| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
Full paper |
1Medical Physics Department, Medical School, University of Athens, 75 Mikras Asias, 115 27 Athens and 2Radiology Department and Medical Physics Unit, Konstantopoulio — Agia Olga Hospital, 3–5 Agias Olga, Nea Ionia, 142 33 Athens, Greece
 |
Abstract |
|---|
 Top Abstract IntroductionMaterials and methodsResults and discussionConclusionReferences |
|---|
|
Introduction |
|---|
|
TopAbstractIntroductionMaterials and methodsResults and discussionConclusionReferences |
|---|
The use of Monte Carlo techniques and previous studies in the field have led to the conclusion thatan indirect reliable estimate of E can be obtained by measuring the entrance surface dose (ESD) or the dose–area product (DAP) and multiplying these by appropriate conversion coefficients. They have been determined for specific X-ray projections, or even for complete examinations, through detailed analysis of each examination procedure [3, 4, 6, 7, 15, 16]. Thus, the estimation of E in diagnostic examinations has been simplified and reduced to a single measurement, encouraging its use for risk assessment in diagnostic radiology.
Intravenous urography (IVU) is the examination of the urinary tract involving up to 20 radiographs (mean of 8.2) [2]. For this reason, even if the IVU frequency is only about 1.3% of the total number of examinations, its contribution to the collective dose is much higher, equal to about 11% [5].
In the "National protocol for patient dose measurements in diagnostic radiology" [5], the measurement of the ESD is proposed for individual radiographs rather than for complete examinations where the DAP is the preferable dose related quantity. IVU does involve a large number of radiographs but the orientation of the patient throughout the examination does not change. Thus, ESD measurements are easy to perform, provided that appropriate dosemeters are available.
If such dosemeters are not available, an estimate of the ESD can by obtained by using measurements of absorbed dose to air, such as those performed during the quality control procedure of an X-ray unit (free-in-air measurements on the central beam axis) and appropriate backscatter factors (BSFs) [5].
The aim of this study is to compare the E values in IVU examinations obtained using ESD measurements with those obtained using DAP measurements. The conversion coefficients of ESD and DAP to E, noted in this study as CCESD and CCDAP correspondingly, have been determined for a mathematical phantom, for a specific geometrical set-up and for a selection of radiation qualities [7, 15]. By definition, both conversion coefficients should give the same E whether ESD or DAP measurements are used, provided that the theoretically assumed conditions are always met. However, since in clinical practice these conditions are not usually met, differences in the E values arising from ESD and DAP measurements are expected.
Furthermore, a secondary aim is to check the aforementioned assumption that in the absence of appropriate dosemeters to measure DAP or ESD, a reliable estimate of the ESD, and consequently of E, could be obtained by recording the exposure data for each X-ray projection and using measurements of absorbed dose to air in combination with literature data on BSFs.
Finally, the mean DAP and E per examination will be compared with the dose reference levels (DRLs) proposed for IVU and with other data from relevant studies [3, 5, 10, 17, 18].
|
Materials and methods |
|---|
|
TopAbstractIntroductionMaterials and methodsResults and discussionConclusionReferences |
|---|
The radiation measurements free-in-air were made using a Radcal Corporation dosemeter (Model 9010 Radiation Monitor Controller, 90X6-6 ionization chamber; Monrovia, CA). According to the manufacturers, all radiation measurements performed during the acceptance testing of the dosemeter employ NIST traceable techniques. The dosemeter has a nominal accuracy of ±4% and a reproducibility of ±1%. Its response to photons from 20 keV to 1.33 MeV (with build-up cap) is flat to within ±5%.
For DAP measurements, a Vacutec DAP meter was used (Vacudap 2000 with a 70157 flat ionization chamber; Dresden, Germany). According to data supplied by the manufacturer, the ionization chamber presents a filtration equivalent to 0.38 mm Al at 80 kVp and its nominal energy response ranges from 1.14 at 40 kV to 1.05 at 150 kV, having a calibrated response equal to 1 at 70 kV. For the tube potential range observed in IVU examinations of this study (70–90 kVp) the response variation is only about ±1%.
