 18 patients for Brilliance 16 and 64 CT scanner respectively. Specifically, Lithium fluoride thermoluminescence dosimeters (TLD100H) were placed on the right eyelid, as an indicator of lens dose, and right lobe of thyroid gland for each patient. Thermoluminescence dosimeters (TLD) are useful because they are easily packaged, identifiable and radiolucent but they are also have systematic errors arise from differences between TLDs and TLD positioning. 65

68 The recommended standard positioning of the head was tucking the patient s chin to the chest. Due to the anatomical orientation of the cortex inside the skull, an angulated sequence of images was used because in case of normal slice sequence eyes exposed to direct radiation (Figure 26) The appropriate position of patient s head in conjunction with angulation of the gantry (30 o ) had as result to avoid irradiation of the eyes. Patients that have not taken the appropriate position of their head resulted lenses of the eye included in the scan field. Figure 26: Planning overview of a cranial tomogram. a, c Normal slice sequence;b, d angulated slice sequence [97] For this reason the above groups were further divided to patients that eye lenses were in the primary beam (N=5 and N=8 for Brilliance 16 and 64 respectively) and eye lenses were out of the primary beam (N=10 for each CT scanner). There are several limitations in our study. Because the lens lies close to the skin surface we exercised the assumption that entrance surface dose over the center of the eye (as measured with TLDs) was a reasonable estimate of the lens dose. Also in case of thyroid gland we have an approximation of thyroid dose by measuring the skin dose. 66

69 10. RESULTS Mean CTDIvol and DLP for brain, chest and abdomen-pelvis examinations derived from a) calculated values utilizing a specialized software tool and b) displayed values at CT scanners console. The mean effective dose was calculated from DLP using conversion coefficients from CC tables, which are specific to the CT scanner and tube potential value with which these images were acquired. These tables have been derived using the Impact CT Patient dosimetry Calculator both for ICRP 60 and 103 organweighting schemes. For the calculation of effective dose value, the percentage of dose due to overrange effect was taken into account and separately calculated as well. Finally, the mean value of skin dose was calculated for brain examinations utilizing Thermoluminescence dosimeters (TLD) in the area of eyes and thyroid for the approximate estimation of the eyes lenses and thyroid dose Calculation of CTDIvol and DLP CT dose descriptors for Brain,Chest and Abdomen-Pelvis examinations Mean values and standard deviations (SD) of scan length, displayed and calculated CTDI vol and DLP values, for the number of examinations have been referred (Table 4) and carried out with the two MSCT scanners, are presented in the table below (Table 7). The mean values of the differences observed among calculated and displayed values are also given for comparison. 67

70 Table 7: Mean values of scan length, CTDIvol and DLP CT dose descriptors (calculated and displayed) for Brain, Chest and Abdomen-Pelvis examinations The comparison of mean values of calculated and displayed CTDIvol and DLP dose quantities as well as the mean value of scan length for brain, chest and abdomenpelvis examinations are also presented in the figures below (Figure 27-35) 68

71 CTDIvol (mgy) Length (cm) Brain 14,0 13,9 13,8 13,7 13,6 13,5 13,4 13,3 13,2 Brilliance 16 Brilliance 64 Figure 27: Mean values of scan length for Brain examinations Brain 80,0 70,0 60,0 Displayed Calculated 50,0 40,0 30,0 20,0 10,0 0,0 Brilliance 16 Brilliance 64 Figure 28: Mean value of CTDIvol dose descriptor for Brain examinations 69

72 Lenght (cm) DLP(mGy.cm) Brain Brilliance 16 Brilliance 64 Displayed Calculated Figure 29: Mean value of DLP dose descriptor for Brain examinations Chest Brilliance 16 Brilliance 64 D-DOM Brilliance 64 Z-DOM Weight (kg) Figure 30: Mean value of scan length for Chest examinations 70

73 Figure 31: Mean value of CTDIvol dose descriptor for Chest examinations Figure 32: Mean value of DLP dose descriptor for Chest examinations 71

74 CTDIvol (mgy) Lenght (cm) Abdomen-Pelvis Weight (kg) Brilliance 16 Brilliance 64 Figure 33: Mean value of scan length for Abdomen-Pelvis examinations Abdomen-Pelvis Brilliance 16 Displayed Calculated Brilliance 64 Displayed Calculated Weight (Kg) Figure 34: Mean value of CTDIvol dose descriptor for Abdomen-Pelvis examinations 72

75 DLP (mgy x cm) Abdomen-Pelvis Brilliance 16 Displayed Calculated Brilliance 64 Displayed Calculated Weight (Kg) Figure 35: Mean value of DLP dose descriptor for Abdomen-Pelvis examinations Calculation of Effective dose and overscan contribution for Brain,Chest and Abdomen-Pelvis examinations The effective dose is one of the most important quantities for the evaluation of exposure. In recent years, tissue weighting factors used for the calculation of the effective dose have been updated in publication number 103 of the International Commission on Radiological Protection (ICRP) [31]. The know-how update of tissue weighting factors influencing the effective dose is clearly important. In this study, we calculated effective doses based on conversion coefficients according to ICRP 60 and 103 organ-weighting schemes, for head, chest and abdomen-pelvis examinations. In table 8 the mean values and standard deviations (SD) of E and average CC for the above examinations are presented as well as the percentage of the overscan contribution. 73

76 Table 8: Mean values and standard deviations (SD) of E and average CC are given for Brain, Chest and Abdomen-Pelvis examinations. The comparison of mean values of Effective dose and average CC, derived from ICRP 60 and 103 organ-weighting schemes for brain, chest and abdomen-pelvis examinations, is also presented in the figures below (Figure 36-41) 74

77 Effective dose (msv) CC [msv/(mgy.cm)] Brain 0,0035 0,003 ICRP 60 ICRP 103 0,0025 0,002 0,0015 0,001 0, Brilliance 16 Brilliance 64 Figure 36: Mean value of conversion coefficients derived from tables utilizing ICRP 60 and 103 organ-weighting schemes for Brain examinations Brain 3 2,5 2 ICRP 60 ICRP 103 1,5 1 0,5 0 Brilliance 16 Brilliance 64 Figure 37: Mean value of Effective dose according to ICRP 60 and 103 organ-weighting schemes for Brain examinations 75

78 Figure 38: Mean value of conversion coefficients derived from tables utilizing ICRP 60 and 103 organ-weighting schemes for Chest examinations Figure 39: Mean value of Effective dose according to ICRP 60 and 103 organs-weighting schemes for Chest examinations 76