You are here

Article Text

Article menu

Download PDFPDF

Neuroimaging
Original research
Practical techniques for reducing radiation exposure during cerebral angiography procedures
  1. Monica S Pearl1,2,3,
  2. Collin Torok3,
  3. Jiangxia Wang4,
  4. Emily Wyse1,
  5. Mahadevappa Mahesh3,
  6. Philippe Gailloud1,3
  1. 1Division of Interventional Neuroradiology, Johns Hopkins University School of Medicine, Baltimore, Maryland, USA
  2. 2Department of Interventional Neuroradiology, Children's National Medical Center, Washington, DC, USA
  3. 3Department of Radiology and Radiological Science, Johns Hopkins University School of Medicine, Baltimore, Maryland, USA
  4. 4Department of Biostatistics, Johns Hopkins Bloomberg School of Public Health, Baltimore, Maryland, USA
  1. Correspondence to Dr M S Pearl, Division of Interventional Neuroradiology, The Johns Hopkins Hospital, 1800 Orleans Street, Bloomberg Building, 7218, Baltimore, MD 21287, USA; msmit135{at}jhmi.edu

Abstract

Purpose DSA remains the gold standard imaging method for the evaluation of many cerebrovascular disorders, in particular cerebral aneurysms and vascular malformations. The purpose of this study was to demonstrate the effect of modifying DSA frame rate, fluoroscopic and roadmap pulse rates, and flat panel detector (FPD) position on the radiation dose delivered during routine views for a cerebral angiogram in a phantom model.

Materials and methods Adult skull and abdomen/pelvis anthropomorphic phantoms were used to compare the radiation dose metrics Ka,r (in mGy), PKA (in μGym2), and fluoroscopy time (in minutes) after modification of fluoroscopic pulses per second (p/s), DSA frames per second (f/s), and FPD position and collimation in three components of a cerebral angiogram: (1) femoral artery access, (2) roadmap guidance, and (3) biplane cerebral DSA.

Results For femoral artery access, DSA protocols resulted in significantly higher doses than those utilizing fluoroscopy alone (p=0.007). Roadmaps using 3 p/s or 4 p/s delivered significantly less dose than higher pulse rates (p=0.008). The ranges of delivered doses for biplane cerebral DSA were 347.3–1188.5 mGy and 3914.54–9518.78 μGym2. The lowest radiation doses were generated by the variable frame rate DSA protocols.

Conclusions Replacing femoral arterial access evaluations by DSA with fluoroscopy, utilizing lower pulse rates during fluoroscopy and roadmap guidance, and choosing variable frame rates for DSA are simple techniques that may be considered by operators in their clinical practices to lower radiation dose during cerebral angiography procedures.

  • Angiography

https://doi.org/10.1136/neurintsurg-2013-010982

Statistics from Altmetric.com

    Request Permissions

    If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.

    Introduction

    Although the use of non-invasive imaging modalities such as MRI and CT is increasing, cerebral DSA remains the gold standard technique for the evaluation of many cerebrovascular disorders, in particular cerebral aneurysms and vascular malformations. Patients harboring these pathologies may undergo multiple neuroangiographic procedures for diagnosis, treatment, and follow-up evaluation, and thus are potentially exposed to significant amounts of radiation. As the number and complexity of interventional neuroradiology procedures increases,1 so does the responsibility of each practitioner to investigate their own practices and to control unnecessary radiation exposure according to the ALARA principle.2 A number of metrics have been developed for measurement of radiation dose incurred during fluoroscopic procedures and include peak skin dose, reference air kerma (Ka,r), kerma area product (PKA), and fluoroscopy time.3

    At our institution, the patient dose delivered during neuroangiographic procedures is displayed on the monitors in the angiography suite and recorded in procedure reports using the radiation dose metrics Ka,r (in mGy), PKA (in μGym2), and fluoroscopy time (in minutes). Ka,r, also known as reference point air kerma and formerly called cumulative dose4 or cumulative air kerma, is the dose to air at the interventional reference point, which is 15 cm from the isocenter toward the x-ray tube. Ka,r is a cumulative approximation of the total radiation dose to the skin, summed over the entire procedure,3 and is a useful metric for the real time monitoring of doses incurred during fluoroscopic procedures and for monitoring of fluoroscopic sentinel events.5 The PKA, which reflects the total radiation energy entering the patient, is an indicator of stochastic risk for the patient, correlates with operator and staff dose, and is a surrogate measure for skin dose, although a limitation of this value is that the same PKA may be generated by a large dose delivered to a small skin area or by a small dose delivered to a large skin area. Fluoroscopy time is the simplest and most widely available yet least useful measure of patient dose.3 The purpose of this study was to demonstrate the effect of modifying fluoroscopic and roadmap pulse rates, DSA frame rate, and flat panel detector (FPD) position on the radiation dose delivered during routine views for a cerebral angiogram in a phantom model.

