Author + information
- Received June 27, 2016
- Revision received October 4, 2016
- Accepted November 17, 2016
- Published online March 6, 2017.
- Kenneth Fetterly, PhDa,∗ (, )
- Beth Schueler, PhDb,
- Michael Grams, PhDc,
- Glenn Sturchio, PhDd,
- Malcolm Bell, MDa and
- Rajiv Gulati, MDa
- aDepartment of Cardiovascular Diseases, Mayo Clinic, Rochester, Minnesota
- bDepartment of Radiology, Mayo Clinic, Rochester, Minnesota
- cDepartment of Radiation Oncology, Mayo Clinic, Rochester, Minnesota
- dDepartment of Diagnostic Radiology, Mayo Clinic, Jacksonville, Florida
- ↵∗Address for correspondence:
Dr. Kenneth Fetterly, Mayo Clinic, Cardiovascular Diseases, 200 First Street SW, Rochester, Minnesota 55905.
Objectives The first aim of this study was to assess the magnitude of radiation dose to tissues of the head and neck of physicians performing x-ray-guided interventional procedures. The second aim was to assess protection of tissues of the head offered by select wearable radiation safety devices.
Background Radiation dose to tissues of the head and neck is of significant interest to practicing interventional physicians. However, methods to estimate radiation dose are not generally available, and furthermore, some of the available research relating to protection of these tissues is misleading.
Methods Using a single representative geometry, scatter radiation dose to a humanoid phantom was measured using radiochromic film and normalized by the radiation dose to the left collar of the radioprotective thorax apron. Radiation protection offered by leaded glasses and by a radioabsorbent surgical cap was measured.
Results In the test geometry, average radiation doses to the unprotected brain, carotid arteries, and ocular lenses were 8.4%, 17%, and 50% of the dose measured at the left collar, respectively. Two representative types of leaded glasses reduced dose to the ocular lens on the side of the physician from which the scatter originates by 27% to 62% but offered no protection to the contralateral eye. The radioabsorbent surgical cap reduced brain dose by only 3.3%.
Conclusions A method by which interventional physicians can estimate dose to head and neck tissues on the basis of their personal dosimeter readings is described. Radiation protection of the ocular lenses by leaded glasses may be incomplete, and protection of the brain by a radioabsorbent surgical cap was minimal.
The potential for adverse health effects from occupational radiation exposure is of concern for interventional cardiologists, radiologists, and surgeons who routinely use x-ray fluoroscopy and angiography to diagnose and treat cardiac and vascular disease (1–4). Interventional physicians have an appetite for information that can help them assess their own health risks associated with occupational radiation dose. One aim of this work is to provide physicians a simple means to estimate radiation dose to tissues of the head, particularly the brain, carotid arteries, and ocular lenses.
Methods to minimize radiation dose to tissues of the head are also desired by interventional physicians to mitigate risk for radiation injury. Appropriate x-ray shielding remains one of the fundamental ways to protect physicians from x-ray scatter. The potential for a ceiling-mounted upper body shield to protect the head and neck of physicians is well known (5–8). Other novel shielding devices (9–11) have not been widely adopted. The potential for leaded glasses to protect the eyes and thereby reduce cataract risk has long been known (12). The remainder of the head and superior portion of the neck frequently do not have dedicated radiation protection. Radioprotective surgical caps have been proposed to reduce dose to the brain (13–15). The second aim of this work was to assess the potential for select wearable radiation protection devices to reduce dose to tissues of the head, particularly the brain and ocular lenses, and provide interventional physicians practical guidance in the effectiveness of these devices.
