| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
Full Paper |
dilek, MD2
1 Ankara University Faculty of Engineering, Department of Engineering Physics, 06 100 Tando
an, Ankara and 2 Ankara University Faculty of Medicine, 
bni Sina Hospital, Department of Interventional Radiology, Shhiye, Hasrclar Str. Ankara, Turkey
 |
Abstract |
|---|
 Top Abstract IntroductionMaterials and methodsResultsDiscussionConclusionReferences |
|---|
|
Introduction |
|---|
|
TopAbstractIntroductionMaterials and methodsResultsDiscussionConclusionReferences |
|---|
Measurement of dose–area product (DAP) or entrance skin dose are the common methods for patient dosimetry and effective doses are then estimated by multiplication of the measured quantity by conversion factors that have been calculated by Monte Carlo techniques. These factors have been determined for specific X-ray projections. However, uncertainties in the effective dose estimates may be introduced because the standard conditions of exposure assumed when deriving the conversion factors will never truly be met fully in practice. Apart from differences in anatomy between the mathematical phantom and patients (in terms of organ sizes and positions), there will also be differences in the X-ray field size and position; this affects the coverage of organs by the X-ray beam and hence their doses and the subsequent effective doses [2, 3]. Entrance dose measurements either with thermoluminescent dosemeters (TLDs) attached to the patient skin or free in the air in the absence of the patient are the pertinent quantities when deterministic effects are concerned. It should be emphasised that DAP measurements cannot provide information about the most irradiated anatomical areas, which carry the highest risk of injuries. However, it is stated that DAP results are more reliable than entrance dose measurements for effective dose estimation, especially for interventional procedures where the irradiation of the patient and direction of the X-ray beam changes continuously [4].
It would therefore be desirable to make simultaneous DAP, air kerma and entrance dose measurements for obtaining information from which deterministic and stochastic effects may be estimated. Calculation of DAP from air kerma measurement or air kerma estimation from DAP measurements are possible. However uncertainties up to 40% have been reported due to the deviations from nominal values [5].
The Diamentor M4KDK (PTW Freiburg, Germany) is a device for the simultaneous measurement of DAP and air kerma in diagnostic radiology. The DAP is measured with a flat ionization chamber mounted directly on the light beam diaphragm housing. This flat chamber is almost transparent to X-rays and a small section (1.7 cm x 1.7 cm) in the centre of this chamber is designed to be used as a second chamber for air kerma measurements [6]. However, this construction causes field size dependency on DAP and air kerma measurements. There is incomplete charge collection created by the 2 mm gap between the two chambers. This becomes important in the case of DAP measurements for small field sizes, but the effect is negligible for fields larger than 200 cm2. Dependency on field size for air kerma measurements also occurs if the exposed area on the patient surface is less than the projected size of the small chamber.
In this study, patient dose measurements for some interventional examinations using M4KDK and TLDs were carried out. In addition to effective dose calculations from DAP and air kerma measurements, surface doses obtained from air kerma and TLD measurements were also compared. Phantom experiments were carried out to investigate the variation of air kerma and DAP with exposure conditions.
Careful recording of exposure parameters for each projection was accomplished during the patient examinations and the most appropriate technical parameters were used for the effective dose calculations in order to reduce discrepancies.
|
Materials and methods |
|---|
|
TopAbstractIntroductionMaterials and methodsResultsDiscussionConclusionReferences |
|---|
Radiological procedures
Angiographic procedures were categorised according to the anatomy imaged [7]. For the effective dose calculations, NRPB data files [3] which tabulate the conversion factors for the estimation of effective dose from DAP and entrance skin dose measurements were used. It is of course not possible to find a standard projection which exactly matches the projection used in clinical applications. There will be deviations from the theoretical models in terms of field size, position and beam quality. However, considering the irradiation of the same area, a theoretical projection which best matched the clinical application was selected in accordance with the data given by McParland [7], Ruiz Cruces [8] and Steele [9].
For some clinical projections there were no similar theoretical views. In these cases the closest projection was used, e.g. posteroanterior (PA) views were used for oblique projections.
