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 Google Scholar Google Scholar Articles by Geijer, H Articles by Beckman, K-W Search for Related Content PubMed PubMed Citation Articles by Geijer, H Articles by Beckman, K-W

Radiation exposure in stent-grafting of abdominal aortic aneurysms

Departments of ¹ Radiology, ² Vascular Surgery and ³ Medical Physics, Örebro University Hospital, Örebro, SE-70185 Sweden

	Abstract

Top Abstract Introduction Material and methods Results Discussion References

 
In recent years, endovascular stent-grafting of abdominal aortic aneurysms has become more and more common. The radiation dose associated with these procedures is not well documented however. The aim of the present study was to estimate the radiation exposure and to simulate the effects of a switch from C-arm radiographic equipment to a dedicated angiographic suite. Dose–area product (DAP) was recorded for 24 aortic stent-grafting procedures. Based on these data, entrance surface dose (ESD) and effective dose were calculated. A simulation of doses at various settings was also performed using a humanoid Alderson phantom. The image quality was evaluated with a CDRAD contrast-detail phantom. The mean DAP was 72.3 Gy cm² at 28 min fluoroscopy time with a mean ESD of 0.39 Gy and a mean effective dose of 10.5 mSv. If the procedures had been performed in an angiographic suite, all dose values would be much higher with a mean ESD of 2.9 Gy with 16 patients exceeding 2 Gy, which is considered to be a threshold for possible skin injury. The image quality for fluoroscopy was superior for the C-arm whilst the angiographic unit gave better acquisition images. Using a C-arm unit resulted in doses similar to percutaneous coronary intervention (PCI). If the same patients had been treated using dedicated angiographic equipment, the risk of skin injury would be much higher. It is thus important to be aware of the dose output of the equipment that is used.

	Introduction

Top Abstract Introduction Material and methods Results Discussion References

 
In recent years, endovascular stent-grafting of abdominal aortic aneurysms has gained considerable attention because it appears to be a promising method of treating aneurysms [1]. One drawback compared with conventional surgery is the radiation exposure that patients and operators are exposed to during the procedure. This radiation exposure leads both to increased cancer risk and risk of skin injury [2]. However, the amount of exposure that the patients are exposed to is not well documented in the literature.

Often, the stent-grafting procedure is performed in an operating theatre using mobile C-arm equipment for X-ray guidance. It was thought that this equipment might lead to higher doses because of greater operational difficulties compared with dedicated angiographic equipment. Would a move to dedicated angiographic equipment simplify the procedure compared with the C-arm, and thus lead to lower doses?

The aims of this study were to document the radiation exposure associated with aortic stent-grafting. The radiation dose – image quality relationship for a mobile C-arm unit was also documented and compared with dedicated angiographic equipment.

	Material and methods

Top Abstract Introduction Material and methods Results Discussion References

 
The study consisted of two major parts: the first in which radiation dose data were recorded in a clinical stent-grafting series, and the second in which a phantom-based simulation of dose and image quality was performed using two different types of X-ray equipment. Furthermore, the cumulative effective dose was calculated for a typical patient undergoing a stent-graft procedure including the pre-operative imaging and 10 years of follow-up.

Patients
Patient dose data were collected from a series of stent-grafting procedures of the abdominal aorta, yielding 24 cases; 22 males (92%) and 2 females. The mean age was 71.8 years (standard deviation (SD) 7.2 years, range 59–85 years), mean height 174 cm (SD 8.6 cm), mean weight 80.5 kg (SD 12.8 kg). The mean aneurysm size was 6.2 cm (SD 1.6 cm).

All procedures were performed jointly by one vascular surgeon and one interventional radiologist. Informed consent was obtained for the stent-grafting procedure, but not separately for the dose measurement since this is routinely included as a part of the procedure.

Angiographic equipment
All clinical procedures were performed in an operating theatre using a BV-300 mobile C-arm unit (Philips Medical Systems, Best, The Netherlands). Available image intensifier field sizes are 17 cm, 23 cm and 31 cm and the filtration is 4 mm Al plus 0.1 mm Cu.