Primary experimental methods used for this work were described elsewhere (Fetterly et al. ). Therefore, only a brief summary of those methods and relevant enhancements to support this work are included here. Preliminary experiments characterized the energy of scattered radiation emitting from a patient (phantom) during an x-ray angiographic procedure. Four different scatter beam qualities were selected to mimic scatter associated with patient sizes ranging from a small child to a large adult. A standard x-ray tube was tuned to mimic the 4 scatter beam qualities by adjusting the peak tube potential (56, 74, 90, and 106 kVp) and half-value layer (3.5, 4.5, 5.5, and 6.5 mm Al) of the beam. The scatter-equivalent beams were directed upon an anthropomorphic phantom (Alderson, RANDO phantom, Radiology Support Devices, Long Beach, California) to estimate scatter radiation dose to the head and neck of an interventional physician (Figure 1). The transverse plane slabs of the phantom were assembled on a table to simulate an upright physician. Radiochromic film (XR-QA2, Ashland, Bridgewater, New Jersey) was placed within transverse planes of the phantom to measure tissue dose. The scatter-equivalent radiographic x-ray beam and phantom were positioned and oriented to mimic scatter incident upon a physician performing a left femoral access cardiac interventional procedure. The phantom was covered with a 0.5-mm lead-equivalent radioprotective apron, and 3 × 3 cm2 pieces of radiochromic film were attached to the outside of the lead apron at locations to mimic a personal dosimeter located at the left and right collars (Figure 1). Our previous work reported the influence of scatter beam quality or energy on the percentage of left collar dose (LCD) (16). This work reports a composite or typical dose that is the average percentage LCD associated with the various scatter beam qualities.
A single scatter-equivalent x-ray beam (90 kVp; half-value layer 5.5 mm Al) was used to assess select radiation safety devices, including 2 types of radioprotective glasses (Figure 2) (glasses 1: 0.75-mm lead equivalent, Liberty MX30, Phillips Safety Products, Middlesex, New Jersey; glasses 2: 0.07-mm lead equivalent, XR-700, Toray Medical, Toray International America, Houston, Texas), a commercially available radiopaque surgical cap (No-Brainer 9100, RadPad Protection, Kansas City, Kansas), and a radiopaque hood fabricated in our laboratory from the material of the radiopaque surgical cap (Figure 3). The surgical cap was positioned on the phantom such that the inferior portion of the cap was in contact with the auricles of the ears and just superior to the location corresponding to the superciliary arches and eyebrows. Similar to that described by Kuon et al. (7), the hood was fashioned to hang along the side of the face of the phantom, around to the back of the neck, and extend from the surgical cap inferiorly to the level of the phantom chin (Figure 3). Anatomic regions protected by the wearable shielding devices were represented by a region of x-ray shadow recorded in the radiochromic film. The reduction in radiation dose to tissues protected by the devices was estimated by comparing radiochromic film dose measurements with and without the protective devices.
After exposure, the radiochromic films were scanned with a flatbed scanner, resulting in 2-dimensional (2D) transverse plane dose distribution maps. The 2D dose maps were normalized by the dose received by the LCD, resulting in 2D maps of dose as a percentage of the LCD. Computed tomographic images of the phantom were overlain onto the 2D dose maps to provide bony landmarks. Dose to select organs and tissues, including the left and right brain, whole brain, brain stem, ocular lenses, and carotid arteries, was calculated.
Percentage of LCD maps corresponding to the 90-kVp scatter-equivalent beam are shown in Figure 4. The color scale of Figure 4 ranges from 0% LCD (dark blue) through 100% LCD (yellow). The percentage LCD decreased rapidly from the anterior left (x-ray beam entrance) surface of the phantom and was consistent with expectations of exponential attenuation of the x-ray beam by the tissue-equivalent phantom. Also, the highly attenuating cranial and facial bones protect underlying tissues from radiation. The x-ray shadow from the bones of the phantom contributes detail in the 2D dose maps corresponding to regions of the otherwise homogenous soft tissue. The steep dose gradient in the posterior region of Figure 4F was due to the x-ray shadow cast by the protective apron. Table 1 provides percentage LCD tissue doses to select organs and tissues calculated as the average of the 4 scatter beam qualities. Tissues that are relatively superficial or left sided received greater dose than tissues that are deep and relatively right sided.