Cerebral angiography
The intracranial carotid and vertebral artery circulation was examined by common carotid and dominant vertebral artery injections. The theoretical "head PA", "head LAT" projections [3] were used for PA, right lateral and left lateral clinical views. Head PA was considered the most suitable for the oblique view.
Carotid arteries
For the carotid arteries, the tip of the catheter was placed in the common carotid artery proximal to the bifurcation under fluoroscopic guidance. The theoretical "cervical spine AP" projection was used for the PA view and "cervical spine lateral" projections were used for right and left lateral clinical views. Similarly "cervical spine PA" was used for the oblique view [3].
Hepatic angiography
Most of the examinations were therapeutic chemoembolisations of hepatic masses. The theoretical projection "abdominal PA" was considered the most suitable for this procedure [3].
Thoracic
Only the patients who had a single view diagnostic examination were included in the study. The theoretical projection "chest PA" was considered the most suitable for this procedure [3].
Renal
In addition to PA diagnostic interventions, balloon angioplasties and renal artery stenting were performed. The theoretical projection "kidney PA" was chosen for this procedure [3].
Upper extremity (shoulder)
Radiographic images were obtained in the PA position. Most of the angiograms were intravenous digital subtraction angiography (DSA) procedures. Arterial and venous pathologies were vascular stenosis. Balloon dilatation and stenting were the interventional procedures in the upper and lower extremities. "Shoulder PA" was considered the most appropriate for this procedure [3].
Lower extremity
Standard angiographic examinations used in patient studies were abdominal aorta, iliac and peripheral arteries. We did not make any exposure partition during the data acquisition in our angiography studies. In most of the cases, irradiation started from the abdomen and continued to the feet. However, following the careful inspection of patient studies, the pelvic area was taken for the effective dose calculation. Our assumption was that 50% of the total DAP was attributed to this area. Therefore the theoretical projection "pelvis" was selected for this procedure [3].
Dosimetric systems
Lithium fluoride TLD chips (3.7 x 3.7 x 0.9 mm) (Model 100; Harshaw Chemical, Solon, OH) in plastic handling pockets (3 for each) were used to measure the skin dose. These were attached to a single point on the patient's skin where the exposure was expected to be at its highest level. These locations were determined by the radiologist for each examination.
A model 3500 Reader (Harshaw Chemical) was used for the TLD readout. TLD calibrations were made against an ion chamber (Rad Check Plus; Victoreen, Cleveland, OH) using the same X-ray beam qualities as in the patient studies. In order to minimize batch to batch variability, a pre-selected group of TLDs were calibrated initially and variation of the sensitivities was kept within ±5% (±1 SD). Annual calibration of this ion chamber was performed in a Secondary Standard Dosimetry Laboratory (SSDL).
The DAP and air kerma (or entrance dose) at a certain distance on the central axis were measured using a dual channel dosemeter Diamentor M4KDK. These chambers were calibrated in situ against an ion chamber (Rad Check Plus) according to the procedure proposed by the manufacturer and others [6]. Since air kerma measurements using a M4KDK are obtained at a specific distance, readings were first distance-corrected using the inverse square law and then multiplied by backscatter factors and the mass energy absorption coefficient ratio for tissue to air (1.06), for the estimation of patient entrance dose [5]. Backscatter factors were selected from the tabulated data according to the half value layer (HVL) and tube potential (kVp) used [2]. Focus–skin distances were calculated from table–focus distances for each projection.
An on-line connection of the M4KDK dosemeter to a personal computer was provided using Diasoft Software (PTW Freiburg). Storage of the complete procedure was possible with this software and fluoroscopy time, air kerma and DAP for fluoroscopy and radiography, air kerma rate and DAP rate and number of images (for radiography) can be extracted following the study. This software was designed to distinguish between the fluoroscopy and radiography on the basis of the rate of increase in DAP and duration of radiographs. These user selectable parameters may cause some errors in the reliable identification of radiographic exposures if relatively low doses are used. In order to prevent software related errors, the person responsible for data collection was informed whenever fluoroscopic or radiographic exposures was initiated. Exposures were recorded in separate files which improved the accuracy of the assessment of relative contributions.
