| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Original Research |
1 CT Department, Konstantopoulio-Agia Olga Hospital, Athens, Greece.
2 Present address: Department of Medical Physics, Agios Savvas Hospital, 171
Alexandras Ave., Athens, Greece 11522.
Received May 25, 2006;
accepted after revision December 29, 2006.
Address correspondence to I. A. Tsalafoutas.
Abstract
 Top Abstract IntroductionMaterials and MethodsResultsDiscussionAPPENDIX 1: Defintions of...References |
|---|
MATERIALS AND METHODS. In total, 49 interventions were retrospectively studied: 14 biopsies, 14 radiofrequency ablations, 14 abscess drainages, and seven nephrostomies. CT images were acquired from the department's PACS system and reviewed to record the scan parameters of each slice. Entrance surface dose and E were estimated using the Impactscan database and the related Monte Carlo conversion coefficients.
RESULTS. Median values of E for biopsies, radiofrequency ablations, abscess drainages, and nephrostomies were 23, 35.3, 16.2, and 11.5 mSv, respectively. Respective ranges were 5.8–46.6, 18.4–57.2, 10.9–31.5, and 5.1–32.7 mSv. The corresponding median values and ranges for the peak absorbed dose were 281, 557, 155, and 145 mGy and 133–982, 147–699, 94–315, and 75–297 mGy. The diagnostic scans obtained before the interventions were responsible for 63%, 33% 40%, and 51% of E, respectively. The largest contribution to the peak absorbed dose was due to positioning of the tissue acquisition biopsy gun in biopsies (48%), the radiofrequency needle in ablations (57%), and the catheter in abscess drainages (41%) and nephrostomies (49%).
CONCLUSION. For the CT interventions studied, and especially for biopsies and radiofrequency ablations, patient effective doses were considerably high. Maximum peak absorbed dose observed was about 1 Gy, considerably lower than the threshold for deterministic effects (2 Gy).
Keywords: biopsy • CT • interventional radiology • radiation dose • radiofrequency ablation
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionAPPENDIX 1: Defintions of...References |
|---|
Although most interventional radiology procedures are performed with specialized angiographic units, CT units (with and without the CT fluoroscopy option) are also used for intervention because of the excellent anatomic visualization. In addition to the diagnostic and therapeutic benefits that CT intervention may offer, an important advantage for units without CT fluoroscopy is that they do not involve exposure to medical personnel. On the other hand, exposure of the patient is an inevitable shortcoming of any diagnostic or therapeutic procedure that uses X-rays, and thus the associated radiation risk for the patient should be always of concern. This concern refers to both the stochastic and the non-stochastic (deterministic) effects of radiation.
At present, the dominant notion for stochastic effects is that no threshold exists and that the probability of occurrence increases roughly linearly with patient dose. To quantify the stochastic risk, effective dose (E) is considered the most suitable dosimetric quantity because it takes into account the doses received by all radiosensitive organs weighted according to their radiosensitivity. Deterministic effects have a threshold below which no direct radiation-induced damage to tissues and organs occurs. Threshold doses for most tissues and organs are quite high—on the order of several Gy—and thus for diagnostic and interventional radiology, only the acute skin reactions like erythema and epilation are of concern. These deterministic effects require a peak absorbed dose to the skin of the patient of at least 2 Gy to occur, and doses of this magnitude are not considered uncommon for some complicated interventional radiology procedures [1].
Although data on patient effective and absorbed doses from interventional procedures performed in angiographic units are widely available, at present, relevant data for CT interventions are quite limited. In fact, concerning CT interventions performed without CT fluoroscopy, only one reference was found in the literature that compares patient doses with those obtained with another CT unit equipped with CT fluoroscopy [2]. CT fluoroscopy is a relatively new technique that allows rapid low-dose images to be obtained with the interventionalist at the patient's bedside. CT fluoroscopy can facilitate the interventionalist in the manipulations, thus reducing the time required for the completion of the procedure and the dose to the patient, but it also involves a radiation dose to the medical staff [2].