Anatomic regions protected by the radiopaque surgical cap are demonstrated in Figure 5. The transverse planes included in Figure 5 represent the 3 most superior planes of the phantom and correspond to planes represented in Figures 4A to 4C. Phantom levels inferior to those shown in Figure 5 (Figures 4D to 4F) were unprotected by the cap. Because the x-ray scatter was incident upon the phantom head from an inferior elevation, the x-ray shadow cast by the cap propagated tangential to the superior aspect of the head rather than over the brain cavity. Therefore, the radiopaque surgical cap reduced dose to only a small volume of tissue within the left anterior-superior region of the cranium. The attenuating material of the cap reduced radiation to tissues covered by the x-ray shadow that it cast by approximately 65% to 70%. However, because the shadow cast by the cap covered only a small volume of the brain and surrounding tissue, the cap provided only 4.9% protection to the left brain, 1.8% protection to the right brain, and 3.3% protection to the whole brain.
Anatomic regions protected by the hood fabricated from the material of the surgical cap are demonstrated in Figure 6. The x-ray shadow cast by the radiopaque hood covered approximately the posterior half of the left brain at the level of the eye brows (Figure 5C) and extended to cover portions of the right brain at levels more superior (Figures 5A and 5B). Measurements demonstrated that this hood offered 70% protection to the left brain, 49% protection to the right brain, and 55% protection to the whole brain.
Protection of the ocular lenses provided by the glasses is demonstrated in Figure 7. For the single geometry tested, glasses 1 reduced percentage LCD to protected tissues by 77%. However, the protective shadow of glasses 1 provided incomplete coverage of the left ocular lens (Figure 7A). Therefore, measured protection of the left lens was 27%, and the right eye was entirely unprotected. Glasses 2 reduced percentage LCD to protected tissues by 62%. Glasses 2 cast a shadow over the entire left ocular lens (Figure 7B), resulting in a higher protection level compared with glasses 1, despite the lower lead-equivalent thickness (0.07 mm for glasses 2 vs. 0.75 mm for glasses 1). Measured protection of the left lens was 61%, but the right eye was also unprotected by glasses 2.
This work describes a straightforward, novel model for accurate assessment of scatter radiation dose to tissues of the head and neck of interventional physicians. By normalizing tissue dose by the dose to the location of the left collar, this work facilitates individualized estimates of dose to head and neck tissues of physicians on the basis of their personal dosimeter readings. In contrast to other studies (13–15), these findings demonstrated that radiation absorbent surgical caps provide minimal brain protection. Furthermore, these measurements affirm that radiopaque eyeglasses provide incomplete protection of the ocular lenses.
A primary purpose of this work was to allow the estimation of interventional physicians’ head and neck tissue dose from personal dosimeter readings. Provided that the spatial relationship between the physician and patient is similar to that used here, tissue dose to a physician may be estimated by multiplying his or her personal dosimeter reading by the percentage LCD presented in Table 1. More detailed results were previously published (16). This work assumes that the personal dosimeter is worn on the collar of the same side from which the scatter originates. If the scatter radiation is incident from the contralateral side, physician tissue dose can be estimated as the product of the personal dosimeter reading times percentage LCD times f, where f = 1.31 is the ratio of the LCD to the right collar dose determined by previous experiments (16). The percentage LCD values reported here agree well with similar measurements reported by Marshall et al. (17), wherein dose to select organs was normalized by dose to the bridge of the nose.
Whereas the spatial relationship between the physician and patient is variable in practice, a limitation of this study is that the protection offered by the wearable devices was assessed for only a single experimental geometry. Therefore, the results presented herein should not be considered absolute but should rather be considered a guide to the general magnitude and distribution of radiation dose to a physician’s head and neck from scatter originating from patients. Also, this experiment did not account for protection offered by other radiation protective devices, including a radioprotective thyroid collar, the upper body shield, or radioabsorbent drapes placed on the patient. Finally, only 2 models of leaded glasses were assessed. Future work to more completely describe the influences and dependencies of these practical variables is warranted.