    Materials and methods

    Adult skull and abdomen/pelvis anthropomorphic phantoms were used to compare the radiation dose metrics Ka,r (in mGy), PKA (in μGym2), and fluoroscopy time (in minutes) after modification of fluoroscopic pulses per second (p/s), DSA frames per second (f/s), and FPD position and collimation in three components of a cerebral angiogram: (1) femoral artery access, (2) roadmap guidance, and (3) biplane DSA. The fluoroscopic settings of 73 kV, pulse width 12.5 ms, and DSA dose of 3.000 μGy/frame was the same for all protocols. All examinations were performed on a clinical biplane system (Artis Zee; Siemens Medical Solutions, Erlangen, Germany).

    Femoral artery access

    An abdomen/pelvis phantom was used to measure the radiation dose metrics Ka,r and PKA in a total of 12 femoral artery access protocols. All had 5 s of fluoroscopy (3 p/s, n=6; 4 p/s, n=6) to simulate visualization of an applied external marker over the femoral head, followed by five additional seconds of either fluoroscopy (3 p/s, n=2; 4 p/s, n=2) or DSA (2 f/s, n=4; 3 f/s, n=4). All protocols were performed in the single posteroanterior plane (phantom in the supine position) and each was performed in two different FPD configurations (A, B) (table 1). All protocols were performed with identical magnification (42 cm), flat panel angulation (0° left anterior oblique, 0° cranial), and table height (0 cm). The two FPD configurations differed in regards to the use of collimation (FPD configuration A: tight collimation applied; FPD configuration B: no collimation applied) and source to image detector distance (SID) (FPD configuration A: 110 cm; FPD configuration B: 120 cm). A two sample Wilcoxon rank sum test was used to compare the doses between femoral artery access protocols with and without DSA.

    View this table:
    Table 1

    Femoral artery access protocols

    Roadmap guidance

    An adult skull phantom was used to measure the radiation doses (Ka,r and PKA) in five single plane roadmap protocols, differing only by fluoroscopic pulse rate: 3, 4, 7.5, 10, and 15 p/s. All other parameters, including table height (−9 cm), FPD angulation (24° right anterior oblique, 10° caudal), SID (99 cm), and magnification (42 cm) were identical. All were performed in a single imaging plane (posteroanterior, phantom in the supine position), each for 5 s. A two sample Wilcoxon rank sum test was additionally used to compare the roadmaps using lower pulse rates versus the higher pulse rates.

    Cerebral DSA

    An adult skull phantom was used to measure the radiation doses (Ka,r and PKA) generated by eight biplane DSA imaging protocols, which varied by DSA frame rate and FPD configuration. Six of the eight protocols tested fixed DSA frame rates (2 f/s, 3 f/s, and 4 f/s, each n=2) and two of the eight protocols tested a variable frame rate DSA containing two phases, phase 1: 2 f/s for 4 s; phase 2: 1 f/s for the remainder of the acquisition, up to 28 s). Two FPD configurations were applied to each of the four types of DSA frame rate protocols, which differed in regards to the use of collimation (FPD configuration A: tight collimation applied; FPD configuration B: no collimation applied) and SID (FPD configuration A: variable per anatomic view, set as close to the phantom as possible; FPD configuration B: 120 cm). The biplane DSA protocols included neck as well as cranial frontal and lateral views to simulate bilateral carotid and vertebral acquisitions during a standard four vessel angiogram. All protocols had identical total DSA acquisition exposure times of 72 s and were performed with identical magnification, flat panel angulation, and table height, as specified for each anatomic view. The simulated cerebral angiogram (table 2) was performed for the fixed (2, 3, and 4 f/s) and variable DSA frame rates in both FPD configurations (A, B) for a total of eight DSA protocols. DSA settings were identical, including the use of automatic exposure control (AEC) and the dose per frame (3.000 μGy/frame). The Wilcoxon sign rank test was used to compare the doses by flat panel position.

    View this table:
    Table 2

    Anatomic views and biplane set up for simulated cerebral angiography

    Results

    Femoral artery access

    Eight of the 12 protocols (3A–6B) utilized DSA for femoral artery evaluation and generated average delivered doses of 1.1 mGy (range 0.8–1.9 mGy) and 24.6 μGym2 (range 6.7–40.2 μGym2). The remaining four protocols (1A–2B) used fluoroscopy alone for femoral artery evaluation without a DSA component and delivered average doses of 0.05 mGy (range 0–0.1 mGy) and 1.0 μGym2 (range 0.7–1.4 μGym2), an average of 95% less dose than those protocols utilizing DSA. The Wilcoxon rank sum test results demonstrated that the protocols with DSA resulted in significantly higher doses than those with fluoroscopy alone (p=0.007). All FPD configuration B protocols (1B–6B) generated doses greater than or equal to the corresponding FPD configuration A protocols (1A–6A) (figure 1).