For ESD measurements, the skin dose monitor (SDM) (SDM-Model 104-101; McMahon Medical Inc., San Diego, CA) was used. It is a relatively new type of scintillation dosemeter, very small and light, that can be directly attached to the skin of the patient. No shadows are cast on the radiograph from the connecting cables, except for a tiny image of the dosemeter crystal unlikely to interfere with the diagnosis. The manufacturers supplied no information on its energy response butan extensive report on its performance characteristics can be found in the literature [19].
A schematic diagram of the instruments' set-up is shown in Figure 1
. The distances from the X-ray focus to the DAP meter position (i.e. the edge of collimation system), to the patient entrance surface and to the film position are noted as FCD, FSD and FFD, respectively, Do is the absorbed dose to air and A is the X-ray field area at the distances specified inside the parentheses. The DAP is the product of absorbed dose to air multiplied by the irradiated area and is constant at any distance from the focus. The relationship between DAP and ESD for a given field size at FFD is given by the following equation:
|
|
|
The SDM calibration was made in free air, 75 cm away from the focus in order to minimize errors from different focus-to-dosemeter distances. The SDM was positioned at the side of the Radcal dosemeter (about 5 cm apart to minimize any possible scatter contribution) and its response was adjusted so that its indication was similar to that of the Radcal dosemeter for80 kVp. Subsequently, an additional set of measurements for various tube potentials was made to assess the energy response of the SDM. These measurements are shown in Table 1, where it can be seen that for the tube potential range of interest (70–90 kV), both dosemeters gave almost the same readings. Outside this range, and especially for lower tube potentials, the response of the SDM decreased in agreement with the observations of Wagner and Pollock [19].
|
It should be also mentioned that the SDM readings were given in milliGrays with only one decimal digit and therefore for doses less than 1 mGy per exposure this may introduce significant inaccuracies. Furthermore, it was observed that when the SDM was turned on, the first one to two readings were about 10% less than the subsequent indications, for exposures made with constant mAs and kVp settings. This was not observed in the Radcal dosemeter or the DAP meter reading.
During the ESD measurements in the IVU examinations, the SDM was not positioned in the centre of the field, but somewhere in the middle ofthe centre-to-edge distance. In all X-ray views except those of the urinary bladder, this corresponded to a position just below the sternum and, with respect to the tube anode–cathode axis direction, to a position towards the anode. This position was chosen to project the SDM sensor crystal onto the spinal cord, where it was almost invisible, in order to avoid any possible complaints from radiologists not familiar with such images. As checked using measurements made employing a Plexiglas phantom, this can give readings of about 5% less because of X-ray field non-uniformity (heel effect), scatter reduction and inverse-square law effects. To make a more direct comparison of measured and calculated ESD, this reduction was accounted for in the ESD calculation, assuming a 0.95 correction factor.
To calculate the ESD, the X-ray O/P was determined afresh in the 50–140 kVp range (in 10 kV steps) using the Radcal dosemeter and with the DAP meter fitted, to account for the reduction in O/P caused by the additional filtration introduced by the DAP meter. The variation of O/P in µGy mAs-1 at 1 m from the focus with kVp is given in Figure 2 in a log(O/P)–log(kVp) chart. It can be seen that this variation can be fairly well described by a straight line (continuous line in Figure 2), allowing the calculation of the O/P for a given kVp. However, better fitting is obtained when considering only the data points in the 60–100 kVp range (dashed line in Figure 2), which fully covers the range of the kVp selections observed in the IVU examinations studied.
|
Furthermore, for each X-ray projection the ESD was calculated using the O/P measurements with the DAP on (dashed line in Figure 2
) and assuming a BSF of 1.39 for all projections, as the BSF variation for the field sizes and kVp used in these examinations is not significant. Calculated ESD values are noted as ESDC to differentiate them from measured values noted as ESD.
