Abstract- This study aimed to quantify effective dose (E) for occupationally exposed persons (workers) in ... exposure parameters: dose-area product and the.
Radiation Protection Dosimetry Vol. 82, No. 2, pp. I05- 114 (1999) Nuclear Technology Publishing
DOSIMETRY OF OCCUPATIONALLY EXPOSED PERSONS IN DIAGNOSTIC AND INTERVENTIONAL ARTERIOGRAPHY. PART II: ASSESSMENT OF EFFECTIVE DOSE P. J. H. Kickent*, G. J. Kemerink t, F. W. Schultz+, J. Zoetelief:j:, J. J. Broerse:j:§ and J. M. A. van Engelshovent t Department of Radiology, University Hospital Maastricht P. Debijelaan 25, 6229 HX Maastricht, The Netherlands :j:TNO Centre for Radiological Protection and Dosimetry Lange Kleiweg 151, 2288 GJ Rijswijk, The Netherlands §Department of Clinical Oncology, Radiotherapy Section University Hospital Leiden Albinusdreef 2, 2333 ZA Leiden, The Netherlands
Received October 16 1998, amended December 11 1998, accepted December 14 1998 Abstract- This study aimed to quantify effective dose (E) for occupationally exposed persons (workers) in diagnostic and
interventional arteriography. Using Monte Carlo radiation transport calculations, new air kerma to organ dose conversion coefficients (DCCs) were determined. Taking attenuation by protective clothing into account, E was estimated from the DCCs and entrance doses measured at forehead , neck, thorax, abdomen, upper arms, hands and lower legs of the workers. Average E was calculated for several types of procedures and for three hospital s. For operators, E was between 0.5 and 7 µSv per procedure for undercouch X ray tubes and between 0.8 and 22 µSv for overcouch tubes. For assistants, E was typically a factor of two lower. Wearing a thyroid collar roughly reduced E by 50%. These values are low , which is due, among other things, to the nearly general use of .a mechanical contrast injector and an undercouch X ray tube.
thighs, arms (2 X), hands (2 X) and lower legs (2 X). Thorax, adomen and thighs were always protected by a lead apron. A thyroid collar was worn sometimes, protective eye glasses very infrequently; the other regions were unprotected. Note that TL dosemeters on the head, neck, thorax and abdomen were placed at the middle front, upon protective clothing if present. The major contribution to the radiation dose was incurred from the lateral-frontal side. Unfortunately, no suitable Monte Carlo dose conversion coefficients (DCCs) were available for such oblique projections. Therefore it was decided to use DCCs for anterioposterior (AP) projections and to apply corrections for the deviating projection direction. For head, neck, thorax and abdomen some DCCs were available, for instance from GSF(J J, generally for projections with a limited field of view, as used in diagnostic radiography, and not covering the full body parts that are of interest here. For the extremities no data had been publi shed at all. Therefore new Monte Carlo radiation transport calcu lations were performed for all regions, using a broad unidirectional AP beam of 80 kV X rays with 2.5 mm Al filtration (first half-value layer 2.8 mm Al). This X ray tube voltage was approximately the average found during fluoroscopy in all three hospitals, and fluoroscopy was the imaging mode responsible for exposure of workers. The simulations were performed at TNOCSD using the anthropomorphic phantoms ADAM and EVA . The resulting DCCs were subsequently aver-
INTRODUCTION In Part I exposure conditions and entrance doses were reported for occupationally exposed persons (workers) in diagnostic and interventional endovascular procedures. These persons had been monitored in three different hospitals (AZM, DW, MA). This paper (Part II) presents new Monte Carlo radiation transport code calculations, effective dose estimates, and relations between effective dose and two easily accessible exposure parameters: dose-area product and the entrance dose at the neck. Remarkably, hardly any estimate of effective dose for the workers considered here could be found in the literature. It appeared that all investigators had exclusively focused on quantifying entrance doses (see Table 7, Part I) . MATERIALS AND METHODS
Dose conversion coefficients As described in Part I, our dosimetric approach distinguished 11 regions on the body of occupationally exposed persons: head, neck, thorax, abdomen, both *Present address: University Hospital Rotterdam , Daniel den Hoed Cancer Center, Groene Hilledijk 301 , 3075 EA Rotterdam , The Netherlands 105
P. J. H. KICKEN, G. J. KEMERINK, F. W. SCHULTZ, J. ZOETELIEF, J. J. BROERS£ and J. M. A. van ENGELSHOVEN
aged over both sexes, and compared with (extrapolated) GSF data if possible. The agreement was generally satisfactory, and only minor changes to the TNO data were introduced: in the thorax region a few TNO and GSF data were averaged, while in the abdominal region a few data were replaced by slightly more conservative GSF data. Using tissue weighting factors given in ICRP 60, an air kerma free-in-air to effective dose conversion coefficient (ECC) for each region was also determined. There was no consensus in the literature on the protection provided by a thyroid collar. To obtain independent information, additional Monte Carlo calculations have been performed for the exposure of a worker wearing, in addition to a normal apron, a thyroid collar of 0.35 mm or 0.5 mm Pb-equivalent thickness. The upper edge of the collar just extended to the top of the thyroid of ADAM and EVA. To reduce computation time a partial irradiation of head, neck and upper trunk was applied. Calculation of effective dose
Estimates of effective dose to workers (E as defined in ICRP 60) were calculated according to the following expression: E
- ~ -
~region
K
~
reg ioff~T
wTDCCregion.TCorrregion,T Att .