This article presents the results of a study on patient doses during four CT-guided interventional radiology procedures performed with a helical CT scanner without CT fluoroscopy. Both effective dose and peak absorbed dose were retrospectively calculated using data stored in the CT department's PACS. The results were analyzed and compared with data available in the literature.
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionAPPENDIX 1: Defintions of...References |
|---|
Forty-nine CT-guided interventions performed in adult patients were retrospectively studied, including 14 biopsies (nine in the liver, two in the kidney, two in the psoas, and one in the pancreas), 14 radiofrequency ablations (all in the liver), 14 abscess drainages (seven in the liver, six in the chest, and one in the kidney), and seven nephrostomies. All these procedures were performed by a radiologist having more than 20 years of experience. The biopsies were always performed with a biopsy gun to acquire an appropriate tissue specimen, whereas the nephrostomies were performed with the Seldinger technique because most were difficult cases. Finally, most abscess drainages were performed with the trocar technique and wide catheters ranging from 8 to 12 French. In all cases, full sterile techniques were used, and appropriate measures for gravitational effects on catheters and biopsy gun during CT scans were taken. Furthermore, a follow-up checkup scan was always obtained at the end of every procedure.
All interventional procedures are routinely stored in the CT department's PACS, from which they can be retrieved and reviewed in detail. Thus, for each individual case the images from all scans are available with all the technical parameters shown. These data were used to retrospectively calculate the patient dose from a single slice or a series of slices and, consequently, the cumulative dose from the whole procedure. Furthermore, when reviewing these images on the workstation monitor, the interventionalist assigned each image to one of the following stages: A, scanogram; B, diagnostic examination; C, positioning of the metal marker or grid; D, positioning of the anesthesia needle; E, positioning of the biopsy specimen acquisition needle, the radiofrequency needle, the drainage catheter, or the nephrostomy catheter; and F, lesion examination after intervention. It must be noted that for stages B, C, and F, a helical scan is usually used, whereas for stages D and E, conventional axial scans are routinely obtained.
CT Dosimetry
CT patient dose calculation is a rather complicated procedure requiring the
use of CT-specific dosimetric quantities, such as the CT dose index (CTDI) and
the dose–length product (DLP). Detailed description of these terms is
given in Appendix 1 and the relevant literature
[3,
4].
E can be derived from CTDI values using Monte Carlo techniques [5]. In our study, patient dose calculations were performed using the ImPACT CT Patient Dosimetry Calculator (CTDosimetry.xls), henceforth referred to as CTDosimetry. CTDosimetry is a Microsoft Excel–based program freely available on the Internet (impactscan.org). This program contains the CTDI values normalized per 100 mAs for air (nCTDIair) and for the center (nCTDIc) and periphery (nCTDIp) of both body and head phantoms, for a wide range of CT scanners and tube potentials, along with correction factors for different collimations. It calculates the weighted CTDI (CTDIw), the volume CTDI (CTDIvol), and the DLP from the scanning parameters applied in a CT examination. When the anatomic area scanned is defined on the mathematic phantom included in the program for simulating human anatomy and the National Radiological Protection Board (NRPB) SR250 data sets (containing the coefficients required for Monte Carlo calculations) are available, a calculation of E can be also obtained.
However, this program is not capable of calculating the E for the scan projection view. A scanogram resembles a conventional radiograph in many ways. However, because the scanogram is performed using very thin collimation (1 mm in our CT scanner), an exposure index equivalent to the tube loading (mAs) of a conventional radiograph can be derived by multiplying the scanogram mAs by the collimation thickness (in mm) and dividing by the scanned length (in mm). Thus, taking into account that during the scanogram the tube is not rotated, the entrance surface dose to the phantom surface can be calculated using the soft-tissue CTDI (CTDIs = 1.07 x CTDIair) for the collimation used for performing the scanogram and correcting for the distance difference of the phantom surface from the isocenter by applying the inverse square law.
