Thyroid shielding in cone beam computed tomography: recommendations towards appropriate use

Abstract

The objective of this paper is to provide recommendations towards the appropriate use of thyroid shielding in dental cone beam CT (CBCT). Based on current evidence of thyroid radiosensitivity, dosimetric data in the presence and absence of shielding, and a depiction of potential adverse effects of thyroid shielding, a concise set of recommendations was prepared. According to current risk models, thyroid sensitivity is particularly high at a young age, and much higher for females. In the literature, involving adult male, female and paediatric reference phantoms, the use of a tightly fitted thyroid collar with a lead-equivalent thickness of at least 0.25 mm has consistently shown a significant reduction (average: 45.9%) of the equivalent dose to the thyroid. It can therefore be recommended that thyroid shielding should be routinely used for children undergoing CBCT scanning and is recommended for adults up to the age of 50. The increase of the X-ray tube current from automatic exposure control systems due to thyroid shielding can be avoided by placing the shielding collar after acquiring the scout images. Should real-time tube current modulation be implemented in dental CBCT imaging in the future, perspectives regarding the appropriate use of shielding may change according to current trends in CT. In view of the manifestation of metal artefacts, shielding is best avoided if radiological evaluation of tissues below the lower border of the mandible is needed.

Introduction

Despite the overall benefit of cone beam CT (CBCT) in dental and maxillofacial practice, there is continuing concern regarding the associated radiation dose. Along with justification (benefit outweighing risk), optimization (keeping the exposure to the minimum necessary to achieve the required diagnostic objective) is a fundamental principle of radiation protection of patients.¹

One organ which is of particular concern in dental X-ray imaging is the thyroid, due to its relatively high sensitivity to stochastic effects from ionizing radiation, particularly in children.^1,2 Whereas general dose-reduction techniques, such as field of view (FOV) collimation³ and mAs reduction,^4,5 will reduce the absorbed dose to the thyroid as well as other organs, the use of a thyroid collar (or shield) to optimize protection has been under consideration. However, despite efforts to standardize the use of thyroid collars in pediatric dental imaging by the Image Gently Alliance,⁶ there is inconsistent guidance regarding the application of thyroid shielding, especially for CBCT. Mixed messages are found in literature regarding its benefit, limitations and caveats, and no consensus can be found in current clinical practice.

The objective of this paper is to provide clarity regarding the applicability of thyroid shielding in CBCT, through a review of thyroid radiosensitivity, dosimetry and potential adverse effects of thyroid shielding. A brief set of recommendations was prepared regarding the proper use of thyroid shielding.

Thyroid radiosensitivity

The main sources of data used to determine the radiosensitivity of human organs are the epidemiological life span study of the survivors of the Hiroshima and Nagasaki atomic bombings and the follow up of individuals exposed due to the Chernobyl accident, whereas several other epidemiological studies are in progress.^1,2,7–9 From these data, cancer incidence and mortality risks have been derived, and estimates of health detriment per organ were determined. This eventually led to the concept of effective dose, a widely used risk metric in radiation dosimetry, defined as the weighted sum of the equivalent dose to each radiosensitive tissue or organ. The aforementioned weights (i.e., tissue weighting factors, w_T) represent the relative stochastic risk for each organ. Whereas the latest revision of tissue weighting factors has led to a reduction of the w_T for the thyroid from 0.05 to 0.04,¹ it is still considered the most radiosensitive organ found exclusively in the head and neck region. However, the lethality fraction for cancer of the thyroid (0.07) is much lower than that of other tissues (0.29–0.95) apart from skin (0.002).¹ Furthermore, it should be noted that, due to considerable rounding and consensus-based approximations,¹⁰ the w_T for the thyroid is considerably higher than its relative detriment, which is 0.022 and 0.013 for the whole population for incidence and mortality, respectively.¹

An important consideration is the effect of age and gender on radiation-induced cancer risk. Whereas the overall radiation-induced cancer risk is inversely (non-linearly) proportional to the age, and higher for females, risk models are highly organ-specific.² For the thyroid gland, the effect of gender is especially pronounced, with females showing a 5.5 times higher risk than males from birth up to the age of 10, after which the risk ratio slightly decreases with age.² Furthermore, whereas most organs show a nominal risk even at a high age of exposure, the risk for radiation-induced thyroid cancer drops below one per million per mSv at the age of 30 for males and the age of 45 for females (Figure 1).² This overall, pronounced effect of age on thyroid radiosensitivity is supported by recent data obtained after the Chernobyl accident, which caused large amounts of radioiodine being released, leading to a considerable increase in thyroid cancer for young children (<5 y.o.) in particular.^8,9

Figure 1. .