The X-ray tube has a fixed anode and is placed under the operating table for posteroanterior imaging. The system employs two radiation dose levels in imaging. The lower-dose level is hereafter called fluoroscopy while the higher-dose level is called digital acquisition. The fluoroscopy is continuous by default and the digital acquisition frame rate is 2 s^–1. The patient is placed on an operating table with a floating table-top (imagiQ; Stille Surgical, Solna, Sweden). In the C-arm unit, dose data are calculated from exposure values and are expressed as dose–area product (DAP). The calculated DAP values from the C-arm unit were calibrated using a transmission ion chamber connected to an electrometer (Doseguard 100; RTI Electronics, Mölndal, Sweden).

For a simulation of the dose changes that would occur if the procedures were to be performed with dedicated angiographic equipment, the dose–image quality relationship of an Integris V-3000 (Philips Medical Systems) angiographic unit was also investigated. This system uses a Super 100 CP generator with an MRC-GS tube. The image intensifier can be set between 17 cm and 38 cm field sizes of which 17 cm, 25 cm and 31 cm sizes were used to get sizes comparable with the C-arm unit. The tube and housing have a filtration of 2.5 mm Al. For fluoroscopy extra filtration of 1.3 mm Al and 0.6 mm Cu is introduced into the beam giving a total of 3.8 mm Al and 0.6 mm Cu. The fluoroscopy is continuous and the digital acquisition frame rate is 2 s^–1. For both units, all exposure values were set using standard clinical protocols as installed by the manufacturer.

Simulation
A simulation was performed on both the C-arm and the angiographic system using a humanoid Alderson Phantom Patient [3]. The phantom was exposed using all available image intensifier sizes as well as with a clinically suitable collimation, giving a field size of 20 cm from left to right. No collimation was used in the craniocaudal direction. DAP values were recorded using the Doseguard instrument. Entrance dose, free in air, was recorded simultaneously with a PMX-III instrument with a solid-state detector (RTI Electronics, Mölndal, Sweden). Posteroanterior projection was used with full AEC control. The source–image intensifier distance was 100 cm on both systems. Detailed data are shown in Table 1. The entrance dose values were converted to ESD using a backscatter factor of 1.4 considering the used field sizes and radiation qualities [4]. From these data it was possible to calculate the DAP/ESD relationship for all field sizes, both for fluoroscopy and digital acquisition.

The DAP connected to the C-arm gives separate output for fluoroscopy and for digital acquisition, and for the three image intensifier field sizes. It does not distinguish, however, how fluoroscopy and digital acquisition are divided into the various field sizes. For the dose simulations it was assumed that each field size had the same proportion of fluoroscopy to digital acquisition as the value for the entire study. Knowing the DAP for the clinical cases on the C-arm and the dose relationships between the C-arm and the angiographic unit from the phantom simulations, it was possible to calculate the theoretical DAP for each case, had it been performed using the angiographic system instead of the C-arm. This calculation was made assuming the same exposure time and the same proportions of digital acquisition and fluoroscopy as with the C-arm unit. Since measurement of a single image in digital acquisition mode is connected with a large uncertainty, measurements were acquired of 10 images plus the test shot before imaging starts. Since the pulse frequency was identical for both systems the results are comparable.

Image quality
Image quality was assessed for both systems using a CDRAD 2.0 contrast-detail phantom [5] (Instrumentele Dienst; Nijmegen, The Netherlands). The phantom consists of an acrylic sheet with drilled holes of different depth and diameter. The more holes that can be seen, the better the image quality. From the resulting data, a numerical value, IQF (Image Quality Figure) can be calculated. These values can be used to evaluate the contrast-detail quality of the imaging chain. The phantom was placed in the middle of a 19 cm thick slab of PMMA to simulate the scattering conditions of the human body. This thickness corresponds to the attenuation of the Alderson phantom patient. The phantom was evaluated during live fluoroscopy. The digital acquisition images were read from a PACS display (Sectra/Imtec; Linköping, Sweden) by one radiologist. Images were also acquired of the Alderson phantom patient for illustration purposes.

Calculation of effective dose for pre-operative and follow-up studies
The Eurostar registry has suggested a time schedule for pre-operative and follow-up studies [6]. These are shown in Table 2. Our calculations are based on a Secura CT scanner (Philips Medical Systems), which is a single-slice spiral scanner, using the settings in clinical use in our institution. The pre-operative CT study was performed with one series with 10 mm slices at 120 kV and 180 mA pre contrast and one series with 3 mm slices at 120 kV and 220 mA post contrast, both series at pitch 1.5. The post-operative CT studies were performed with one series post contrast with 5 mm slices at 120 kV and 200 mA, pitch 2. Based on these data, effective dose was calculated using the ImPACT CT Patient Dosimetry calculator version 0.99s [7] together with Monte Carlo dose data sets published in report NRPB-SR250 [8].