Following data acquisition, the total effective dose estimates from DAP, ESD and ESDC, noted as E(DAP), E(ESD) and E(ESDC), respectively, were calculated for each examination by summing the corresponding figures for each projection, using the CCESD and CCDAP values given in NRPB Report 262 [7] for a filtration of 4 mm Al, which more closely matches that of the X-ray unit used. The coefficients for the tomographic views were assumed to be equal to those corresponding to the anteroposterior kidney or abdominal views, depending on the selected area [3]. The conversion coefficients (CCs) are shown in Figure 3 as a function of kVp setting, and can be reproduced by the equation:
|
|
|
Apart from this, geometric deviations from the conditions assumed may also affect the measured quantity and therefore the E estimate. A different FSD will affect only the ESD and consequently the E(ESD), whereas a change in A(FFD) will affect only the DAP and consequently the E(DAP). While the effect of a different FSD and A(FFD) onthe ESD and DAP, respectively can be determined accurately, their effect on E(ESD) and E(DAP) values is unknown. It is not clear whether thedependence of the measured quantity on a geometric condition will partially compensate for the errors introduced by the approximate conversion factor or will add more errors to the E estimate.
If, for example, we consider a patient larger than the mathematical phantom, the FSD would be smaller than that assumed in NRPB Report 262 [7] and the field size required to cover the anatomical regions of interest would be probably larger. Considering only the effect of the smaller FSD, the ESD per mAs and consequently E(ESD) would also be increased compared with E(DAP). This does not necessarily mean that E(ESD) would be more accurate than E(DAP), as it has not has been taken into account that the field size on the surface of the patient is reduced (by the same factor dose is increased, so that the DAP remains constant). Even if this effect is ignored (as a larger patient would normally require a larger field size), it is also not taken into account that because of the greater patient thickness and the subsequent greater X-ray attenuation the mean doses of certain radiosensitive organs per mAs may also be reduced compared with the mathematical phantom. On the other hand, the use of a field larger than that assumed in NRPB Report 262 [7] for imaging a large patient would result in a proportional increase of DAP and E(DAP). It is possible, however, that this larger field for a larger patient would result in mean organ doses similar to those in the mathematical phantom for the standard field. Therefore, the dependence of DAP on field size settings instead of correcting the E(DAP) estimate could also introduce errors.
The advantages and disadvantages of DAP aresummarized in the European Commission's "Radiation Protection 109" [20], where the need for ESD measurements in cases of non-standard patients (e.g. paediatric) or fluoroscopic procedures is also stressed. The ESD is a more straightforward dosimetric quantity and is more appropriate when deterministic effects or organ dose calculations are of concern. In most available DAP meters, dose and area cannot be separated, but more sophisticated dosimetric systems have been manufactured that can simultaneously give all these quantities along with chamber to patient distance, thus enabling the ESD estimation.
It is a fact, however, that in clinical practice, field sizes different than those assumed in NRPB Report 262 [7] are used, not only because of the patient size but also as a result of the personal preference of the examining doctor or errors of the radiation technologist in diaphragm aperture settings. To quantify how field size variations affect E(DAP), the following equations are presented, assuming that all the other NRPB Report 262 [7] assumptions are met except the field area at FFD, which is A' instead of A, for a specific X-ray projection. The parameters corresponding to the area A' at FFD have been assigned the superscript '.
|
|
Equation (3) shows that when using the tabulated CCDAP values (determined for an assumed field size A) for an actual field size A', the effective dose will be equal to that expected if the field size was equal to that assumed, multiplied by the ratio of the field size used to the fieldsize assumed (thereafter referred to as fieldsratio). Now, if we take into account that when theassumed conditions are met we have E=E(DAP)=E(ESD) we can write:
|
|
|
|
Equations (3) to (5) illustrate that the increase or decrease of DAP, which is proportional to the fields ratio, has the same effect on E as if this wasdetermined by ESD measurements and the tabulated CCESD were corrected by the fields ratio. It is questionable however, if such a correction can be considered accurate.
