BJR
HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
First published online January 3, 2007
British Journal of Radiology (2007) 80, 362-366
© 2007 British Institute of Radiology
doi: 10.1259/bjr/95349672
This Article
Right arrow Abstract Freely available
Right arrow Figures Only
Right arrow Full Text (PDF)
Services
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrow reprints & permissions
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Bridcut, R R
Right arrow Articles by Winder, R J
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Bridcut, R R
Right arrow Articles by Winder, R J

Full paper

Patient dose from 3D rotational neurovascular studies

R R Bridcut, BSc, MSc1, E Murphy, BSc, MSc2, A Workman, BSc, MSc1, P Flynn, MRCP, FRCR2 and R J Winder, BSc, MSc, PhD3

1 Northern Ireland Regional Medical Physics Agency, Forster Green Hospital, Belfast BT8 6HD, 2 Department of Neuroradiology, Royal Victoria Hospital, Belfast BT12 6BA, 3 Health and Rehabilitation Research Institute, University of Ulster, Shore Road, Newtonabbey BT37 0QB, UK

Correspondence: Miss Ruth R Bridcut, Radiation Protection and Imaging, Northern Ireland Medical Physics Agency, Forster Green Hospital, 110 Saintfield Road, Belfast, N. Ireland, UK. E-mail: ruth.bridcut{at}mpa.n-i.nhs.uk


   Abstract
Top
 Abstract
Introduction
 Methods and materials
 Results and discussion
 Conclusion
 References
 
The use of image-guided interventional radiological techniques is increasing in prevalence and complexity. Imaging system developments have helped improve the information available to interventionalists to plan and guide procedures. Information on doses to patients resulting from alternative imaging techniques or protocols is useful for both the process of justifying particular procedures and in optimizing the resultant exposures. Such information is not always available, especially for new or developing imaging techniques. We have undertaken a study of doses to patients associated with two alternative imaging methods for pre-intervention assessment of intracranial aneurysms. In the first technique the aneurysm is assessed from a series of digital subtraction angiography (DSA) runs taken at different imaging projections. The second technique involved acquiring images from one single image run while the imaging system rotated 180° around the patient's head. In this technique, the aneurysm was then evaluated from a 3D reconstruction of the projection images. Effective doses were calculated using a computer model to simulate the exposure geometry and parameters. The mean dose from the DSA protocol used at our centre was 3.4 mSv and from the 3D rotational angiography (RA) technique was 0.20 mSv.


   Introduction
Top
 Abstract
 Introduction
Methods and materials
 Results and discussion
 Conclusion
 References
 
The presence of a cerebral aneurysm is a serious and potentially life-threatening medical condition. Often these remain undetected until they burst and the patient suffers a subarachnoid haemorrhage with potentially life-threatening consequences. A cerebral aneurysm can be identified by diagnostic imaging techniques such as digital subtraction angiography (DSA). In recent years there have been significant advances in interventional radiological techniques which allow such lesions to be treated without the need for traditional neurosurgery [1]. The development of such techniques would have been impossible without the use of imaging for the assessment, guidance and verification of the treatments. An accurate assessment of the anatomy of the aneurysm and surrounding vessels is necessary before proceeding with any intervention. The ideal imaging system for the diagnosis of intracranial pathology would allow safe, rapid, high resolution and high contrast imaging of the vascular structure with the ability to manipulate the obtained images in multiple projections to clarify anatomic ambiguity. Until recently the assessment of the necessary treatment was carried out using two-dimensional (2D) projection DSA studies. The purpose of this was to identify the aneurysm and determine the anatomy of the aneurysm including the dimensions of the aneurysm neck, the ratio of the neck to the size of the sac and the relationship to surrounding vessels. Evaluation of intracranial aneurysms using 2D projection DSA requires multiple views to delineate the morphology of the aneurysm and surrounding vessels. Further, if the anatomy of the aneurysm and surrounding structure is complicated, additional views will be necessary thus increasing the radiation and contrast media doses to the patient, the examination time and the overall cost of the procedure.