	Results

Top Abstract Introduction Material and methods Results Discussion References

 
19 patients received an Excluder stent-graft (Gore; Flagstaff, AZ) and five received a PowerLink stent-graft (Endologix; Nieuwegein, The Netherlands). One patient died on the operating table immediately after a successful procedure due to a myocardial infarction. One patient underwent a transfemoral re-intervention after 1 month due to an attachment site (type I) endoleak. Five patients had local complications in the groin or leg such as pseudoaneurysm, haematoma or thrombosis. At follow-up, four patients were shown to have a collateral re-perfusion (type II) endoleak. Except for the complications in the immediate post-operative period, no other complications caused re-operations during a follow-up period of average 10.3 months (SD 4.9 months).

The mean DAP for this group was 72.3 Gy cm², divided about equally into fluoroscopy and digital acquisition. Dose data are shown in detail in Table 3.

Table 1 shows tube potential, DAP and ESD values with an Alderson phantom for the two types of equipment separated into fluoroscopy and digital acquisition.

Collimation to a field size of 20 cm left to right at the image intensifier reduced the DAP for an Alderson phantom from 44.8 mGy cm² s^–1 to 36.7 mGy cm² s^–1 (18%) at the largest field size, 31 cm, for the C-arm and from 26.1 mGy cm² s^–1 to 23.1 mGy cm² s^–1 (12%) for the angiographic equipment. At the middle field size, 23 cm for the C-arm, there was hardly any dose reduction since the collimators were barely visible in the image.

Image quality results with a CDRAD phantom are shown in Figure 1 with corresponding images of an Alderson phantom in Figure 2.

View larger version (21K): [in this window] [in a new window]   Figure 1. Image Quality Figure (IQF) with a CDRAD phantom, fluoroscopic images viewed during live fluoroscopy and digital acquisition images viewed as still images. Lower IQF values indicate better image quality.

View larger version (127K): [in this window] [in a new window]   Figure 2. Sample images with an Alderson phantom. (a) C-arm unit, 31 cm field size. (b) Angiographic equipment, 31 cm field size. (c) C-arm unit, 17 cm field size. (d) Angiographic equipment, 17 cm field size.

	Discussion

Top Abstract Introduction Material and methods Results Discussion References

 
The mean DAP for the stent-grafting procedures was 72 Gy cm², which is comparable with percutaneous coronary intervention (PCI) [9–11]. The skin dose expressed as ESD is quite low, and for all cases well below the level of 2 Gy at which skin erythema might be seen [2]. One reason for the low ESD is the heavy filtration of the beam in the C-arm unit. Another reason might be the low output of the mobile C-arm unit which simply makes it impossible to use any high dose rates. This means that the risk of skin injury is low using the equipment and conditions described in this paper. The mean ESD is also similar to that obtained by Lipsitz et al [12], in which C-arm units were also used.

Radiation exposure is known to cause cancer and also leads to acute skin injury. The patients subjected to aortic stent-grafting are in general in the higher age range (mean age 72 years in the present study), which means that the cancer risk is not the greatest concern considering the usual latency time of 10–20 years after exposure. Instead, the major radiation risk is acute skin injury, which is covered in ICRP report 85 [2]. This type of injury is usually seen a few weeks after a procedure when the skin dose at a point exceeds about 2 Gy.

