This Article Abstract Figures Only Full Text (PDF) Services Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Citing Articles Citing Articles via HighWire Citing Articles via Google Scholar Google Scholar Articles by Siiskonen, T Articles by Vartiainen, E Search for Related Content PubMed PubMed Citation Articles by Siiskonen, T Articles by Vartiainen, E

Monte Carlo simulations of occupational radiation doses in interventional radiology

STUK – Radiation and Nuclear Safety Authority, PO Box 14, FIN-00881 Helsinki, Finland

Correspondence: Dr Teemu Siiskonen, Stuk–Radiation and Nuclear Safety Authority, PO Box 14, FIN-00881 Helsinki, Finland. E-mail: teemu.siiskonen{at}stuk.fi

	Abstract

Top Abstract Introduction Monte Carlo simulations Results and discussion Conclusions References

 
Occupational radiation doses in interventional radiology can potentially be high. Therefore, reliable methods to assess the effective dose are needed. In the present work, the relationship between the personal dose equivalent, H_p(10), the reading of a personal dosimeter and the effective dose of the radiologist were studied using Monte Carlo simulations. In particular, the protection provided by a lead apron was investigated. Emphasis was placed on sensitivity of the results to changes in irradiation conditions. In our simulations a 0.35 mm thick lead apron and thyroid shield reduced the effective dose, on average, by a factor of 27 (the range of these data was 15–41). Without the thyroid shield the average reduction factor was 15 (range 6–22). The reduction sensitively depended on the projection and the X-ray tube voltage. The dosimeter reading, when the dosimeter was worn above the apron and a thyroid shield was used, overestimated the effective dose on average by a factor of 130 (range 44–258) when the dosimeter was located on the breast closest to the primary X-ray beam. Without the thyroid shield the average overestimation was 69 (range 32–127). If the dosimeter was worn under the apron its reading generally underestimated the effective dose (on average by 20% with the thyroid shield). Our study indicates that, even though large variations are present, the often used conversion coefficient from the dosimeter reading above the apron to the effective dose, around 1/30, generally overestimates the effective dose by a factor of two or more.

	Introduction

Top Abstract Introduction Monte Carlo simulations Results and discussion Conclusions References

 
Radiologists (or cardiologists) who perform interventional X-ray procedures have to work close to the primary X-ray beam where the scattered photon field from the patient and equipment is most intense. Catheters and other equipment must often be manipulated while using fluoroscopy, resulting in long exposure times and appreciable doses to the radiologist. Several protective actions can be taken to reduce the radiation burden of medical staff. For example, a lead apron, thyroid shield and protective eye glasses are reasonably comfortable to use and they can reduce the dose significantly. In addition to such personal protective shielding it is also important to put emphasis on work techniques (using protective screens of the X-ray equipment, keeping as far from the radiation beam as possible and minimizing the fluoroscopy time, field size and dose rate used in the procedure) and to use technical dose-saving methods in modern fluoroscopy systems, such as pulsed fluoroscopy and the last image hold feature.

Regular occupational dose monitoring is needed for the verification of compliance with legal dose limits and to assess the effectiveness of the protective measures. For that purpose, a personal dose equivalent at 10 mm depth, H_p(10), is often measured with a dosimeter worn either above and/or under the protective clothing. Many authors have assessed the exposure of the radiologist, using measured or simulated data, and give a relation between the dosimeter reading and the effective dose E. Among others, Niklason et al [1] have proposed means for combining the readings of dosimeters to obtain an estimate of E when dosimeters are worn both below and above the lead apron. Various practical approaches have been discussed by Schultz and Zoetelief [2].

The role of protective clothing and screens has been discussed by, e.g. Tsapaki et al [3] and Vañó et al [4]. In the paper by Vañó et al, it was shown that ceiling-mounted protective screens, if used properly in addition to a protective apron, can reduce the effective dose of the radiologist by a factor of three. Williams [5] concluded that the dose–area product (DAP) is a useful indicator of the effective dose of the radiologist and helps to identify procedures which may potentially lead to high radiation doses.

Measurements of absorbed doses from scattered radiation in phantoms that are shielded with a protective apron are cumbersome because the dose rates in the phantom are very low, and accurate measurements require long irradiation times. However, Monte Carlo simulations provide a means of assessing the effective dose of the radiologist for cases where the DAP value or the dosimeter reading is known. Earlier, Schultz et al [6] have used Monte Carlo simulation to evaluate the radiologist's effective dose due to radiation scattered by a 5-year-old cardiac patient. Other publications on the subject have been based on more approximate dosimetric calculations and/or measurements.