For the calculation of E from scanograms, an alternative method was used based on the observation that the conversion coefficients tabulated in the NRPB-R262 report [6] for deriving E from entrance surface dose for the anteroposterior projections of kidneys, abdomen, pelvis/colon, lumbar spine, and thoracic spine are almost proportional to the field area at the image receptor and the phantom midplane. Thus, the ratios of the conversion coefficients to the corresponding field areas are quite similar and at 120 kVp (the largest kVp considered in the NRPB-262 report) range from 0.00012 to 0.00013 mSv · mGy–1 cm–2 with respect to the field areas at the receptor and from 0.000163 mSv · mGy–1 cm–2 (for abdomen) to 0.000186 mSv · mGy–1 cm–2 (for pelvis/colon) with respect to the field areas at the phantom's midplane. Therefore, E can be estimated by multiplying the entrance surface dose by the product of the scanogram's length (in cm) and the diameter of the standard body phantom (i.e., 32 cm) and by a factor of 0.00017 mSv · mGy–1 cm–2. Taking into account that the scanogram contributes only a small fraction of the dose during diagnostic or interventional CT procedures, any errors that may be introduced by this approximate method are small.
To estimate E, the corresponding anatomic region scanned on the patient must be defined on the mathematic phantom. Because the table position was available on each slice, the only additional thing needed was to use the table position indication of a slice containing a characteristic anatomic structure that could be identified in the mathematic phantom. The characteristic anatomic structure most often used in our study was the border between the heart and the liver that corresponds to 43 cm on the mathematic phantom. In a few cases in which a second scanogram was performed with different patient positioning, or the patient positioning was altered for any reason, care was taken to reduce all the new table position indications to the first series and from there to the mathematic phantom. Furthermore, to determine the peak absorbed dose and the anatomic region where that occurred, one must calculate the overlap of all scans and their respective entrance surface doses.
For this purpose we developed an appropriate spreadsheet in Microsoft Excel in which the patient sex, exposure data, collimation, and table position for all slices were entered. With this spreadsheet, the skin cumulative exposure from all scans was calculated, and the anatomic position of the maximum entrance surface dose (peak absorbed dose) was identified. Furthermore, using commands written in Microsoft Visual Basic, these data were input to the CTDosimetry, and the calculated DLP and E were fed back into the spreadsheet. Because each slice or series of slices was assigned to one of the six stages to which each procedure was analyzed, it was also possible to calculate for each patient the contribution of each stage to the peak absorbed dose and E. Finally, to estimate the time required for each intervention to be completed, the times noted on the first and the last images were used. Although data concerning the duration of interventions were not further analyzed, they can serve as a straightforward index of the difficulty and complexity of each intervention type.
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionAPPENDIX 1: Defintions of...References |
|---|
|
|
As shown in Table 2, the diagnostic part (stage B) of the above procedures was responsible for the largest part of E, with the median values being 63%, 33%, 40%, and 51%, for biopsies, radiofrequency ablations, abscess drainage, and nephrostomy, respectively. However, for radiofrequency ablations, a similar contribution to that of stage B was also observed for stages E and F. It must be noted that the number of scans acquired in each stage greatly varied from case to case, and on a few occasions some of the stages were omitted.