Data acquisition
Personal data for each patient (name, gender, weight, height) were registered. Fluoroscopic and radiographic exposures from the M4KDK were stored on separate files on Diasoft for each projection for ease of retrospective interpretation. During the fluoroscopic examinations, positioning of the patient was observed and average values for tube potential were noted for each projection through the observation of the acquisition display. The staff collecting the data were informed whenever the mode of the image intensifier and the collimation of the X-ray beam were changed.
The source-to-image intensifier and table-to-focus distances were continuously indicated on the system for the vertical position of the C-arm. This information was similarly provided by the assistant during the course of the study. These data were then used to calculate the source-to-skin distances for each view. However for the lateral position of the gantry, an average source-to-skin distance was determined from the patient study observations.
The duration of fluoroscopy and number of radiographic frames were derived for each view from the stored data and this latter parameter was also checked from the archive data of the Polytron Processing system. Additionally kVp, mA, pulse duration and dose factors used during the collection of radiographic images were also extracted from the archive data of this system.
Effective doses were calculated for each projection from DAP and air kerma readings using NRPB data files and XDose program [3, 10]. In the case of multiple projections, effective dose was calculated for each projection. TLD results obtained from the complete examination were similarly used for effective dose calculations for comparison purposes. With the exception of some oblique and lateral views separate TLD packets were used for each projection.
Phantom studies
Phantom experiments were carried out in order to see the variations of DAP and air kerma results with some exposure parameters.
A tissue equivalent lucite phantom of thickness 15 cm was exposed at different focus–phantom, image intensifier–focus distances, X-ray beam collimations, and image intensifier magnification factors. Only one parameter at a time was changed whilst the other factors were kept constant. DAP and air kerma values as well as post exposure data (kVp, mA, exposure time) were recorded for retrospective evaluation. During the collimation of the X-ray beam, the beam size at the phantom entrance plane was measured with a scale placed on the table top. Similar experiments were repeated using the Rando Phantom (Alderson Research Laboratories, Stamford, CO) and its abdominal region was used for all the exposures.
Moreover, a radiologist was asked to perform angiographic examinations of kidney, carotid, cerebral, and abdomen on the Rando phantom. Irradiation geometry, fluoroscopy time and number of radiographic frames were similar to patient studies for each procedure. Effective doses were calculated from DAP, air kerma and TLD measurements.
|
Results |
|---|
|
TopAbstractIntroductionMaterials and methodsResultsDiscussionConclusionReferences |
|---|
shows the number of patients for each examination, total DAP, fluoroscopic and radiographic DAP values and their percentages of the total DAP, average values of kVp, number of radiographic frames, total fluoroscopy time and its percentage of the total irradiation time. Results are presented as mean values with ranges given in the parenthesis.
|
shows effective doses calculated from air kerma, TLD and DAP measurements as well as entrance surface doses obtained from air kerma and TLD measurements. No TLD measurements were made for the oblique view of cerebral and carotid and also lateral carotid examinations.
|
shows ratios of effective doses derived from air kerma and DAP measurements (EAK/EDAP) for patient and phantom studies, for a variety of examinations.
|
compares effective doses determined in this study with previously published data.
|
compares total DAP and fluoroscopy times for several examinations.
|
In the simulation of angiographic procedures on Rando phantom, effective doses for each procedure were found within the range of patient results.
|
Discussion |
|---|
|
TopAbstractIntroductionMaterials and methodsResultsDiscussionConclusionReferences |
|---|
Wise et al [12] investigated the sensitivity of the conversions from entrance skin dose and kerma–area product (KAP) to the likely variations in tube potential, field size, patient size and sex for chest (posteroanterior (PA), left lateral (LLAT) view) and lumbar spine examinations (anteroposterior (AP) and LLAT views). Deviation of field size and position from the assumed theoretical model caused variations of E/KAP and E/entrance skin dose ratios up to a factor of 2 and 3, respectively. Marshall et al [13] found that calculated effective doses for the abdominal PA projection using the DAP method were on average 40% lower than those using entrance surface doses derived using normalized organ dose data. Both techniques agreed to within 10% for the PA view. Their explanation for these discrepancies was the use of differences in organ dose conversion factors for the effective dose calculations.