    Figure 1
    Figure 1

    Radiation doses incurred during the eight femoral artery access protocols show significantly higher doses (p=0.007) when DSA was utilized for femoral artery access evaluation (protocols 3A–6B vs 1A–2B in (A) and (C)). All flat panel detector (FPD) configuration B protocols generated doses greater than or equal to the corresponding FPD configuration A protocols ((B) and (D)).

    Roadmap guidance

    A total of five protocols were evaluated, which varied only by fluoroscopic pulses per second. The radiation doses ranged from 0.2 to 0.9 mGy and from 4.9 to 22.5 μGym2. Roadmaps using 3 or 4 p/s generated significantly less dose than higher pulse rate protocols (p=0.008) by the Wilcoxon rank sum test.

    Cerebral DSA

    Eight biplane imaging protocols were created. The ranges of delivered doses were 347.3–1188.5 mGy and 3914.54–9518.78 μGym2. The greatest difference in dose was between the variable frame rate FPD configuration A (347.4 mGy) and the 4 f/s FPD configuration B (1188.5 mGy) protocols, in which the former delivered 59% less dose than the latter. The FPD configuration A consistently generated less radiation dose than FPD configuration B when comparing by DSA frame rate (p=0.07). The lowest radiation doses were generated by the variable frame rate DSA protocols (figure 2).

    Figure 2
    Figure 2

    Radiation doses incurred during the eight biplane cerebral DSA protocols depict the doses in mGy (A) and μGym2 (B). The lowest radiation doses were generated by the variable frame rate DSA protocols. When comparing doses by flat panel detector (FPD) configuration at each DSA frame rate (C), the FPD configuration A consistently generated less radiation dose than the FPD configuration B (p=0.07).

    Discussion

    Neurointerventional methods are being increasingly utilized to treat more complex lesions,6 and while interventional procedures typically expose patients to larger doses, diagnostic studies can also result in surprisingly high patient doses.2 Published radiation doses from diagnostic cerebral angiograms range from 350 to 4100 mGy.7–10 A number of factors contribute to this variability, including patient habitus and vascular anatomy, clinical indication, technical factors, and operator performance. Technical factors include the number of DSA acquisitions, DSA frame rate, DSA dose per frame, fluoroscopic and roadmap pulse rates, and image intensifier or FPD positioning and the use of collimation. The use of manufacturer specific dose saving techniques, such as automatic exposure control, the ability to store detector positioning, and the use of virtual markers on the screen to simulate location and collimators position without the need for fluoroscopic exposure, may also play a role.

    Operator related variables include experience, failure to stop fluoroscopy when there is no manipulation of the catheter or wire, and the operator's commitment to actively limiting radiation exposure. The physicians performing the examination must make a conscious effort to employ techniques aimed at reducing radiation dose where appropriate and reasonably achievable, and should be aware of the specifics of their angiographic equipment. Physicians should be cognizant of the default DSA settings, particularly delivered dose per frame, as this value is not displayed during the procedure, but rather can be found on inspection of the dose parameters on the console outside the angiography suite. Furthermore, set-up configurations vary among institutions and manufacturers, and should be reviewed after all software upgrades or services, as an operator's optimized settings may have been modified. Delivered radiation doses should be evaluated throughout each case, by looking for the displayed Ka,r (in mGy) and PKA (in μGym2), and after each case by examining the examination protocols in order to better understand one's practice and the delivered doses incurred during each DSA acquisition.

    As of 2011, there are no federal regulatory requirements in the USA concerning recording or reporting of radiation dose data for interventional procedures. Recently, however, the Standards of Practice Committee of the Society of Interventional Radiology has created guidelines for recording patient radiation exposure during fluoroscopic procedures.3 As radiation exposure data are collected and recorded, an analysis of one's practices can be made to guide improvements in clinician training and delivery of patient care. Optimizing operator technique is an important factor in order to reduce delivered radiation dose and cumulative exposure for patients and staff.

    Limitations of this study include analysis of radiation doses generated in a phantom model rather than from patient cases, and the exclusion of pediatric protocols and specific pediatric techniques (such as removing x-ray scatter grids). Despite these limitations, these data illustrate the impact on radiation dose by modifying the individual parameters of fluoroscopic pulses per second, DSA frame rate, and FPD position. Furthermore, these techniques are easy to incorporate into daily practice.