There is agreement in the published research and among these authors that radiation protective glasses should be worn by operators of x-ray fluoroscopic systems (18). However, similar to the findings of Geber et al. (19), the results of this work indicate incomplete protection of the ocular lenses by the glasses. In particular, neither model of glasses protected the ocular lens contralateral to scatter source. This is because dose to the contralateral lens enters the eye obliquely through the face, whereas the glasses preferentially protect against radiation incident from the front. Further work to assess the real-world implications of these findings and the influence of glasses design on ocular lens protection is warranted.
This work demonstrates that the radiation absorbing surgical cap provides essentially no protection to the brain of an interventional physician. This result is readily explained by geometry. Because the scatter originates from a location inferior to the physician’s head, the x-ray shadow cast by the cap propagates nominally tangential to the superior aspect of the skull rather than over the cranial cavity. This finding contradicts marketing materials and published works suggesting that these or similar types of caps offer substantial brain protection (13–15). The experimental methods used in these other works measured attenuation of the surgical cap rather than dose to tissues. As demonstrated herein, the optimistic perspective that radiopaque surgical caps substantially reduce brain dose is misleading.
Although this work demonstrates limitations of selected wearable radiation safety devices, there are several established methods to facilitate head and neck dose reduction in interventional laboratories. Because scatter radiation is directly proportional to the radiation dose delivered to the patient, methods to reduce radiation dose to the patient can be expected to have a commensurate effect on physician dose (20). The x-ray shadow cast by a radiopaque thyroid collar, typically worn in conjunction with a lead apron or vest, can be expected to provide substantial protection of the inferior portion of the neck. The merit of a well-positioned ceiling-mounted upper body shield has been described (5–8). Radiopaque drapes placed strategically upon the patient have been shown to reduce physician dose (21–23). The potential benefit of the drapes is greatest when the upper body shield cannot be used and is dependent upon their design and strategic placement upon the patient (8,22,23) Marshall et al. (17) reported 81% brain dose reduction associated with a lead-acrylic face mask. The x-ray shadow cast by the face mask could be expected to provide similar protection to all head tissues. Kuon et al. (7) described the potential use of a lead-equivalent hood. Novel radiation shielding systems that facilitate whole-body protection can be expected to provide good protection of the head and neck (8–10) and may offer the user the ergonomic benefit of not wearing lead garments. Finally, robotic interventional systems may facilitate substantial physician dose reduction for select procedures (24,25).
The magnitude of brain dose from scatter is of interest to practicing physicians. Following is an example of how the percentage LCD values presented here can be used to estimate brain dose and then applied to facilitate a discussion of associated tumor risk. An interventional physician who receives the U.S. (European) recommended annual maximum effective dose of 50 mSv (20 mSv) may have a personal dosimeter reading of approximately 167 mSv (67 mSv) (26–28). For this example, assume that the annual left collar dosimeter reading of an interventional physician is 50 mSv · year−1 (5,000 mrem · year−1). On the basis of the measured 8.4% of LCD in the present study, the average dose to the whole brain would be 4.2 mSv · year−1 (50 mSv · year−1 × 8.4%).
A recent summary of interventional physicians who developed brain cancer has raised awareness of the potential for radiation risk (2). However, studies demonstrating a causal effect between radiation of adults and brain tumor risk are sparse (29). Preston et al. (30) described a causal relationship between brain dose and neurological tumors in atomic bomb survivors. Brain dose to the cohort of 80,160 people in the study ranged from <0.005 to >1 Sv. The natural incidence rate of fatal neurological tumors (including malignant and benign) in that group was about 1 in 400, and the excess relative risk (ERR) from acute exposure to radiation was 1.2 Sv−1 (95% confidence interval: 0.6 to 2.1 Sv−1). If the ERR estimate 1.2 Sv−1 is applied to a group of occupationally exposed persons whose estimated brain dose is 4.2 mSv · year−1, then the annual ERR incurred by the group is 0.005 year−1. If extended over a 25-year interventional career, the lifetime ERR would be 0.13. The predicted fatal neurological brain tumor incidence rate of a group of people so exposed would be about 1 in 350. For a group of 10,000 interventional physicians so exposed, 25 may be expected to develop fatal brain tumors from causes other than occupational exposure, and an additional 4 may be expected to develop fatal brain tumors from occupational radiation dose. For this hypothetical group, the estimated risk for developing a fatal brain tumor from occupational radiation exposure is 0.04%.