We noticed that some technical parameters also cause further variations in DAP and air kerma results and this was proven with phantom experiments. It was easy to appreciate the variations of DAP and air kerma in uniform phantom experiments since the function of the automatic brightness control (ABC) was only related to the technical factor under investigation rather than the non-homogeneous structure of the exposed areas as in the case of Rando phantom. However, the Rando phantom was also used in these experiments for the confirmation of lucite phantom results. In order to record the post exposure data, only digital image acquisition was performed for the phantom studies, so that the effect of ABC on tube output was observable.
When the focus–phantom entrance surface distance was increased, DAP readings remained constant as expected but air kerma decreased with the inverse square law. Consistent variation of air kerma and DAP was observed in the case of focus–image intensifier distance changes. These results were also confirmed from the inspection of post exposure data in which variation with the tube current was noticed.
Collimation of the X-ray beam affects the tube output due to the function of the ABC. On our system, reducing field size results in increasing both tube current and pulse width. Therefore when exposing the uniform phantom at increasing collimation there was also a reduction in DAP values as expected. However, this decrease was not proportional to the reduction of the exposed field size due to the function of the ABC. On the other hand, air kerma values at different collimations increased consistently with the increase in the tube current. However this situation may differ in patient examinations, since the amount of tissue or bony structures that will remain in the X-ray beam may change considerably with collimation and cause variations in the ABC response.
There was also a consistent variation of air kerma and DAP values with the function of the ABC when the magnification factor of the image intensifier was switched from one field to another one.
E values obtained from air kerma measurements were higher than those calculated from DAP readings both for patient and phantom studies (Table 3
). If the ratios of effective doses (EAK/EDAP) for each procedure are taken then the mean of these data is found as 3.38 (1.15–5.15).
We also calculated E using TLD results for comparison purposes. Only a limited number of TLDs were used at certain positions on the skin and due to the variation of irradiation geometry they did not remain within the exposed area for the entire examination. Therefore E results calculated from TLD measurement gave considerably lower values.
During the patient studies, careful recording of exposure parameters and use of the most appropriate technical parameters have resulted in a decrease of effective dose ratios for angiographic studies. The remaining differences may be attributable to variations in correction factors.
Since the DAP measurements include a field size component, conversion from DAP to E is less sensitive than that from air kerma measurements. This can also be shown with the use of data in Tables 1 and 2. Considering only the single projection examinations, correlation coefficients (r) are found to be 0.96 and 0.75 for EDAP vs DAP and EAK vs air kerma, respectively.
Higher effective doses were found for therapeutic procedures in the majority of the examinations. Renal and hepatic studies gave the maximum values. We think the complexity of a procedure (irrespective of being diagnostic or therapeutic) may influence the patient dose. For multiprojection studies, effective doses were found individually for each projection. Their contributions were then added to give an insight into the whole examination. Although the acquisition of the exposure data and other related parameters on-line is not easy, more reliable information can be obtained regarding the distribution of doses over the body.
Comparison of E values with other studies are easy for the hepatic, renal, upper extremity and thoracic examinations since the exposed body portion is well defined. In general, our findings are comparable with other published results (Table 4). However, it is not easy to make direct comparisons of effective doses with the other studies for the lower extremity angiography unless the partition of DAP between different regions of the body and selected conversion factors are clearly indicated [9, 14]. If we consider the similar studies, our results (3.5 mSv) are found to be in close agreement with previously published data.
In the case of multiple projections (cerebral and carotid), comparison of the results has to be done carefully due to the use of different conversion factors and averaging of projections. Our fluoroscopy times and DAP values are approximately half of McParland's data (5 min and 23.8 Gy cm2 vs 10.3 min and 49.3 Gy cm2) for the carotid procedures. However our effective doses are exactly the same (4.9 mSv). This is because McParland took the average of the conversion factors for cerebral and carotid projections instead of factors specific to carotid procedures [7]. If the conversion factors of lateral and AP views of cervical projections are compared, large differences are seen. There is also some discrepancy in the lateral and PA projections for the head. Averaging of projections may introduce errors in effective dose results.
The contribution of lateral views to our total DAP was around 25–30% for cerebral and 15–20% for carotid studies. We calculated the effective dose for each projection and found a total of 3 mSv, which is at the lower range of reported results. It should be noticed that effective dose is not an additive quantity. The purpose of addition of each projection result is to be able to make a simple comparison with the literature data which are usually given for the whole examination. On the other hand summation of E values would propagate the uncertainties and increase the overall error in the results.