Even though the above-mentioned studies give insight into occupational effective doses, more detailed data are needed, in particular on the effect of the lead apron. Often, conversion from the measured air kerma or personal dosimeter reading to effective dose is complicated by the unknown spectral fluence and the complex and poorly specified irradiation geometry, i.e. the actual conditions differ greatly from the conditions where the measurement instruments were calibrated. Different (empirical) correction factors must then be used to account for the influence of the lead apron. In Monte Carlo simulations these conversion and correction factors can be consistently calculated (i.e. within one model), therefore potentially reducing the uncertainty of the calculated dose.

The aim of the present study is to use Monte Carlo simulations to estimate the personal dose equivalent, effective dose-and radiation protection effectiveness of the lead apron in different radiological procedures (eight cardiac and two cerebral exposure conditions). The effectiveness is estimated for each situation by calculating the radiologist's effective dose with and without the lead apron. We also present data for estimating the radiologist's effective dose from the dosimeter reading (the personal dose equivalent), when the dosimeter is worn above or under the protective clothing and from the DAP data. The former information is valuable especially when dosimeter readings close to the dose limits are rated and the latter is a useful indicator of the radiologist's dose in individual examinations. Often, the only available dosimetric information for estimating the dose to the radiologist is the dosimeter reading which may be a sum over a large number of different procedures. In actual practice, accurate information on the exposure conditions is rarely available and the effective dose cannot be calculated accurately. Therefore, particular emphasis is placed on the sensitivity of the results with respect to variations in simulation parameters: geometrical factors, the X-ray beam filtration, the tube voltage and the field size.

	Monte Carlo simulations

Top Abstract Introduction Monte Carlo simulations Results and discussion Conclusions References

 
Simulations were done with the MCNPX Monte Carlo code version 2.5f [7], running under the Linux operating system. Photon cross-sections were taken from ENDF/B-VI, release 8 (see [8]). Energy deposition (equivalent doses) was calculated by track length heating estimation (F6 tally in MCNPX language). The personal dose equivalents (H_ext and H_p(10) below) were calculated with point detectors. Bremsstrahlung and photon coherent scattering were ignored in the simulations. Typical calculation times varied from 1 to 3 days per case (with a 1.6 GHz processor), depending on the projection and use of an apron. The X-ray spectra were calculated with the method of Birch and Marshall [9].

Geometry
Mathematical MIRD-type hermaphrodite phantoms of the patient and the radiologist were used. The phantoms were based on the reports by Cristy and Eckerman [10] and Eckerman et al [11] and constructed for the calculations with BodyBuilder^TM software [12]. Legs were not included in the patient model, since their contribution to the scattered X-ray field was negligible in the irradiation conditions studied. For the same reason, the patient's head was not included in simulations of procedures aimed at the trunk. Similarly, the patient's trunk was not included in simulations of cerebral procedures. The phantoms consisted of soft tissue, lungs and skeleton. The organs composed of soft tissue were not modelled separately in the patient phantom, but were considered in the phantom representing the radiologist. Both persons were modelled as standard sized phantoms of weight 73.5 kg (for the full phantom with legs, trunk and head) and a trunk length of 70 cm, width of 40 cm and thickness of 20 cm. Head and the trunk of the patient phantom were taken from this full standard phantom. The composition of the materials was taken from [10] (skeletal and lung tissue) and [13] (soft tissue excluding the lungs), based on the ICRU data [14].

The models of the X-ray source, the patient in a supine position and the radiologist standing on the right-hand side of the patient (Figure 1) were defined in an air-filled sphere (75.5% of N, 23.2% of O, 1.3% of Ar and 0.01% of C). The X-ray tube focal point was placed at positions shown in Figure 2 and the beam of photons was limited to a rectangular cone with perfectly absorbing collimators. The exposure conditions are listed in Table 1. The field size was defined in the plane which is perpendicular to the X-ray beam central axis and includes the intersection point of the X-ray beam central axis and the patient's surface. The X-ray beam central axis was normal to the longitudinal axis of the patient. Additional equipment, like the patient support couch, the antiscatter grid and the image intensifier, was not modelled.

View larger version (18K): [in this window] [in a new window]   Figure 1. On the left(a), the radiologist performing a cardiac procedure with the lead apron and the trunk of the patient, viewed from above. The grey square indicates the position of the X-ray field. On the right (b), the same arrangement is viewed from the left-hand side of the radiologist. The position of the thyroid is indicated by the black rectangle. The coordinate system and the (suggestive) length scale are indicated in the figure. Here, the focal spot of the X-ray tube (black dot) is at z = –70 cm and corresponds to the posteroanterior (PA) projection. The primary X-ray beam is indicated by the dashed cone. The external personal dosimeter is indicated with a grey sphere.