The CCESD for the anatomical area of interest is almost proportional to the area, since the CCESD is about 0.0001 mSv (mGy cm2)-1 (at 80 kV, 4 mm filtration) as was deduced from kidneys, abdomen, pelvis/colon, lumbar spine and thoracic spine tabulated values [7]. The bladder is an exception since CCESD is about 0.00015 mSv (mGy cm2)-1, because of the inclusion of gonads [7]. Thus, for the majority of the X-ray views used in IVU the tabulated CCESD multiplied by the fields ratio could be considered as a fairly good approximation for the CCESD that should be used. For the bladder, this correction may lead to overcompensation or undercompensation, as the very radiosensitive organs may or may not be included in a particular field size and therefore the changes in the CCESD are not exactly proportional to the field size.
|
Results and discussion |
|---|
|
TopAbstractIntroductionMaterials and methodsResults and discussionConclusionReferences |
|---|
, whereas the comparison of measured with calculated ESD is shown in Figure 5. The statistical quantities concerning the DAP, ESD and E distributions are summarized in Table 2 for comparison with relevant studies.
|
|
|
, it is obvious that the E(DAP) values were in all cases larger than the corresponding E(ESD) and E(ESDC) values. For the 25 examinations studied, the mean E(DAP) is on average about 38% higher than themean E(ESD) and about 28% higher than E(ESDC).
To justify those differences, the effect of field size was considered and the E'(ESD) values werededuced for each X-ray projection, using Equation (4) and in Equation (5) setting the field size A'(FFD) to be equal to the size of the cassette used. This, however, did not significantly reduce the differences between E(DAP) and E'(ESD) estimates, except for the urinary bladder views where the 24 x 30 cm2 cassette used is larger (43%) than the 21 x 24 cm2 field assumed. As the remaining differences could be explained only if the X-ray fields were larger than the cassette size, the possibility that such an error had occurred was investigated.
The DAP divided by the mAs and the cassette size is the absorbed dose to air at FFD, and multiplied by (FFD/100)2 should agree with O/P values determined by the Radcal dosemeter against which the DAP meter was calibrated. However, the O/P calculated from DAP was often larger compared with the expected values, indicating that the radiation field was set larger than the cassette size as a result of errors in the manual setting of the diaphragm aperture. By dividing the O/P calculated by DAP readings by the expected O/P values and multiplying by the cassette size, the actual X-ray field area at FFD was approximated. The above procedure revealed many cases where the radiation field extended 2 cm over each cassette edge, which in the case ofa 30 x 40 cm2 cassette corresponds to a 24% increase of the irradiated area, and consequently of the DAP reading, without any possible diagnostic gain. Since the increase in radiation field size is unintentional in such cases, and could not be attributed to the patient size, it should be expected that E(DAP) values are more accurate than E(ESD), for the reasons considered in "Dose measurements and calculations". Substituting the approximated actual field size for A'(FFD) in Equation (5) significantly reduced the differences between E'(ESD) and E(DAP) values, as can be seen in Figure 6 and Table 2.
|
As far as the differences between measured andcalculated ESD are concerned, they are on average about 10%, excluding the last three examinations where some problems with the SDM occurred. For those examinations performed on the same day and, in particular, for the final examination, the ESD data are not reliable. The SDM during the last examination gave erratic readings because of the destruction of the SDM sensor, as was proven following a self-test procedure offered by the instrument. It can be seen in Figure 5 that in the last examination the difference between measured and calculated ESD values is 50%. This was not unexpected as the SDM sensors are expendable and after a number of examinations may present decreased dose readings and should at that point be replaced by new ones. Thus, to verify the reliability of measurements made with the SDM, the self-test procedure should be carried out at the beginning and at the end of each examination.
The generally good agreement between measured and calculated ESD values suggests that, in the absence of appropriate dosemeters, the ESD could be calculated with good accuracy by simply recording the mAs, tube potential (kVp) and FSD for each exposure. It should be noted, however, that for almost all cases, the measured ESD was smaller than calculated and this cannot be explained with certainty. It may be the result of the above-mentioned problems observed during the calibration of the SDM or due to a systematic error occurring in the calibration procedure, but it may also have been caused by the overestimation of BSF in relatively thin patients or in patients with an excess of bowel gas because of inadequate preparation [21]. If a BSF of 1.3 had been used, as Wagnerand Pollock [19] suggested, the ESDC would be on average only 3.7% larger than measured values.