Three-dimensional rotational angiography (3D RA) is a relatively new imaging technique which provides accurate assessment of intracranial aneurysms [2, 3]. 3D RA is designed to provide a comprehensive information data set from a single imaging run, thus minimizing both radiation and contrast media doses to the patient. In this technique individual 2D angiographic projections are acquired as the imaging system C-arm subtends a 180° arc axially around the patient's head. All projection images are obtained during a single intra-arterial injection of contrast media. The projection images, following correction for distortions in the imaging system, are then used to reconstruct a 3D volume data set using a cone beam reconstruction algorithm [4]. The 3D data set may then be presented for viewing as a surface shaded display or maximum intensity projection. The volume reconstruction can be rotated and viewed at any angle allowing optimum delineation of the pertinent anatomy. This can improve the understanding of the 3D morphology of the vasculature and assist in the selection of the optimum imaging projection to be used for guidance of the intervention.

3D RA has clearly provided radiologists with useful clinical information to assist in interventions to treat cerebral aneurysms [2, 5]. However there is no information available on the radiation dose received by patients from this technique or how doses compare with those from conventional projection DSA used for the assessment of intracranial aneurysms. The purpose of this study was to develop a means of determining the radiation dose to patients undergoing 3D RA and to compare the doses with that from the conventional DSA technique used at our centre to assess aneurysms. It is acknowledged that DSA protocols may vary between centres and our comparison is limited to the DSA protocol described. The evaluation has been limited to those parts of the patient examinations which involved the aneurysm assessment phase. It is recognized that this only forms part of the overall patient examination. There are other parts of the examination such as the catheterization and intervention stages which involve similar procedures, independent of the technique used to evaluate the aneurysm. Since the purpose of this study was to compare doses from the different assessment procedures, we decided to confine our dose evaluation to these procedures rather than to consider the entire examination. Effective dose (ED) [6] provides a means of comparing the stochastic risks from different radiological procedures. Since the two techniques being compared in this study involve quite different irradiation geometries it was felt that effective dose would provide the best metric for comparing the radiation risk to patients.


   Methods and materials
Top
 Abstract
 Introduction
 Methods and materials
Results and discussion
 Conclusion
 References
 
Neurovascular studies are performed in the Neuroradiology Department, Royal Victoria Hospital, Belfast. The examination room is equipped with a Philips BV5000 Bi-plane vascular system (Philips Medical Systems, Eindhoven, The Netherlands). This equipment has an isocentric C-arm with a capability of acquiring rotational angiographic images. Projection images are transferred to a Silicon Graphics Octane® workstation (Silicon Graphics Inc., Mountain Hill, CA) for 3D reconstruction.

A prospective study was performed over a 6 month period where 46 patients were identified. Of these, 28 patients had 3D RA and 18 had conventional 2D projection DSA for aneurysm evaluation. In both methods assessment was carried out on only one vessel (either the left or right carotid artery).

For conventional 2D projection DSA runs, 8 cm3 of contrast medium (Iomeron 300; Bracco, Milan, Italy) was rapidly injected by hand from a 10 cm3 syringe. The normal protocol for this type of imaging sequence was an image acquisition rate of 2 or 3 frames per second until early venous phase is depicted. The default cerebral angiography DSA acquisition protocol provided on the BV5000 system was used. This program was set up by the manufacturer and the dose per image was not adjustable by the user. Radiographic factors were recorded for each angiographic run including kV, mAs, projection angle, frame rate, acquisition run length, number of runs and image intensifier field selection. This process was facilitated as most of these data were recorded by the system during the examination and were available as a print-out following completion of the examination.

The 3D RA acquisition protocol was the default provided on the system and was not adjustable by the user. During 3D RA acquisitions the patient is positioned such that the anatomy of interest is at the isocentre of rotation, with the focus to image intensifier distance (FID) fixed at 120 cm. A 17 cm image intensifier field size is always used. During acquisition the C-arm moves the imaging system around the patient in a 180° arc beginning in the left lateral projection, moving though the posteroanterior (PA) projection and finishing in the right lateral projection. A frame rate of 12.5 images per second is used during acquisition with 100 images of 10 ms exposure time being acquired and the whole acquisition process taking about 8 s. For 3D RA studies typically 26 cm3 of contrast was injected at 3 cm3s–1.