On the other hand, if the procedures had been performed using the dedicated angiographic equipment with its factory settings, the DAP values would be about eight times as high and the ESD values about seven times. 16 out of 24 patients in the clinical series would have exceeded the ESD limit of 2 Gy and thus be at risk for acute skin injury. The risk would probably be quite high since the irradiated field on the back of the patient does not change much during the procedure and the dose load is thus not divided over a large skin area. The high dose for the angiographic equipment is mainly because of the tube current and low filtration used for digital acquisition. In digital subtraction angiography (DSA), the dose must be high to get a low quantum noise for the subtraction, otherwise the image quality would be too low. It is possible to reduce the dose in digital acquisition if a lowered image quality is accepted, but this is not a trivial task with these highly computerized systems and it requires close cooperation between operators, physicists and engineers. In fluoroscopy, a heavy filtration including copper is introduced into the beam, which gives a lower dose rate than the C-arm system. The heavy filtration reduces the dose but also reduces contrast. The operator can choose between three different dose settings in fluoroscopy, but this study has only evaluated the setting with the lowest dose, which is the setting that is used in daily clinical practice. However, it must be remembered that these figures are estimations and not measurements on real patients. This estimation was made on the assumption that the amount of fluoroscopy and imaging would be unchanged, which is not necessarily the case. Perhaps equipment which is easier to operate would lead to a reduction of fluoroscopy time. Still, these results indicate that it is extremely important to be aware of the dose output for the equipment that is used for these high-dose procedures, which, indeed, are also performed in angiographic suites [13]. We do not know if these results are transferable to other brands of angiographic units, but we do suspect that they are. To be certain, measurements must be made on each type of equipment.

In interventional cardiology an entity called "DAP trigger level" of 300 Gy cm² has been proposed by Neofotistou et al [14]. This trigger level indicates the DAP level at which there is a risk for possible skin injury. In aortic stent-grafting, this trigger level would probably be lower since, compared with interventional cardiology, the exposure is concentrated on a smaller area of the skin, which means that it would be easier to get a skin injury from a certain DAP level.

The ESD and effective doses were calculated entities based on phantom simulations and not measured. The relationship between ESD and DAP is not trivial and depends on a number of factors. If collimation is used in the clinical cases and not in the simulation, the ESD will be underestimated. However, collimation was hardly used at all in the clinical series because of cumbersome handling of the controls on the C-arm unit. On the other hand, the calculated ESD can be seen as a maximum value assuming no movement of the irradiated field. In practice there is at least some variation of the field during a clinical study, and this would reduce the risk of skin injury. The source–object distance also influences the ESD. We took great care to position the Alderson phantom that was used in the simulations in the same position that the clinical patients had done. Of course, exposure factors influence both DAP and ESD. In all imaging, AEC control was used. Of course, the results of such a simulation must be interpreted with some caution. Still, we believe they are reasonably accurate.

In the RAD-IR study Miller et al [15, 16] give extensive skin dose data for a multitude of interventional procedures, unfortunately not including aortic stent-grafting. Certain procedures that were associated with high overall patient skin doses were renal/visceral angioplasty with stent placement, transjugular intrahepatic portosystemic shunt (TIPS) creation, and embolisation procedures of all kinds. It is not unreasonable to assume that aortic stent-grafting might also fit into this category since it includes both embolisation procedures such as closing an internal iliac artery before stent-grafting, and stent placement in the abdomen, which is known to lead to much higher doses than procedures in the extremities [17].

Measurements of DAP and ESD with an Alderson phantom (Table 1) show a well-known relationship: when reducing the image intensifier field size the DAP is reduced (or even unaffected with some types of equipment) while the ESD increases. This is because the equipment is set to increase tube output with the smaller field sizes to maintain a constant quantum noise level. This also means that it is easier to reach a high skin dose (ESD) with a small field size such as is used quite frequently, at least in our institution, when checking for endoleaks in the upper part of the stent-graft.

Collimation is an effective way to reduce both DAP and scattered dose to the operators. Collimation also increases image quality since the scattered radiation decreases. It has to be balanced by the reduced field-of-view that results. In clinical practice and in this study, collimation has hardly been used at all. The major reason for this is that it is very cumbersome to adjust the C-arm unit since the controls have to be operated by an assistant who is not dressed in sterile clothing. Collimation would probably be easier to use with dedicated angiographic equipment because there the controls are usually mounted at the table-side. A clinically acceptable collimation would reduce DAP by about 20% for the larger field sizes, but not change the ESD. The risk of skin injury would thus not be reduced, however the area at risk would be smaller because of the smaller field size.

Continuous fluoroscopy was used in the study since this is what was available with the two systems. Pulsed fluoroscopy would reduce the fluoroscopy dose further and possibly increase image quality [18, 19].