The mean DAP value for the IVU examinations studied is 11.7 Gy cm2; well below the DRLs of 40 Gy cm2 [5, 17] and 20 Gy cm2 [18] proposed. Compared with the mean of 10.17 Gy cm2 recently reported for a sample of 205 patients [10], the mean DAP value of the present study is slightly higher. However, taking into account that in the referred study the average number of films per examination was only 3.7, the mean DAP per X-ray view in this study (where 9.3 films were used on average) is much lower. This should be attributed to the lower mean tube potential used in the referred study, that is 60 kV compared with 80 kV in the present study.
Finally, the average E(DAP) and E'(ESD) estimates of 3 mSv and 2.8 mSv, respectively are very close to the E value of 2.64 mSv, as was determined from detailed organ dose measurements in an anthropomorphic phantom for a simulated IVU examination [3], even though the mean values of applied peak potential (64 kV) and number of films (6.5) were different from the corresponding figures in this study.
|
Conclusion |
|---|
|
TopAbstractIntroductionMaterials and methodsResults and discussionConclusionReferences |
|---|
As in this study the larger field size used is not attributed to patient size, E(DAP) can be thought of as being more accurate than E(ESD). This is in agreement with Le Heron [16], who reported that"dose area product best mirrored effective dose", and the National Radiological Protection Board statement that "it is likely that dose–area product is more closely related to effective dose, since it takes into account the X-ray beam area which affects the number of organs irradiated" [7]. Nevertheless, the better accuracy of E(DAP) compared with E(ESD) cannot be generalized as E is dependent on many factors and the main advantage of DAP remains its practicality.
For ESD measurements in diagnostic radiology, the use of the SDM as an alternative to thermoluminescent dosemeters (TLDs) seems quite appealing as it is simpler to use and its main advantage of directly giving the dose indication could be surpassed only if the TLDs provide better accuracy and sensitivity, especially for low doses. However, no definite conclusions can be drawn without comparative studies in phantoms and in clinical practice.
Furthermore, it was also shown that a reliable ESD calculation in IVU could be obtained by simply recording exposure data and the FSD, a cost-free procedure that requires only data from the quality control of the X-ray machine.
Finally, as far as conforming with the DRLs is concerned, the average DAP values were well below the proposed values and, as shown, further decreases can be obtained by limiting the X-ray field sizes used for each X-ray projection.
Received for publication February 25, 2000. Revision received January 22, 2001. Accepted for publication April 4, 2001.
|
References |
|---|
|
TopAbstractIntroductionMaterials and methodsResults and discussionConclusionReferences |
|---|
This article has been cited by other articles:
 |
|
 |
|
  T. Niemann, T. Kollmann, and G. Bongartz Diagnostic Performance of Low-Dose CT for the Detection of Urolithiasis: A Meta-Analysis Am. J. Roentgenol., August 1, 2008; 191(2): 396 - 401. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
T. A. Jaffe, A. M. Gaca, S. Delaney, T. T. Yoshizumi, G. Toncheva, G. Nguyen, and D. P. Frush Radiation Doses from Small-Bowel Follow-Through and Abdominopelvic MDCT in Crohn's Disease Am. J. Roentgenol., November 1, 2007; 189(5): 1015 - 1022. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 C J Martin Effective dose: how should it be applied to medical exposures? Br. J. Radiol., August 1, 2007; 80(956): 639 - 647. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 R. D. Nawfel, P. F. Judy, A. R. Schleipman, and S. G. Silverman Patient Radiation Dose at CT Urography and Conventional Urography Radiology, July 1, 2004; 232(1): 126 - 132. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
D Bor, T Sancak, T Olgar, Y Elcim, A Adanali, U Sanlidilek, and S Akyar Comparison of effective doses obtained from dose-area product and air kerma measurements in interventional radiology Br. J. Radiol., April 1, 2004; 77(916): 315 - 322. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| BJR | DMFR | IMAGING | ALL BIR JOURNALS |
ESD%=(ESDC-ESD)/max(ESDC,ESD)].