The kV–mA response of the automatic exposure control system for the RA scan was measured by placing varying amounts of copper filtration at the X-ray tube head and carrying out a rotational scan. Copper was used rather than PMMA or another tissue equivalent material as it could be fixed to the tube head during rotation and so provide the same attenuation throughout the entire scan. This allowed determination of the mAs used by the system for a given selected kVp. The imaging system did not record the exposure factors used during the 3D RA run (Drazenko Babic, Philips Medical Systems, private communication, 2002) and only printed the kV and mA used for the final frame in the rotational imaging run. These values were recorded for each run.

To estimate the change in radiographic factors during a 3D rotational run the kV and mA were recorded during the scan of a water phantom. The phantom simulated a head, having an elliptical cross-section, measuring 18.5 cm in the anteroposterior dimension and 14 cm in the lateral dimensions. The dimensions of this phantom are close to the mean head dimensions measured by Huda et al [7] for adult patients undergoing CT head examinations. They reported an anterior–posterior dimension of 18.9±0.9 cm and a lateral dimension of 14.8±0.8 cm. Scanning the phantom simulated the scanning of a head, allowing the system to respond to the variation of thickness with projection angle. The response of the system to heads of different sizes was determined by acquiring additional rotational scans of the water phantom with aluminium filters attached to the tube head. For each patient, the recorded kV value, being the kV of the final image in the run, was matched to the phantom run which had the same final kV. kV and mA values throughout the run were assumed to match those recorded during phantom simulations. It was difficult to confirm these values in the clinical setting as the numbers on the console changed rapidly and only one run was acquired.

The radiation output characteristics of both X-ray tubes as a function of kV and mAs were measured using a Keithley Triad system (Innovision, Cleveland, OH). Outputs were measured using a 15 cm3 chamber which had a calibration traceable to PTB (Physikalish-Technische Bundesanstalt, Braunschweig, Germany). An estimate of the total filtration of the tube was obtained using HVL data and voltage ripple measurements [8]. X-ray tube output characteristics and the recorded exposure factors enabled the patient entrance air kerma to be calculated for each angiographic projection. The image intensifier dose per image frame was measured for both 2D DSA and 3D RA by positioning a 150 cm3 chamber on the surface of the image intensifier unit, above the anti-scatter grid, during an angiographic run with a 2 mm copper filter placed at the X-ray tube to provide patient equivalent exposure factors.

PCXMC (STÜK, Finish Centre for Nuclear Safety, Helsinki, Finland) is a commercially available PC-based Monte Carlo simulation program for determining radiation doses to patients from medical X-ray examinations. The software simulates the irradiation of a mathematical hermaphrodite phantom based on the Medical Internal Radiation Dose (MIRD) type adult phantom [9] and allows the modelling of arbitrary projection geometries. The results of the simulations are given as organ and effective doses. To estimate the effective dose for each angiographic run the radiographic projection geometries (including projection angles, position and field dimensions) and exposure factors (including entrance air kerma without backscatter, kVp and X-ray tube total filtration) were entered into PCXMC.

The simulation in PCXMC uses a rectangular irradiation field; however, the irradiation field used by an image intensifier, unless collimated within the image field, is usually circular or hexagonal. In the simulation we use a square irradiation field with the same area as the hexagonal image intensifier irradiation field. This approximation is justified since, for the projections used, the use of the square field does not involve the irradiation of any other radiation-sensitive organs than would be irradiated had a hexagonal field been used in the simulation.

In 3D RA 100 projection images are acquired. The simulations to calculate effective dose were carried out at 11 different projection angles spaced at 18° intervals, covering an 180° arc. The patient entrance air kerma for each of these exposures was scaled by the ratio of the original number of projections to the simulated number of projections to allow for the smaller number of simulated projections. The effective dose from each simulated projection was summed to give the total effective dose for 100 images.