Given the many and potentially large uncertainties associated with estimating dose and assessing risk, an absolute risk estimate as presented here should be considered with healthy skepticism. Practical sources of variability in the clinical practice include location at which the radiation dosimeter is worn, the spatial relationship of the physician and patient, and the proper use of accessory shielding devices. It has been suggested that tissue dose estimates such as presented in this work should be considered to include relative uncertainty in the range 0.66× to 1.5× to account for this variability (31). Also, with the exception of radiation-induced cataracts, the association between occupational exposure to radiation and adverse health effects is uncertain (29).
Radiation safety remains a concern for interventional cardiologists, radiologists, and surgeons. This work provides physicians a method to estimate dose to head and neck organs and tissues from their own personal dosimeter readings and thereby provides a means to assess the potential for associated health risks. Critical assessment of selected wearable radiation safety devices demonstrated that radiopaque surgical caps can be expected to provide only minimal brain protection because the x-ray shadow they produce is not cast over the brain. Ocular lens protection offered by leaded glasses can be expected to depend on glasses design. This work highlights a need for continued development of radiation safety devices that are both effective and practical. Furthermore, it provides a method by which the potential for radiation safety devices to protect tissues can be directly assessed.
WHAT IS KNOWN? Practical information regarding radiation dose to tissues of the head and neck is required to guide radiation safety practices for interventional physicians.
WHAT IS NEW? This study provides a method by which radiation dose to tissues of the head and neck can be estimated from radiation dosimeters that are routinely worn by physicians. Furthermore, this work identifies limitations of radiopaque surgical caps intended to reduce dose to the brain.
WHAT IS NEXT? Future work to incorporate the findings of this work into radiation safety practices is warranted.
The authors have reported that they have no relationships relevant to the contents of this paper to disclose.
- Abbreviations and Acronyms
- excess relative risk
- left collar dose
- Received June 27, 2016.
- Revision received October 4, 2016.
- Accepted November 17, 2016.
- American College of Cardiology Foundation
- Andreassi M.G.,
- Piccaluga I.,
- Gargani L.,
- et al.
- Andreassi M.G.,
- Piccaluga E.,
- Guagliumi G.,
- Del Greco M.,
- Gaita F.,
- Picano E.
- Kuon E.,
- Birkel J.,
- Schmitt M.,
- Dahm J.B.
- Fetterly K.A.,
- Magnuson D.J.,
- Tannahill G.M.,
- Hindal M.D.,
- Mathew V.
- Dragusin O.,
- Weerasooriya R.,
- Jaïs P.,
- et al.
- Fattal P.,
- Goldstein J.A.
- Reeves R.R.,
- Ang L.,
- Bahadorani J.,
- et al.
- Alazzoni A.,
- Gordon C.L.,
- Syed J.,
- et al.
- Fetterly K.,
- Schueler B.,
- Grams M.,
- Sturchio G.
- Marshall N.W.,
- Faulkner K.,
- Clarke P.
- Sturchio G.M.,
- Newcomb R.D.,
- Molella R.,
- Varkey P.,
- Hagen P.T.,
- Schueler B.A.
- Geber T.,
- Gunnarsson M.,
- Mattsson S.
- Lange H.W.,
- Von Boetticher H.
- ↵U.S. Nuclear Regulatory Commission. 10 CFR Part 20—standards for protection against radiation. US NRC 56 FR 23391, 1991.
- ↵Conference of Radiation Control Program Directors. Suggested state regulations for control of radiation, volume i, part D. March 2003. Available at: http://www.crcpd.org/SSRCRs/dpart.pdf. Accessed June 2, 2016.
- International Commission on Radiological Protection
- National Research Council
- Preston D.L.,
- Ron E.,
- Yonehara S.,
- et al.
- International Commission on Radiological Protection