View larger version (10K): [in this window] [in a new window]   Figure 2. Definition of the projections used in this work. The arrows indicate the direction of the primary X-ray beam. The focal point of the X-ray tube is located on the dashed ellipse. The positions of the patient trunk and head (not to scale) are shown by black ovals (the view is towards the feet of the patient, to the negative x-axis direction). See Figure 1 for details on coordinate system and positioning of the patient and radiologist. Abbreviations for the projections are: PA, posteroanterior; LLAT, left lateral; RLAT, right lateral; AP, anteroposterior; RAO, right anterior oblique; RPO, right posterior oblique projection. For the latter two projections, "30" refers to the angle in degrees between the projection and the AP and PA directions, respectively.

View this table: [in this window] [in a new window]   Table 1. Exposure conditions: X-ray tube potential, X-ray field size at the patient's skin and projection (see Figure 2) are given. The focus-to-skin distance is 70 cm at the centre of the X-ray beam. A 4 mm aluminium total filtration is used. All beams are directed normally to the longitudinal axis of the patient

The back of the patient was at the groin level (z = 0 cm) of the radiologist. A rectangular lead apron covered the front and sides of the radiologist but left the arms unprotected. The apron approximately extended to the knees of the radiologist. The apron was modelled as a 0.35 mm thick lead layer. At the neck level, the apron had a tilted rectangular plate (34x13x0.35 cm³) which extended above the thyroid.

In addition to the exposure conditions of Table 1 the variability of the calculated doses was examined by varying the focus-to-skin distance (40 and 70 cm), the field size (7x7 and 14x14 cm²), the X-ray tube filtration (2.5 and 4 mm of Al) and the position of the radiologist in the x-direction. Different positions of the personal dosimeter were investigated (left and right breast). In addition, personal dose equivalents and effective dose were calculated when a curved lead apron was used, i.e. the frontal part of the apron was modelled as an elliptical cylinder. For the curved apron, two different thicknesses were considered, 0.35 mm and 0.5 mm. These variations were based on exposure conditions 1 and 3 in Table 1. The same thyroid shield (the tilted lead plate) was present in all these calculations.

Dose calculations
The model of the radiologist was equipped with a personal dosimeter which was located on the breast above the lead apron. Instead of modelling the actual device, we calculated the absorbed dose (in kerma approximation, see below) in the centre of a 2 cm diameter soft tissue ball placed at the personal dosimeter's position. The dose (tissue kerma) was calculated from the photon fluence using the NIST data [13] for mass energy absorption coefficients. This dose was used to approximate the personal dose equivalent that would be measured above the apron; we shall refer to this dose as the external dosimeter reading in this paper and denote it by H_ext. The same position was also used when the apron was absent. In addition, the personal dose equivalent was calculated in phantom soft tissue at 10 mm below the skin surface at the same position. This is referred to here as the H_p(10) value. Breast was chosen for the location of this dose calculation point in order to ascertain that there is only soft tissue close to it. The radiologist's organ equivalent doses were calculated for the organs listed by ICRP [15], except for extrathoracic airways which were, in the MIRD-type phantom, included in muscle tissue.

For the sake of computational efficiency the kerma approximation was used, i.e. electrons (or other charged particles) were not transported. A test calculation did not show any significant difference in organ equivalent doses if electron transport was taken into account. Kim and Kim [16] concluded that the kerma approximation is acceptable at photon energies below 200 keV at depths greater than 0.07 mm. The approximation is valid up to 3 MeV in photon energy for depths greater than 10 mm. The approximation may not be valid near the bone surface. However, we did not separately consider bone surface tissue in our model. The dose to the bone surface was approximated by the average dose in the whole skeleton.

Our phantom model did not define the red bone marrow (RBM) as a separate tissue. The absorbed dose to RBM in a specific bone group was calculated from the dose to that bone group, multiplied by the mass energy absorption coefficient ratio of RBM to bone material (the RBM composition was assumed to be equal to that of the soft tissue). Finally, the average over the skeletal system was taken, weighted with the fraction of RBM in each bone group [11]. The kerma approximation may underestimate the dose in RBM by a few percent, because RBM is located in small cavities of bone [17].

The dose to the gonads was calculated as the average of the dose to testicles and ovaries. The colon was divided into upper and lower parts. The dose to the colon was obtained as a weighted sum of the doses in the upper and lower parts of the colon. The weighting of the dose to the upper part was 0.57 and to the lower part 0.43 [15]. In calculating the effective dose the tissue weighting factor for the remaining organs, 0.05, was applied for the mass-averaged dose to these organs (in all cases studied in this work, none of the specified remainder organs received a higher dose than any organ for which a weighting factor is specified).