The image quality as studied with the CDRAD phantom (Figure 1) yielded results largely proportional to the dose settings. The digital acquisition settings had the highest image quality, which is natural considering the higher doses. The difference in image quality between the C-arm and the angiographic equipment is, however, rather small considering the very large difference in dose. On the other hand, fluoroscopic image quality is surprisingly good especially for the C-arm unit, even if the comparison of image quality between a static digital acquisition image and live fluoroscopy must be made with caution, since in fluoroscopy the human visual system makes a temporal integration which increases the perceived image quality. The angiographic unit had the lowest image quality for fluoroscopy, which might be a result of the heavy filtration of the beam which reduces image contrast. The image quality was further reduced at the largest field sizes, which might be due to the lower magnification.

This obviously raises the question: which image quality is best suited for the procedure? The fluoroscopic image quality is lowest for the angiographic system, but the dose and image quality can easily be increased with the push of a button. The operator can therefore obtain the fluoroscopic image quality that is needed for the moment. The image quality for digital acquisition is sometimes felt to be too low with the C-arm. This is partly due to problems with lighting and optimal placement of the viewing monitors in the crowded operating theatre. On the other hand, the acquisition dose for the angiographic system is so high that it probably should be reduced before performing this type of procedure. We have carried out informal tests wherein we reduced the dose request (and the dose) for the system to half of the original. The resulting images are noisier than the original, but probably of acceptable image quality.

The cumulative effective dose of 115 mSv for all radiographic studies including 10 years of follow-up CT studies is high. This is not surprising considering the large number of studies, 15 CT studies and eleven abdominal X-ray studies plus the stent-grafting procedure. Unlike angiography, CT does not lead to acute skin injury since the skin dose is spread evenly around the patient. Using a common risk estimation of 5% risk of lethal cancer per Sv exposure [20], the dose received during 10 years of follow-up would incur a risk of cancer development of about 1/200 for a middle-aged patient, probably less in older patients as in the present study. This risk must be balanced with the risk of aneurysm rupture if untreated and the operative risk if operated on with the conventional open method. If angiography is included in the pre-operative investigation, it would contribute to the cumulative effective dose, with around 5 mSv.

The mean follow-up time of less than a year is short, and too short to draw any conclusions concerning clinical results, but this was not the aim of the study. Regarding complications such as endoleaks, extra diagnostic angiographies might be needed as well as re-interventions, which would both lead to increased dose. Registry data have shown a transfemoral re-intervention rate of 14% during a mean follow-up of 20 months [21]. The collateral perfusion (type II) endoleaks did not cause secondary interventions due to a less aggressive approach towards this type of endoleaks in our institution [22].

In summary, this study shows that a clinical series of abdominal aortic stent-grafting procedures using a C-arm unit results in dose values in the same dose range as coronary intervention. If the same patients had been treated using dedicated angiographic equipment with higher radiation output, the risk of skin injury would be much higher. It is thus of the utmost importance to be aware of the X-ray output of the equipment in use and the many factors upon which this depends.

Received for publication September 21, 2004. Revision received February 9, 2005. Accepted for publication May 4, 2005.

	References

Top Abstract Introduction Material and methods Results Discussion References

 