The system did not allow dose–area product (DAP) to be recorded for each individual run. For each projection, the DAP was calculated by multiplying the air kerma per exposure scaled to the input plane of the image intensifier by the irradiated field area measured in the same plane. The total DAP per patient was calculated by summing the DAP for each projection, and the mean DAP per patient was calculated for each technique. A DAP to effective dose conversion factor was in turn calculated for each technique, by dividing the mean effective dose by the mean calculated DAP value.


   Results and discussion
Top
 Abstract
 Introduction
 Methods and materials
 Results and discussion
Conclusion
 References
 
For the patients in this study who underwent 2D projection DSA (n = 18), the mean number of image runs undertaken was 7.6 (minimum 4, maximum 12). In these sequences generally the 17 cm image intensifier field for the frontal image intensifier system was used although much more infrequently (15% of projections) the 22 cm selection for the lateral image intensifier system was used. Typically the protocol used either 2 or 3 frames per second acquisition speed with a mean number of images per run of 18.4 (minimum 12, maximum 34). The mean HVL and tube filtration was measured as 3.3 and 2.7 mm Al respectively. Acquisition data are summarized in Table 1Go for both 2D DSA and 3D RA procedures. The kV–mA response curve for rotational angiography is shown in Figure 1Go.


View this table:
[in this window]
[in a new window]

  		Table 1. Summary of data for two-dimensional digital subtraction angiography (2D DSA) and three-dimensional rotational angiography (3D RA) acquisitions

 

Figure 1
View larger version (9K):
[in this window]
[in a new window]

  		Figure 1. kV–mA response for exposures during rotational angiography. Moving from left to right along the curve corresponds to increasing attenuation of the X-ray beam.

 
For 2D projection DSA the mean effective dose was 3.4 mSv (range 1.6–4.9 mSv). For 3D RA the mean effective dose was 0.20 mSv (range 0.15–0.26 mSv). The distribution of effective doses for both 2D projection angiography and 3D RA is shown as a histogram in Figure 2Go. The dose scale of the abscissa is shown as a log scale to allow the accommodation of both distributions on the same plot. The mean calculated DAP for each technique was 48 Gy cm–2 and 2.0 Gy cm–2 for 2D DSA and 3D RA respectively.


Figure 2
View larger version (14K):
[in this window]
[in a new window]

  		Figure 2. Distribution of patient effective doses for 3D RA and 2D DSA during the aneurysm evaluation stage of the examination. The abscissa has a logarithmic scale in order to better illustrate the nature of the distributions.

 
The 3D RA results in a dose of less than 10% of that of a comparable 2D DSA series of runs used for the evaluation of the aneurysm. This is due to the reduced mAs per frame for 3D RA, resulting from the reduced image intensifier dose per frame (Table 1Go). While on average a similar number of frames are acquired for each technique (136 for 2D DSA and 100 for 3D RA), the average mAs per frame for 3D RA (1.8 mAs) is less than 10% of that for 2D DSA (25 mAs).

Direct comparison of dose for the two techniques using DAP may not be appropriate as the two techniques use different irradiation geometries. Effective dose is a better indicator of radiation risk, though its magnitude cannot be measured directly. Methods have been developed to determine effective dose using Monte Carlo methods to simulate the interaction of X-rays in an anthropomorphic mathematical phantom [10]. The approach used in this work was to compute the effective dose for each projection using a commercial Monte Carlo based dosimetry program, PCXMC. Results using this programme have been verified against other Monte Carlo codes for paediatric cardiac catheterizations and shown to produce accurate determinations of effective dose [11]. The authors of this program have also shown good agreement of their results with previously published Monte Carlo determined organ and effective doses for medical X-ray examinations.

In 3D RA the system acquires 100 projection images over an arc of 180°. It would be inefficient and tedious to simulate each of these projections with PCXMC in order to determine the effective dose. Instead it was felt that since for the neuroexamination under study there was no projection where radiologically sensitive organs which contributed to the effective dose received greater irradiation than others, and since the exposure factors varied slowly with change in projection angle, it was not necessary to model each individual projection at each angle. Instead we modelled a smaller number of projections at equiangular spacing around a 180° arc and weighted the entrance air kerma from each of these projections to allow for the reduction in the number of projections.