In the Monte Carlo model the phantom representing the radiologist and the model of the lead apron is not optimal. Especially, the calculation of the dose to the thyroid is not realistic. One of the reasons is that the thyroid position is not realistic in the MIRD-type phantom [18]. Also, in the model the lead apron would partially shield the thyroid even when the separate thyroid shielding part (the tilted lead plate) was absent (see Figure 1). Therefore, in order to study the situation where the radiologist does not use a thyroid shield, we estimated the effective dose by using organ doses as calculated for the lead apron case but replaced the thyroid dose with data from calculations without the lead apron. We also note that the geometry of the simulation may overestimate the shielding effect of the apron for the dose in the brain (see Figures 1 and 3) because the apron protrudes between the irradiated area of the patient and the head of the radiologist. However, the contribution to the effective dose from the brain is small. Furthermore, we did not use the neck location for the external dosimeter because of the unrealistic shape of the MIRD-type phantom at the shoulder region.

View larger version (13K): [in this window] [in a new window]   Figure 3. Organ doses without the apron divided by the dose with the apron and thyroid shield for stomach, brain, testicles and skin for different exposure conditions(Table 1). Open circles are EN/EAT from Table 4. Lines are drawn only to guide the eye.

	Results and discussion

Top Abstract Introduction Monte Carlo simulations Results and discussion Conclusions References

 
Dose to the radiologist
The calculated personal dose equivalents and effective doses of the radiologist, all normalized to the DAP of the X-ray beam, are given in Tables 2 and 3. Table 2 shows the effective doses for all three cases studied: without the apron, with the apron but without the thyroid shield, and with both the apron and the thyroid shield. Without the apron, the overcouch X-ray tube positions (AP, RAO30) and right lateral (RLAT) position of the cardiological procedures yield the highest effective doses. In these projections more scattered radiation enters the middle and upper portions of the trunk where most of the radiosensitive organs are located, although simultaneously the shielding by the patient reduces the dose to the testicles in overcouch tube positions (Figure 3). When the X-ray tube is in the undercouch position, the testicles receive the highest dose. However, the doses to the stomach, liver and lungs are reduced compared with the overcouch tube position. In the left lateral projection (LLAT) the absorption in the trunk of the patient filters the X-ray beam, resulting in a harder energy spectrum in the transmitted radiation. At the same time, the intensity of the transmitted radiation is lower than in other projections and the effective doses of the radiologist are relatively low (Table 2). Effective doses in the cerebral procedures are lower than in corresponding cardiac procedures. This is mainly due to the larger distance of the radiologist from the scattering region and a larger amount of bone material in the X-ray beam, leading to a stronger absorption in the patient.

Personal dose equivalents
The results in Table 3 show that, without the apron, the personal dose equivalent H_p(10) is equal to the external dosimeter reading H_ext within the stated uncertainties. Therefore, we can conclude that our model for the dosimeter depicts H_p(10) realistically in the no-apron case. When the 0.35 mm thick apron is introduced, a slight reduction, typically of the order of 10–15%, is seen in H_ext. This results from a reduction in the backscattered radiation from the body of the radiologist. The lead apron itself does not produce much backscattering at the energies used in this work and the L X-rays from lead are notably attenuated before reaching the 10 mm depth in the dosimeter. With the apron the H_p(10) value is strongly reduced, as expected, on average by a factor of 340.

Figure 4 shows that the dosimeter position has a strong influence on both H_p(10) and H_ext values. Values obtained when the dosimeter is worn on the breast closest to the X-ray beam (i.e. left breast) are roughly two times higher than when the dosimeter is on the other (right) breast. This is mainly due to the distance effect: the scattered radiation field is more intense closer to the scattering region.

View larger version (11K): [in this window] [in a new window]   Figure 4. Ratio of personal dose equivalentHp(10) calculated in the right breast to that calculated in the left breast (without the apron) for the cardiac procedures. Exposure conditions are indicated below each data point and on the horizontal axis (Table 1). One sigma error bars represent the statistical uncertainty of the ratio. For abbreviations, see Figure 1.

The results in Table 4 show that, on average, H_ext measured on the left breast overestimates the effective dose by a factor of 130 when the apron and thyroid shield are used. The variation in the ratio H_ext/E_AT is, however, large, and ranges from 44 to 258 in the conditions considered in this paper. Without the thyroid shield (H_ext/E_A) the overestimation is smaller, a factor of 69 (range from 32 to 127).