1. Greenhalgh RM, Brown LC, Kwong GP, Powell JT, Thompson SG. Comparison of endovascular aneurysm repair with open repair in patients with abdominal aortic aneurysm (EVAR trial 1), 30-day operative mortality results: randomised controlled trial. Lancet 2004;364:843–8.[CrossRef][Medline]
2. ICRP. Avoidance of radiation injuries from medical interventional procedures. ICRP Publication 85. Oxford: Pergamon Press; 2000;30 (2).
3. ICRU. Phantoms and computational models in therapy, diagnosis and protection. ICRU Report 48. Bethesda, MD: International Commission on Radiation Units and Measurements, 1992;48.
4. Wall BF, Harrison RM, Spiers FW. Patient dosimetry techniques in diagnostic radiology. Institute of Physical Sciences in Medicine, Report No. 53. York: IPEM, 1988.
5. Thijssen MAO, Thijssen HOM, Merx JL, Lindeijer JM, Bijkerk KR. A definition of image quality: the image quality figure. In: Moores BM, Wall BF, Eriskat H, Schibilla H, editors. BIR Report 20: Optimization of image quality and patient exposure in diagnostic radiology. London: The British Institute of Radiology; 1989: 29–34.
6. Eurostar. Stent-graft techniques for abdominal aortic aneurysm repair. European collaborators group. Case record form. 2000.
7. Impactscan.org [homepage on the Internet]. London: ImPACT group [updated 07/05/2003, cited 2003 Sep 12]. Available from http://www.impactscan.org.
8. Jones DG, Shrimpton PC. Normalised organ doses from X-ray computed tomography calculated using Monte Carlo techniques. NRPB SR-250. Chilton: National Radiological Protection Board, 1993.
9. Geijer H, Beckman K-W, Andersson T, Persliden J. Radiation dose optimization in coronary angiography and percutaneous coronary intervention (PCI). II. Clinical evaluation. Eur Radiol 2002;12:2813–9.[Medline]
10. Bernardi G, Padovani R, Morocutti G, Vañó E, Malisan MR, Rinuncini M, et al. Clinical and technical determinants of the complexity of percutaneous transluminal coronary angioplasty procedures: analysis in relation to radiation exposure parameters. Catheter Cardiovasc Interv 2000;51:1–9.[CrossRef][Medline]
11. Fransson SG, Persliden J. Patient radiation exposure during coronary angiography and intervention. Acta Radiol 2000;41:142–4.[CrossRef][Medline]
12. Lipsitz EC, Veith FJ, Ohki T, Heller S, Wain RA, Suggs WD, et al. Does the endovascular repair of aortoiliac aneurysms pose a radiation safety hazard to vascular surgeons? J Vasc Surg 2000;32:704–10.[CrossRef][Medline]
13. Balzer JO, Doss M, Thalhammer A, Fieguth HG, Moritz A, Vogl TJ. Urgent thoracic aortal dissection and aneurysm: treatment with stent-graft implantation in an angiographic suite. Eur Radiol 2003;13:2249–58.[CrossRef][Medline]
14. Neofotistou V, Vañó E, Padovani R, Kotre J, Dowling A, Toivonen M, et al. Preliminary reference levels in interventional cardiology. Eur Radiol 2003;13:2259–63.[CrossRef][Medline]
15. Miller DL, Balter S, Cole PE, Lu HT, Schueler BA, Geisinger M, et al. Radiation doses in interventional radiology procedures: the RAD-IR study: part I: overall measures of dose. J Vasc Interv Radiol 2003;14:711–27.[Medline]
16. Miller DL, Balter S, Cole PE, Lu HT, Berenstein A, Albert R, et al. Radiation doses in interventional radiology procedures: the RAD-IR study: part II: skin dose. J Vasc Interv Radiol 2003;14:977–90.[Medline]
17. Ramalanjaona GR, Pearce WH, Ritenour ER. Radiation exposure risk to the surgeon during operative angiography. J Vasc Surg 1986;4:224–8.[CrossRef][Medline]
18. den Boer A, de Feyter PJ, Hummel WA, Keane D, Roelandt JRTC. Reduction of radiation exposure while maintaining high-quality fluoroscopic images during interventional cardiology using novel X-ray tube technology with extra beam filtering. Circulation 1994;89:2710–4.[Abstract/Free Full Text]
19. Wagner LK, Archer BR, Cohen AM. Management of patient skin dose in fluoroscopically guided interventional procedures. J Vasc Interv Radiol 2000;11:25–33.[Medline]
20. ICRP. 1990 recommendations of the International Commission on Radiological Protection. ICRP Publication 60. Oxford: Pergamon Press; 1991;21(1–3).
21. Laheij RJ, Buth J, Harris PL, Moll FL, Stelter WJ, Verhoeven EL. Need for secondary interventions after endovascular repair of abdominal aortic aneurysms. Intermediate-term follow-up results of a European collaborative registry (EUROSTAR). Br J Surg 2000;87:1666–73.[CrossRef][Medline]
22. van Marrewijk CJ, Fransen G, Laheij RJF, Harris PL, Buth J, and for the Eurostar collaborators. Is a type II endoleak after EVAR a harbinger of risk? Causes and outcome of open conversion and aneurysm rupture during follow-up. Eur J Vasc Endovasc Surg 2004;27:128–37.[CrossRef][Medline]

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 Google Scholar Google Scholar Articles by Geijer, H Articles by Beckman, K-W Search for Related Content PubMed PubMed Citation Articles by Geijer, H Articles by Beckman, K-W