Comparison of our DAP values and fluoroscopy times for each procedure with previously published results is made in Table 5. Although the use of DAP and fluoroscopy times are the useful parameters for comparison of patient exposures, the dependency of DAP on beam quality, position of the projection and potential variation on the skill of the radiologists complicates this comparison. If we consider the average of these published results, our DAP and fluoroscopy times are close to this range for renal, thoracic, cerebral studies and remarkably less for the hepatic, carotid and lower extremity examinations.
Skin doses obtained from air kerma measurements gives the worst case result from the point of deterministic effects. The maximum individual findings among the single projections appears on hepatic studies (1470 mGy) and cerebral oblique projections (1390 mGy). However, even for a single projection, the X-ray beam is not directed to same area during the course of the whole examination. ESD measured with TLDs gives a better estimation of deterministic effects. The maximum skin dose for the cerebral PA projection was 441 mGy, which is a factor of three less than the air kerma findings and quite far from the deterministic threshold. Due to the large variations in X-ray field size during the examination, significant differences occurred between the air kerma and TLD results. Because of the practical limitations, a larger number of TLDs could not be used. Positioning errors of TLDs at the points where the maximum dose is expected, may also create some uncertainties.
We can also compare our skin dose values with the results of other studies given in the literature. Reported values for hepatic angiography are 1052 mGy [20], 157 mGy [15] and 340 mGy [7] and they are all considerably higher than our TLD results. Regarding cerebral angiography, 1520 mGy [16], 615 mGy [17], 350 mGy [18] and 340 mGy [7] are reported; our average TLD measured skin doses for diagnostic and therapeutic procedures are 354 mGy and 448 mGy, respectively, which is close to the lower end of the range of these results. Verdun [19] presented a low value for cerebral angiography (200 mGy). However his result for carotid (376 mGy) and thoracic (107 mGy) are higher than our findings.
|
Conclusion |
|---|
|
TopAbstractIntroductionMaterials and methodsResultsDiscussionConclusionReferences |
|---|
Estimation of E using DAP measurements may be more accurate than using air kerma (or entrance surface dose) measurements as DAP allows for variations in field size.
We also compared our results with the data given in the literature and found generally good agreement, although lower DAP and effective dose results were found in comparison with other studies.
|
Footnotes |
|---|
Received for publication January 17, 2002. Revision received June 24, 2003. Accepted for publication September 18, 2003.
|
References |
|---|
|
TopAbstractIntroductionMaterials and methodsResultsDiscussionConclusionReferences |
|---|
This article has been cited by other articles:
 |
|
 |
|
  S. Suzuki, S. Furui, I. Yamaguchi, M. Yamagishi, A. Watanabe, T. Abe, and I. Kobayashi Effective Dose during Abdominal Three-dimensional Imaging with a Flat-Panel Detector Angiography System Radiology, February 1, 2009; 250(2): 545 - 550. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 C. J. Martin RADIATION DOSIMETRY FOR DIAGNOSTIC MEDICAL EXPOSURES Radiat Prot Dosimetry, March 28, 2008; (2008) ncm495v1. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
D. Bor, T. Sancak, T. Toklu, T. Olgar, and S. Ener Effects of radiologists' skill and experience on patient doses in interventional examinations Radiat Prot Dosimetry, March 1, 2008; 129(1-3): 32 - 35. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
D. Bor, S. Cekirge, T. Turkay, O. Turan, M. Gulay, E. Onal, and B. Cil Patient and staff doses in interventional neuroradiology Radiat Prot Dosimetry, December 1, 2005; 117(1-3): 62 - 68. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 S. Suzuki, S. Furui, I. Kobayashi, T. Yamauchi, H. Kohtake, K. Takeshita, K. Takada, and M. Yamagishi Radiation Dose to Patients and Radiologists During Transcatheter Arterial Embolization: Comparison of a Digital Flat-Panel System and Conventional Unit Am. J. Roentgenol., October 1, 2005; 185(4): 855 - 859. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| BJR | DMFR | IMAGING | ALL BIR JOURNALS |