The spread in effective doses for the 3D RA procedures seen in Figure 2Go arises due to variations in the dimension of the patients' heads. The spread in doses for the 2D projection DSA studies arise due to variation both in patient head dimension and in the examination technique related to the specific patient's condition. Due to differences in the presentation of the aneurysm and the complexity of the structure, more or longer DSA imaging runs may be required to obtain the necessary morphological information on the aneurysm and its surrounding vessels. Clearly such patient variability results in a greater variability in the doses for the 2D projection DSA technique.

The DAP to effective dose (ED) conversion factor for the 2D DSA and 3D RA procedures was calculated to be 0.07 and 0.09 mSv (Gy cm–2)–1 respectively. Given the potential errors in calculating both quantities (i.e. DAP and ED) it is probably appropriate to use the mean value of 0.08 mSv (Gy cm–2)–1 for both techniques. Use of this factor may allow other centres to estimate the effective dose and hence the risk associated with their procedures. Marshall et al [12] determined a conversion factor of 0.087 mSv (Gy cm–2)–1 for cerebral angiography and McParland [13] calculated a conversion factor of 0.1 mSv (Gy cm–2)–1 for diagnostic and therapeutic cerebral angiography procedures. For 3D RA, the ratio of effective dose to DAP did not vary significantly with projection (typically a maximum deviation of 8% from the mean ratio), indicating further that the use of a DAP to effective dose conversion may be appropriate for comparing doses for 3D RA with DSA for neuro-angiography examinations.

The mean total DAP value for cerebral aneurysm coiling procedures was 160 Gy cm–2. Using the determined conversion factor, a mean effective dose for cerebral aneurysm coiling of 13 mSv was calculated. The dose for the aneurysm evaluation stage of the procedure using the 2D DSA technique (3.4 mSv) is on average approximately 25% of the dose for the total procedure. Therefore using 3D RA for evaluation of the aneurysm has the potential to reduce the dose for the entire procedure by approximately 25%.

A number of authors have evaluated doses and estimated effective dose in neuroradiological studies for diagnostic angiography and arterial embolism procedures [12–14]. Marshall et al [12] determined an effective dose of 3.6 mSv for diagnostic four vessel cerebral angiography. McParland [13] used factors for converting DAP measurements to effective dose to calculate effective dose for diagnostic cerebral angiography and cerebral embolisation from DAP measurements. He reported a mean ED of 7.4 mSv for cerebral angiography and a mean ED of 10.5 mSv for cerebral embolisation.

Effective dose can be used to estimate the stochastic risks from exposure to ionizing radiation. However for complex interventional radiology examinations where protracted exposure to a localized region of the patient's skin may occur there is a potential that the skin dose may reach levels where deterministic radiation effects occur [14]. The use of 3D RA in the aneurysm assessment phase reduces the overall dose for the interventional procedure and hence reduces the potential likelihood of skin doses reaching levels where deterministic radiation effects occur. Furthermore the dose to the skin with 3D RA is distributed rather than localized.

The results of this study apply to the particular equipment used in our department and hence to the particular imaging protocols used. These protocols were the default selections available on the imaging system for the particular imaging procedure. It was not possible for the user to adjust factors such as the dose per frame to the image intensifier for either 2D DSA or 3D RA. There may be scope to optimize the exposures for both of these procedures; however a study of this would require the evaluation of both the radiation dose and the diagnostic information available. The present study concentrated on evaluating and comparing radiation doses for patients from the default protocols available on this imaging system.


   Conclusion
Top
 Abstract
 Introduction
 Methods and materials
 Results and discussion
 Conclusion
References
 
This patient dosimetry study has shown that the patient effective dose from 3D RA used in the assessment of cerebral aneurysms is significantly lower than that for conventional techniques which use a series of 2D projection DSA runs. Although the assessment is only one stage in the overall endovascular treatment of the aneurysm, the use of 3D RA in this process reduces the overall radiation dose to the patient due to the procedure. The 3D technique also has the advantage that it presents the interventionalist with a 3D representation of the anatomy of the aneurysm.