Concerning the peak absorbed dose values, these were mostly due to the last stage of the interventional part of the procedure (i.e., stage E) as a result of scan overlapping at the anatomic region of interest. Indeed, the largest contribution to the peak absorbed dose was due to the positioning of the specimen acquisition needle in biopsies (48%), the radiofrequency needle in ablations (57%), and the catheter in abscess drainages (41%) and nephrostomies (49%). Schematic diagrams showing the absorbed dose distribution in the cases with the highest peak absorbed dose for each of the four intervention types studied are given in Figures 1A, 1B, 1C, 1D.
|
|
|
|
The peak absorbed dose in all anatomic positions was calculated using the CTDIp values as an approximation of entrance surface dose (except for scanograms) that refers to the dose at 1 cm below the surface of a body phantom with a standard diameter of 32 cm. Thus, for both the larger and smaller patients included in this study (as indicated by the anteroposterior and lateral diameters given in Table 1), the actual peak absorbed dose can be different from that calculated and, in addition, larger for the lateral than for the anterior and posterior entrance surfaces. When the inverse square law correction is applied to account for the different distances of focus-to-phantom and focus-to-patient entrance surfaces, the actual peak absorbed dose can be more accurately approximated, provided that the patient is centrally positioned in the gantry.
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionAPPENDIX 1: Defintions of...References |
|---|
The mean E values reported by Teeuwisse et al. [2] are three times lower for biopsies but only 20% lower for drainages compared with the mean values of 23 and 16.2 mSv calculated in our study. In an effort to explain these differences, the technical parameters were compared. The tube potential and loading used in both studies were similar: 225–250 mAs with 120 or 140 kVp in the referred study and about 200 mAs (160–300 mAs) with 120 kVp in our study, whereas in both studies the mAs were not lowered during the interventional part of the procedures. In both studies the collimation was 5–7 mm, however, in the referenced study a pitch factor of 1.4 was used, whereas in our study the pitch factor was always 1. Furthermore, the CTDI values of the Philips LX scanner are quite lower compared with our scanner (according to the CTDosimetry for 120 kVp and 5-mm collimation, the nCTDIair for the Philips LX is 19.3 mGy, whereas for the GE Prospeed it is 38 mGy; the respective values for nCTDIw in the body phantom are 7.4 and 10.1 mGy).
Concerning the accuracy of the CTDI values used in the preceding comparisons, please note that although small differences in the actual CTDI values may exist among different scanners of the same model, according to the results of the NRPB survey in the United Kingdom [7], a generally good agreement between measured and tabulated CTDI values (included in CTDosimetry) should be expected. Indeed, for our scanner in which the X-ray tube has been replaced six times since the CT scanner was first installed, differences between measured CTDI values and those given by CTDosimetry never exceeded a –5% to 10% range.
To appreciate the effect of the different CTDIs and pitch factors, consider this example. For a full abdominal or pelvis CT scan ranging from 0 to 45 cm on the mathematic phantom of the CTDosimetry (without determining sex) with exposure factors 120 kVp, 200 mAs, 5-mm slice thickness, and pitch factor equal to 1, the E would be 14 mSv with the GE Prospeed and 11 mSv with the Philips LX, which, with pitch factor equal to 1.4, would be reduced to 7.9 mSv. The unisex phantom assumption was emphasized above because the corresponding E values for men and women are different (for the example given for the GE Prospeed scanner, the respective values for men and women are 12 and 16 mSv). The larger pitch factor and the lower CTDI values of the Philips LX CT scanner used in the interventions studied by Teeuwisse et al. [2] may explain in part the differences in E. It seems, however, that significant differences between the two studies should also exist in terms of the technique and the number of slices acquired in biopsies and drainages, but such information is not reported in the referenced study for comparison with our data.
In contrast to what should be expected from the lower E values, the mean values of peak absorbed dose reported by Teeuwisse et al. [2] compared with our values are almost equal for biopsies and almost 2.5 times higher for drainages, whereas the maximum peak absorbed dose values reported in that study are almost 1.5 and 5 times larger than the values of 982 and 315 mGy observed in our study for biopsies and drainages, respectively. Although these differences may at first seem difficult to explain, they may denote that in the procedures studied by Teeuwisse et al., the scans were more focused around the region of interest.