Variability of the simulated doses
Table 5 shows that none of the doses (H_p(10), H_ext or effective dose, either with or without the apron), normalized to the DAP, changes notably when the focus-to-skin distance is varied. When filtration is reduced, all these doses decrease slightly, as can be expected. When the field size is reduced these doses are reduced notably. This is likely to result from the X-ray beam being more tightly limited to the phantom's spine area, which is more absorbing than the nearby areas. However, none of these changes (focus-to-skin distance, field size or filtration) has a notable effect on the ratio E_N/E_AT or H_p(10)/E. On the other hand, H_ext/E decreases somewhat during these changes in parameters. (E_N is the effective dose to the radiologist without the apron. E is used as a generic symbol to represent the effective dose and subscripts are used for indicating the degree of shield usage.) We note that recently McVey [20] investigated the effect of phantom type, beam filtration, X-ray tube voltage, voltage ripple, field size and field position on the X-ray scattering from a patient phantom. Large variations in the amount of scatter were found especially with changing phantom type, tube voltage and field size.

View this table: [in this window] [in a new window]   Table 5. Variation of the doses with respect to variations in simulation parameters. Exposure conditions(see Table 1) on which the variation is based are given in parentheses. The thyroid shield was used in calculations with the apron. All doses are normalized to the DAP value of the X-ray beam. Standard deviations (statistical uncertainty, in %) are given in parentheses

As already mentioned above, we also studied the effect of changing the model of the apron, in order to see if the details of the apron are important in the simulation. When the frontal part of the apron is shaped as an elliptic half cylinder, higher doses (in AP projection, exposure condition 3) to internal organs are obtained than with the rectangular apron, causing a 13% increase in effective dose. This is due to the smaller effective thickness of the elliptically shaped apron in the direction of impinging photons. The effect on the effective dose is moderate, because E is, also in this case, dominated by the unprotected parts of the body. Of the dose quantities considered, the change in the apron shape affects most notably the H_p(10), which increases by a factor of two. In practice, the movements of the radiologist make the situation more ambiguous and the resulting effective thickness of the apron is difficult to estimate.

A 0.5 mm thick apron is often used. Such an apron with the elliptically shaped frontal part decreases E_AT by 31% compared with the 0.35 mm thick elliptical apron. The H_p(10) value is reduced by 74%. Compared with the rectangular 0.35 mm thick apron the reduction in E_AT is 23% and in H_p(10) 44%.

Uncertainty estimates
Statistical uncertainties in simulated effective doses are less than 5% (one sigma uncertainty, Table 2). The statistical uncertainties in H_p(10) and H_ext are less than 10%. Uncertainties related to the physical modelling in a Monte Carlo calculation are more difficult to estimate. Potential error sources include the cross-sections and various approximations in radiation transport, e.g. using the kerma approximation, and not including coherent scattering and Bremsstrahlung in the physical interaction processes. Our estimate for this type of uncertainty in E, H_ext and H_p(10), based on the test calculation with coherent scattering and electron transport included, is 15%. We assume that all uncertainties related to physical modelling can be considered as normally distributed.

Our treatment of the energy deposition in the red bone marrow may entail a few percentage error because of the bone marrow cavity effect. However, the resulting error in the effective dose of the radiologist is smaller. The uncertainties related to the phantom geometry, e.g. wrongly positioned organs, are partly cancelled when the ratios of effective doses are calculated, but affect the conversion factors, such as E/DAP or H_ext/E. Ferrari and Gualdrini [21] and Kramer et al [22] have examined the organ doses and effective doses with several voxel and MIRD-type phantoms in photon fields. When the phantoms were irradiated in the AP geometry the difference in E between the male MIRD-type phantoms and the voxel phantoms was less than 10% for the energy range considered here (30–100 keV), although the differences in the doses of specific organs varied significantly more.

The variation of the doses related to the apron type must be accounted for. The calculated 13% difference in E_AT between rectangular and elliptical aprons (with a thickness of 0.35 mm) is considered as an upper limit of this variation (see the discussion above). Therefore, we can conclude that the total uncertainty in conversion coefficients arising from the mathematical modelling of the problem is of the order of 30%. In particular, this estimate applies to the conversion coefficients H_ext/E.

When the distance of the radiologist from the beam centre is reduced by moving the patient 20 cm to the negative x-direction, H_ext/E is reduced by nearly 50% in both the apron and no-apron cases (Tables 4 and 5). Thus, the largest variation found here originates from the position and movements of the radiologist and also, partly, from the position of the dosimeter relative to the irradiated patient volume. The stated 50% variation is a rough estimate since it strongly depends on the personal practices of the radiologist. When this is combined with 30% uncertainty related to the mathematical modelling, a total uncertainty of 60% is obtained for the dose ratios in a given projection and X-ray tube voltage setting (we assume that all uncertainties and variations are normally distributed). This estimate may be used for the conversion coefficient E/DAP. The uncertainty estimate for the dose ratios does not explicitly include the variation of H_ext or H_p(10) with respect to dosimeter position (see Figure 4).