Received for publication July 4, 2005. Revision received June 28, 2006. Accepted for publication August 18, 2006.


   References
Top
 Abstract
 Introduction
 Methods and materials
 Results and discussion
 Conclusion
 References
 

  1. Guglielmi G, Vinuela F, Sepetka I, Macellari V. Electrothrombosis of saccular aneurysms via endovascular approach. Part 1: Electrochemical basis, technique and experimental results. J Neurosurg 1991;75:1–7.[Medline]
  2. Missler U, Hundt C, Wiesmann M, Mayer T, Bruckmann H. Three-dimensional reconstructed rotational digital subtraction angiography in planning treatment of intracranial aneurysms. Eur Radiol 2000;10:564–8.[CrossRef][Medline]
  3. Moret J, Kemkers R, Op de Beek J, Koppe R, Klotz E, Grass M. 3D rotational angiography: Clinical value in endovascular treatment. Medicamundi 1998;42:8–14.
  4. Feldkamp LA, Davis LC, Kress JW. Practical cone-beam algorithm. J Opt Soc Am A 1984;1:612–19.[CrossRef]
  5. Anxionnat R, Bracard S, Ducrocq X, Trousset Y, Launay L, Kerrien E, et al. Intracranial aneurysms: clinical value of 3D digital subtraction angiography in the therapeutic decision and endovascular treatment. Radiology 2001;218:799–808.[Abstract/Free Full Text]
  6. International Commission on Radiological Protection. 1990 Recommendations of the International Commission on Radiological Protection, ICRP Publication 60. Pergamon Press, Oxford. Ann ICRP 1991;21
  7. Huda W, Lieberman KA, Chang J, Roskopf ML. Patient size and X-ray technique factors in head computed tomography examinations. I. Radiation doses. Med Phys 2004;31:588–94.[CrossRef][Medline]
  8. Cranley K, Gilmore BJ, Fogarty GWA, Desponds L, Sutton DG. Catalogue of diagnostic X-ray spectra and other data. IPEM Report 78 (CD-ROM). York: IPEM, 1997
  9. Cristy M. Mathematical phantoms representing children of various ages for use in estimates of internal dose, NUREG /CR-1159, ORNL/NUREG/TM-367. Oak Ridge National Laboratory, 1980
  10. Hart D, Jones G, Wall BF. Estimation of effective dose in diagnostic radiology from entrance surface dose and dose–area product measurements. NRPB-R262. Chilton: NRPB, 1994
  11. 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]
  12. Marshall NW, Noble J, Faulkner K. Patient and staff dosimetry in neuroradiological procedures. Br J Radiol 1995;68:495–501.[Abstract/Free Full Text]
  13. McParland BJ. A study of patient radiation doses in interventional radiological procedures. Br J Radiol 1998;71:175–85.[Abstract]
  14. Mooney RB, McKinstry CS, Kamel HAM. Absorbed dose and deterministic effects to patients from interventional neuroradiology. Br J Radiol 2000;73:745–51.[Abstract]



This article has been cited by other articles:


Home page
 RadiologyHome page
A. M. McKinney, C. L. Truwit, C. S. Palmer, M. Teksam, K. Papke, C. K. Kuhl, M. Fruth, C. Haupt, M. Schlunz-Hendann, D. Sauner, et al.
Intracranial Aneurysms: Is the Diagnostic Accuracy Rate of Multidetector CT Angiography Equivalent to That of Three-dimensional Rotational Conventional Angiography?
Radiology, March 1, 2008; 246(3): 982 - 984.
[Full Text] [PDF]

This Article
Right arrow Abstract Freely available
Right arrow Figures Only
Right arrow Full Text (PDF)
Services
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrow reprints & permissions
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Bridcut, R R
Right arrow Articles by Winder, R J
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Bridcut, R R
Right arrow Articles by Winder, R J

HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
BJR DMFR IMAGING  ALL BIR JOURNALS