In general, it is difficult to make reliable comparisons between different CT facilities in terms of patient E and peak absorbed dose from CT-guided procedures, especially when the sample sizes are small. CT-guided procedures, even when performed in the same CT facility and by the same interventionalist, present great variability in the difficulties encountered during lesion localization and device placement, depending on lesion size, site, density, and contrast enhancement. Furthermore, gravitational effects on devices such as biopsy guns and trocar catheters considerably increase difficulties in determining the appropriate entrance angle. Indeed, this is also shown by the large range of E and peak absorbed dose values observed in both our study and that by Teeuwisse et al. [2]. However, the results of both studies should be considered indicative of the level and the variability of patient doses involved in CT-guided interventional procedures.
Although the above comparisons concern only the biopsies and abscess drainages, it is quite obvious from the E values given in Table 1 that the effective doses recorded in our study are considerably high. Consequently, ways to reduce patient dose without impairment of the diagnostic and therapeutic outcome of these procedures should be considered. Given the scanner type available, to reduce patient dose the most obvious actions relate to the reduction of the number of slices acquired during each stage, with a corresponding reduction of the anatomy scanned to that strictly necessary and/or the increase of the pitch factor. Although these actions may seem commonplace, it should not be taken for granted that they are always applied by the interventionalists and the radiation technologists involved in these procedures.
Apart from these options, the use of lower mAs, especially for the scans concerning the positioning of the metal marker, the anesthesia needle, the biopsy gun, the radiofrequency needle, or the catheter seems the next reasonable step to be taken. Because in a CT-guided interventional procedure an initial diagnosis already has been made with a previous diagnostic CT or other alternative imaging technique, stage B can be omitted (as when a recent diagnostic CT is available) or at least be confined to the region of interest. Except for the cases in which a detailed reevaluation of the patient pathology is necessary, a tube loading of about 100 mA may suffice (depending on the patient size) for stages B and F, whereas even lower mA values could be adequate for stages C, D, and E, in which image quality is not the major issue, except probably for certain biopsies when a lesion may contain different heterogeneous regions that must be clearly distinguished. Reduction of the tube loading from 200 to 100 mAs could reduce the doses by a factor of at least 2 without need for increasing the pitch factor. Furthermore, because when reviewing the electronic records the interventionalist has identified many scans that could have been avoided, he or she can be expected in the future to be more cautious and avoid unnecessary patient exposure.
Although theoretically by applying these simple rules, a reduction of dose by a factor of more than 2 seems feasible, this remains to be proven in practice. Because the major concern of interventionalists during such procedures is the successful outcome of the intervention, patient radiation protection issues may easily be overlooked in order to increase the image quality, even when not really necessary. Therefore, interventional protocols should be prepared with the lowest possible mAs that interventionalists have agreed can produce an image quality sufficient for their purposes. The aforementioned actions for optimizing patient doses in CT interventions performed in our CT facility are currently under investigation.
APPENDIX 1: Defintions of CT Specific Dosimeter Quantities
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionAPPENDIX 1: Defintions of...References |
|---|
The weighted CTDI (CTDIw) is used for approximating the average
dose over a single slice and is defined separately for the head and body
phantoms using the following equation:
 | (1) |
The volume CTDI (CTDIvol) takes into account the parameters that
are related to a specific scanning protocol and is defined by the following
equation:
 | (2) |
Finally, dose–length product (DLP) is used to calculate the dose for
a series of slices or a complete examination and is defined by the following
equation:
 | (3) |
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionAPPENDIX 1: Defintions of...References |
|---|
CiteULike 
Complore 
Connotea 
Del.icio.us 
Digg 
Reddit 
Technorati What's this?
This article has been cited by other articles:
|
|
 |
|
  I. A. Tsalafoutas, V. Tsapaki, C. Triantopoulou, C. Pouli, V. Kouridou, I. Fagadaki, and J. Papailiou Comparison of Measured and Calculated Skin Doses in CT-Guided Interventional Procedures Am. J. Roentgenol., November 1, 2008; 191(5): 1601 - 1607. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