Comparison with earlier studies
In their work, Faulkner and Marshall [23] studied a range of exposure conditions by using thermoluminescent dosimeter (TLD) measurements in an anthropomorphic phantom and experimental broad beam transmission data in lead for scattered radiation. They concluded that a dosimeter, worn on the 0.35 mm thick lead apron, may overestimate the effective dose by a factor of roughly 25 to 60, depending on the irradiation conditions. In addition, they state that an accurate estimate for the effective dose is impossible to obtain with a single dosimeter. The present study yielded more pronounced overestimation (H_ext to E_AT ratios), from 44 to 258 (Table 4) when a thyroid shield was used. Without the thyroid shield (H_ext to E_A) the range was from 32 to 127.

Niklason et al [24] concluded that the dosimeter reading on the apron should be divided by 25 to obtain the effective dose. The dosimeter was placed at the neck so that a direct comparison with the present results may not be accurate. Kicken et al [25] used Monte Carlo-derived air kerma-to-organ dose coefficients along with a number of correction factors to estimate the effective dose in interventional arteriography, and stressed the role of the thyroid shield and the unprotected extremities. A correction factor was used to account for the protection offered by the 0.5 mm thick lead apron. Kicken et al suggested that a dosimeter worn at the neck can be used to estimate the effective dose. In this case the dosimeter reading should be divided by a factor of 24–45 if a thyroid shield is used and by a factor of 12–15 if the thyroid shield is not used. Although the location of our external over-the-apron dosimeter is slightly different from the neck location used by Kicken et al [25], it seems that their estimate of the effective dose of the radiologist is higher than ours by a notable factor (see Table 4).

Kicken et al [25] also present conversion coefficients from DAP to the effective dose of the radiologist, and suggest values of 0.1–0.2 µSv Gy^–1 cm^–2 for the PA projection and 0.8 µSv Gy^–1 cm^–2 for the AP projection in the case when a thyroid shield is not used. Also these figures are higher than our results, but a comparison is not straightforward because their data are not given for a specified distance from the radiation field. Anyway, it seems that their method of calculating the effective dose of the radiologist results in higher dose values than the direct Monte Carlo method used here.

The Monte Carlo simulations by Schultz et al [6] yielded a wide range for the ratio E_N/E_A (Schultz et al call this the lead apron protection factor). The frontal apron model used by Schultz et al leaving the arms and sides unprotected, gave lead apron protection factors of 16.5 (0.25 mm of Pb) or 20.7 (0.5 mm of Pb). Their wrap-around apron model yielded lead apron protection factors 41.3 and 62.0 for 0.25 mm and 0.5 mm of Pb, respectively. The X-ray tube potential was 59 kV and the field size was 5.2x6.6 cm². This is in reasonable agreement with the value E_N/E_AT = 39 and E_N/E_A = 22 for the exposure condition 7 in the present work, taking into account the differences in the apron model and irradiation conditions.

The two-dosimeter model of Niklason et al [1] (equation 5, op. cit.) uses collar and waist dosimeters in estimating the effective dose. Even though the positions of the dosimeters are different in Niklason et al and the present work, it is interesting to see how well the two-dosimeter formula predicts the effective dose in this different irradiation geometry. When H_ext is used instead of the over-apron collar dose and H_p(10) instead of the under-apron waist dose in the formula of Niklason et al the resulting effective dose overestimates E_AT in our Table 2 by a factor of 1.3 to 5.6, depending on the exposure conditions. Without the thyroid shield the overestimation ranges from 2.1 to 7.7. Again, if H_ext had been measured on the collar instead of the breast, the overestimation would be slightly smaller.

	Conclusions

Top Abstract Introduction Monte Carlo simulations Results and discussion Conclusions References

 
Monte Carlo simulations are a practical tool for estimating occupational effective doses in interventional radiology. A caveat in using simulations is the potential sensitivity of the results to small changes in the modelled irradiation conditions. Therefore, an analysis of the uncertainty budget is indispensable. Measurements can be done in some special cases in the laboratory environment, e.g. to obtain the reduction in H_p(10) when the lead apron is used. However, measurements of the effective dose in real exposure conditions are practically unachievable.

The results from the Monte Carlo simulations reveal that the effective dose to the radiologist and the protection provided by the lead apron vary significantly with changing irradiation conditions. Moreover, the location of the personal dosimeter has a major impact on the simulated, as well as the measured, H_ext value. The thyroid shield often reduces the effective dose by a factor of two.