(1)
reg1on.T
where wT is the tissue weighting factor for a specified or remainder organ T according to ICRP 60. For remainder organ j, the tissue weighting factor W; was calculated using mass weighting, i.e. w; = 0.05 m/lm; , with m; the mass of remainder organ i. The weighting factor wT was calculated for men and women separately and then averaged. DCCregion.T is the dose conversion coefficient for organ T in body part 'region', with ' region ' any of the l l body parts described in the previous section. In the Monte Carlo data used in this study DCCs were calculated as DCC = Dorga/Ka1r. free in air. at entrance posi1ion (Gy/Gy). Krcgion is the air kerma, free-in-air, at the skin entrance position of body part 'region'. In Part I it was stated that the TL dosemeters were calibrated in air kerma. Since the dosemeters were worn on the body, a contribution of backscattered radiation is included in the TLD readings (Kair(TLoi>· However, many Monte Carlo dose conversion coefficients (DCCs), including the set calculated in this study, give the relation between organ absorbed dose and air kerma free-in-air at skin position. Application of such DCCs requires, therefore, first correction of the TLD dose data for backscatter. This correction can be obtained from the following consideration. An air kerma free-in-air of 1 Gy at the entrance position corresponds to a tissue absorbed dose of Z = (µe/p),;ssuJCµen/p).;, Gy, with (µ e/P) the mass attenuation coefficient. In this conversion bremsstrahlung has 106
been neglected. Since the DCCskin obtained in Monte Carlo calculations is the ski n dose per unit air kerma free-in-air, and because this occskin includes backscatter, the desired correction factor for backscatter is given by: DCCskin/Z. This correction was applied for the head, neck. upper arms, hands and lower legs, according to Kregion = Kair(TLo/(DCCsk;/Z). It was assumed that DCCskin is constant for a given region, and that Z = 1.06 for diagnostic X rays. Values for DCCskin were taken from GSF tables if available, otherwise no correction was applied. Backscatter corrections for the TLD readings on the lead apron were considered to be negligible. Our measurements on copper had shown that the backscatter contribution was less than 3%, and according to Paix et a[< 5 > lead behaves similarly to copper in this respect. Corrregion.T is a correction factor used for a few different purposes. It was set to unity, unless stated otherwise below: (a)
Exposure of workers predominantly occurs from the lateral-frontal side. This has two effects: first, the mid-frontally measured entrance dose underestimates the maximum val ue, and second, DCCs derived for AP projections are too high. According to measurements presented in Part I, the ratio between the entrance dose at the side and that at the front of the trunk is between 1.5 and 2.5 . An indication for the change in DCCs with angle, as reflected in the effective dose (E), was obtained from data published by Yamaguchi and Zankl et a/Ol: EAp/E45 ooblique = 1.3- 1.5. Combining the two opposing effects, Corrregion.T was taken to be 1.5 for all organs in the thorax and abdomen regions. (b) In the hospital MA the entrance dose at the side of the head was about a factor of 4 higher than that at the forehead (Part I) . For a given entrance dose, the contribution to the effective dose by exposure of the head hardly depends on the projection direction. In the analysis of data from the MA hospital, Corrregion.T was therefore chosen as 4 for the head region. (c) Finally, lead aprons genei::?lly have neck holes that leave a small part of the upper thorax unprotected. As the area of the neck hole is about the same as that of the neck, this was accounted for by setting Corrregion,T to 2 for the organs red marrow, oesophagus, skin, bone surface and muscle in the neck region. Since the thyroid does not extend into the neck hole region, Corrneck.thyroid was set to unity. Attregion,T is the effective attenuation factor of the lead apron, applicable to the regions thorax, abdomen and thighs. In a previous study the attenuation of 0.5 mm Pb-equivalent aprons under clinical conditions was found to be approximately 200. However, the effective photon energy also increases through the spectral selective filtering by lead, causing DCCs for the transmitted
OCCUPATIONAL DOSIMETRY IN ARTERIOGRAPHY: EFFECTIVE DOSE
photons to be higher than for the incident radiation. Using data by Seuntjes et af