    Conclusion

    In this study, we have demonstrated the effect of modifying DSA frame rate, fluoroscopic and roadmap pulse rates, and FPD position on the radiation dose delivered during routine views for a cerebral angiogram in a phantom model. Based on our findings and their successful implementation into our own clinical practice, we suggest the following guidelines.

    1. Femoral artery access evaluation (in the setting of determining eligibility of a femoral closure device) should be performed with fluoroscopically acquired images (2 or 3 p/s), which can then be saved, rather than with a DSA acquisition. The addition of DSA can be helpful when a diagnostic question arises.

    2. Fluoroscopic and roadmap guidance for wire and catheter navigation should be performed with the lowest adequate pulse rate for any given patient and clinical situation— for example, 3 or 4 p/s—although tailoring of these parameters to address specific challenges of any given case are expected.

    3. DSA variable frame rates (eg, 2 f/s for 4 s during the arterial phase followed by 1 f/s for the venous phase) and fixed rates of 2 f/s rather than 3 f/s should be routinely considered. For high flow lesions (requiring greater hemodynamic stratification), faster DSA frame rates of 4–7.5 f/s may be utilized during selective angiography of the vessel(s) supplying the lesion. The frame rate should then be returned to the default lower frame rate for the remainder of the study.

    4. FPD positions and collimation during each fluoroscopic scene and DSA acquisition should be configured to reduce the size of the exposed field of view and decrease the SID.

    References

    1. ↵
      1. Gnanalingham KK,
      2. Apostolopoulos V,
      3. Barazi S,
      4. et al
      . The impact of the international subarachnoid aneurysm trial (ISAT) on the management of aneurysmal subarachnoid haemorrhage in a neurosurgical unit in the UK. Clin Neurol Neurosurg 2006;108:117–23.
      OpenUrlCrossRefPubMedWeb of Science
    2. ↵
      1. Alexander MD,
      2. Oliff MC,
      3. Olorunsola OG,
      4. et al
      . Patient radiation exposure during diagnostic and therapeutic interventional neuroradiology procedures. J Neurointerv Surg 2010;2:6–10.
      OpenUrlAbstract/FREE Full Text
    3. ↵
      1. Miller DL,
      2. Balter S,
      3. Dixon RG,
      4. et al
      . Quality improvement guidelines for recording patient radiation dose in the medical record for fluoroscopically guided procedures. J Vasc Interv Radiol 2012;23:11–18.
      OpenUrlCrossRefPubMedWeb of Science
    4. ↵
      International Electrotechnical Commission. Report 60601 medical electrical equipment–part 2–43: particular requirements for the safety of x-ray equipment for interventional procedures. Geneva 2000.
    5. ↵
      1. Vance AZ,
      2. Weinberg BD,
      3. Arbique GM,
      4. et al
      . Fluoroscopic sentinel events in neuroendovascular procedures: how to screen, prevent, and address occurrence. AJNR Am J Neuroradiol 2013;34:1513–15.
      OpenUrlAbstract/FREE Full Text
    6. ↵
      1. Bor D,
      2. Cekirge S,
      3. Turkay T,
      4. et al
      . Patient and staff doses in interventional neuroradiology. Radiat Prot Dosimetry 2005;117:62–8.
      OpenUrlAbstract/FREE Full Text
    7. ↵
      1. Mooney RB,
      2. McKinstry CS,
      3. Kamel HA
      . Absorbed dose and deterministic effects to patients from interventional neuroradiology. Br J Radiol 2000;73:745–51.
      OpenUrlAbstract
    8. ↵
      1. Struelens L,
      2. Vanhavere F,
      3. Bosmans H,
      4. et al
      . Skin dose measurements on patients for diagnostic and interventional neuroradiology: a multicentre study. Radiat Prot Dosimetry 2005;114:143–6.
      OpenUrlAbstract/FREE Full Text
    9. ↵
      1. Vano E,
      2. Gonzalez L,
      3. Fernandez JM,
      4. et al
      . Patient dose values in interventional radiology. Br J Radiol 1995;68:1215–20.
      OpenUrlAbstract/FREE Full Text
    10. ↵
      1. Theodorakou C,
      2. Horrocks JA
      . A study on radiation doses and irradiated areas in cerebral embolisation. Br J Radiol 2003;76:546–52.
      OpenUrlAbstract/FREE Full Text

    Footnotes

    • Scientific Paper presented at the 50th Annual ASNR Meeting, New York, NY, 26 April 2012.

    • Contributors All authors included in this manuscript have contributed appropriately and fulfill the criteria of authorship. No one else who fulfills the criteria has been excluded as an author.

    • Competing interests None.

    • Provenance and peer review Not commissioned; externally peer reviewed.

    • Data sharing statement Low dose three-dimensional DSA protocols, not investigated in this study, are available for data sharing. No other unpublished data from this study are available for data sharing.