If an accurate estimation of the effective dose is needed, the exposure conditions should be known accurately. In particular, knowledge of the location of the radiologist, the dosimeter position, the X-ray tube voltage and the projection are essential. In reality, dosimeter readings are often sums over a long time period during which many types of procedures are carried out. Then, an average value of the conversion coefficient from H_ext to E needs to be used – while understanding that the accuracy of the conversion is modest. Evaluating the effective dose of the radiologist from the DAP data may be useful as well, but the results are much more uncertain, because the distance of the radiologist from the X-ray beam will be a major factor influencing the evaluation.

In the present work, special emphasis is placed on the uncertainty estimate of the simulated results. This estimate includes both the statistical uncertainty stemming from the Monte Carlo calculation itself and the variation of the results with respect to changes in the irradiation geometry. The results indicate that moderate changes in the irradiation geometry do not have a major impact on the lead apron protection factor, E_N/E_A. However, the ratio H_ext/E is more sensitive to changes in geometry.

The exposure conditions in this and earlier studies are not exactly comparable. Therefore, a direct comparison between the results may not be reasonable. Our results indicate, however, that the often used conversion coefficients from H_ext to E_A, around 1/30 (see also [2]), somewhat overestimates the effective dose. Since the conversion coefficient varies greatly from one exposure condition to another, a generally applicable accurate relationship between the dosimeter reading(s) and the effective dose of the radiologist cannot be established based on the present data. Such a relationship would require detailed information on the actual exposure conditions, combined with similar data as presented here. However, as a rough estimate, a conversion coefficient H_ext to E_A around 1/60 and H_ext to E_AT around 1/120 may be used while remembering that an underestimation of the effective dose may occur with some exposure conditions.

	Acknowledgments

 
We thank Dr M. Kortesniemi (Helsinki University Central Hospital) for providing information on the examination parameters.

Received for publication April 7, 2006. Revision received August 18, 2006. Accepted for publication September 19, 2006.

	References

Top Abstract Introduction Monte Carlo simulations Results and discussion Conclusions References

 

1. Niklason LT, Marx MV, Chan HP. The estimation of occupational effective dose in diagnostic radiology with two dosimeters. Health Phys 1994;67:611–15.[Medline]
2. Schultz FW, Zoetelief J. Estimating effective dose for a cardiac catheterisation procedure with single or double personal dosemeters. Radiat Prot Dosim 2006. 118:196–204.
3. Tsapaki V, Kottou S, Patsilinakos S, Voudris V, Cokkinos DV. Radiation dose measurements to the interventional cardiologist using an electronic personal dosemeter. Radiat Prot Dosim 2004;112:245–9.[Abstract/Free Full Text]
4. Vañó E, González L, Guibelalde E, Fernández JM. Radiation exposure to medical staff in interventional and cardiac radiology. Br J Radiol 1998;71:954–60.[Abstract]
5. Williams JR. The interdependence of staff and patient doses in interventional radiology. Br J Radiol 1997;70:498–503.[Abstract]
6. Schultz FW, Geleijns J, Spoelstra FM, Zoetelief J. Monte Carlo calculations for assessment of radiation dose to patients with congenital heart defects and to staff during cardiac catheterizations. Br J Radiol 2003;76:638–47.[Abstract/Free Full Text]
7. Pelowitz DB. MCNPX^TM user's manual, version 2.5.0. Los Alamos National Laboratory, Los Alamos, NM, USA, 2005
8. White MC. Photoatomic data library MCPLIB04: a new photoatomic library based on data from ENDF/B-VI release 8. Los Alamos National Laboratory internal memorandum X-5:MCW-02-111, New Mexico, USA, 2002
9. Birch R, Marshall M. Computation of Bremsstrahlung X-ray spectra and comparison with spectra measured with a Ge(Li) detector. Phys Med Biol 1979;24:505–17.[CrossRef][Medline]
10. Cristy M, Eckerman KF. Specific absorbed fractions of energy at various ages from internal photon sources. I. Methods. Oak Ridge National Laboratory Report ORNL/TM-8381/VI, USA, 1987
11. Eckerman KF, Cristy M, Ryman JC. The ORNL mathematical phantom series. Oak Ridge National Laboratory Report, 1996. Available at http://homer.hsr.ornl.gov/VLab/VLabPhan.html
12. BodyBuilder for Windows, version 1.30. White Rock Science, 2003
13. National Institute of Standards and Technology (NIST) database. Available at http://physics.nist.gov/PhysRefData/XrayMassCoef/cover.html
14. International Commission on Radiation Units and Measurements. Tissue substitutes in radiation dosimetry and measurement. ICRU Report 44, Bethesda, MD, USA: International Commission on Radiation Units and Measurements, 1989
15. The International Commission on Radiological Protection. Age-dependent doses to members of the public from intake of radionuclides: Part 4, Inhalation dose coefficients. ICRP Publication 71. Oxford, UK: Pergamon Press, 1995
16. Kim JO, Kim JK. Dose equivalent per unit fluence near the surface of the ICRU phantom by including the secondary electron transport for photons. Radiat Prot Dosim 1999;83:211–19.[Abstract]
17. Kerr GD, Eckerman KF. Neutron and photon fluence-to-dose conversion factors for active marrow of the skeleton. In: Schraube H, Burger G, Booz J, editors. Proceedings of the fifth symposium on neutron dosimetry, Volume I, Radiation protection aspects, EUR 9762 EN, CEC, Luxembourg, 1985
18. Lee C, Lee J, Lee C. The effect of unrealistic thyroid vertical position on thyroid dose in the MIRD phantom. Med Phys 2004;31:2038–41.[CrossRef][Medline]
19. The International Commission on Radiological Protection. Conversion coefficients for use in radiological protection. ICRP Publication 74. Oxford, UK: Pergamon Press, 1996
20. McVey G. The effect of phantom type, beam quality, field size and field position on X-ray scattering simulated using Monte Carlo techniques. Br J Radiol 2006;79:130–41.[Abstract/Free Full Text]
21. Ferrari P, Gualdrini G. An improved MCNP version of the NORMAN voxel phantom for dosimetry studies. Phys Med Biol 2005;50:4299–316.[CrossRef][Medline]
22. Kramer R, Khoury HJ, Vieira JW. Comparison between effective doses for voxel-based and stylized exposure models from photon and electron irradiation. Phys Med Biol 2005;50:5105–26.[CrossRef][Medline]
23. Faulkner K, Marshall NW. The relationship of effective dose to personnel and monitor reading for simulated fluoroscopic irradiation conditions. Health Phys 1993;65:502–8.
24. Niklason LT, Marx MV, Chan HP. Interventional radiologists: occupational radiation doses and risks. Radiology 1993;187:729–33.[Abstract/Free Full Text]
25. Kicken PJH, Kemerink GJ, Schultz FW, Zoetelief J, Broerse JJ, van Engelshoven JMA. Dosimetry of occupationally exposed persons in diagnostic and interventional arteriography. Part II: Assessment of effective dose. Radiat Prot Dosim 1999;82:105–14.[Abstract]

This article has been cited by other articles:

				O. O. Lie, G. U. Paulsen, and T. Wohni ASSESSMENT OF EFFECTIVE DOSE AND DOSE TO THE LENS OF THE EYE FOR THE INTERVENTIONAL CARDIOLOGIST Radiat Prot Dosimetry, December 4, 2008; (2008) ncn296v1. [Abstract] [Full Text] [PDF]

				H. Jarvinen, N. Buls, P. Clerinx, S. Miljanic, D. Nikodemova, M. Ranogajec-Komor, L. Struelens, and F. d'Errico COMPARISON OF DOUBLE DOSIMETRY ALGORITHMS FOR ESTIMATING THE EFFECTIVE DOSE IN OCCUPATIONAL DOSIMETRY OF INTERVENTIONAL RADIOLOGY STAFF Radiat Prot Dosimetry, August 30, 2008; (2008) ncn239v1. [Abstract] [Full Text] [PDF]

				H. Jarvinen, N. Buls, P. Clerinx, J. Jansen, S. Miljanic, D. Nikodemova, M. Ranogajec-Komor, and F. d'Errico OVERVIEW OF DOUBLE DOSIMETRY PROCEDURES FOR THE DETERMINATION OF THE EFFECTIVE DOSE TO THE INTERVENTIONAL RADIOLOGY STAFF Radiat Prot Dosimetry, May 14, 2008; (2008) ncn082v1. [Abstract] [Full Text] [PDF]

				T. Siiskonen, M. Tapiovaara, A. Kosunen, M. Lehtinen, and E. Vartiainen Occupational radiation doses in interventional radiology: simulations Radiat Prot Dosimetry, March 1, 2008; 129(1-3): 36 - 38. [Abstract] [Full Text] [PDF]

This Article Abstract Figures Only Full Text (PDF) Services Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Citing Articles Citing Articles via HighWire Citing Articles via Google Scholar Google Scholar Articles by Siiskonen, T Articles by Vartiainen, E Search for Related Content PubMed PubMed Citation Articles by Siiskonen, T Articles by Vartiainen, E