LAR for radiation-induced thyroid cancer, based on the BEIR VII model (National Research Council of the National Academies 2006), linearly extrapolated to yield the risk per 1 mSv equivalent dose. LAR, lifetime attributable risk.

Thyroid dose in CBCT: effect of shielding

Absorbed dose to thyroid from CBCT

A considerable amount of dosimetric data have been published since the onset of dental CBCT. According to a recent review,¹¹ for the thyroid, median equivalent doses reported in literature for adults were 504 µSv for large FOVs (>15 cm height), 434 µSv for medium FOVs (10–15 cm), 143 µSv for small FOVs (<10 cm) excluding the lower jaw, and 385 µSv for small FOVs including the lower jaw. Using pediatric phantoms, median thyroid doses were 1003 µSv, 227 µSv and 659 µSv for FOVs of >10 cm,<10 cm (maxilla) and <10 cm (mandible), respectively.¹¹ Maximum equivalent doses for thyroid found in literature were 6333 µSv for adults and 4265 µSv for children¹¹; the latter value corresponds to a LAR for thyroid cancer induction of 179 per million (i.e. one per ~5600) for a 5-year-old female.²

Rationale of using thyroid shielding

The dose to the thyroid can be attributed to primary radiation (i.e. X-rays coming straight from the X-ray tube) and scattered radiation (i.e. predominantly Compton scattering) (Figure 2). When the thyroid is at least partially inside the FOV, its absorbed dose will be much higher due to primary radiation. In this case, the use of thyroid shielding will result in a considerable dose reduction by blocking primary X-rays; however, special care is needed to ensure that metal artefacts do not interfere with the diagnostic image quality and that automatic exposure control or tube current modulation systems do not inadvertently boost the tube output (see below; “Adverse effects of thyroid shielding”).

Figure 2. .

Exposure to the thyroid (T) from dental CBCT examinations can be due to primary radiation (P), internal scatter (I) and external scatter (E). A tightly fitted thyroid shielding (S) will only protect from external scatter and, if overlapping with the X-ray beam, from primary radiation. CBCT, cone beam CT.

If the thyroid is not positioned within the primary X-ray beam, it can still receive a non-negligible amount of scattered radiation, depending on its distance to the edge of the FOV (e.g. at 1 cm inferior to the FOV edge, ~15% of the maximum dose found in the primary beam area has been measured for dental CBCT).¹² However, most of this scatter is internal, in which case thyroid shielding does not serve any protective effect. Nonetheless, a certain fraction of radiation scattered in the anterior mandibular region will exit the patient below the chin, and re-enter at the level of the thyroid (Figure 2).

Dosimetric evidence

Several studies have focused on the dose-reduction potential of thyroid shielding in CBCT (Table 1). On average, the use of shielding reduced the equivalent dose to the thyroid by 45.9%. For all but three measurements (two of which involve the use of loosely fitted collar, and one the use of a disposable shield with a relative low lead-equivalent thickness of 0.125 mm), dose reduction was over 20%, indicating the importance of a closely fitted collar with sufficient thickness. One study showed near-identical thyroid doses when 0.25 or 0.5 mm lead-equivalent shielding thickness was used.¹⁷ An important remark is that prior studies used large FOVs, at least covering a full jaw. For small FOVs, evidence regarding the efficacy of thyroid shielding is lacking.

Table 1. .

Overview of equivalent doses (µSv) to the thyroid for CBCT, with and without the use of shielding, reported in literature

Reference	CBCT model	FOV (cm)	Lower border of FOV	Phantom	Equivalent dose (µSv) to thyroida			Type of shieldingb
					Without shielding	With shielding	% Dose reductionc
Qu et al.13	NewTom 9000	15 × 15	Below mandible	Adult	775	848	−9.4	Thyroid collar (0.35 mmPb) loosely on front
					775	733	5.5	Two thyroid collars (0.35 mmPb x2) loosely on front and back
					775	398	48.7	Thyroid collar (0.35 mmPb) tightly on front
					775	415	46.5	Two thyroid collars (0.35 mmPb x2) tightly on front and back
Qu et al.14	DCT PRO	20 × 19	Middle of C5	Adult	1895	625	67.0	Thyroid collar (0.35 mmPb) tightly on front
					1895	728	61.6	Two thyroid collars (0.35 mmPb x2) tightly on front and back
		16 × 10	Middle of C5	Adult	2700	767.5	71.6	Thyroid collar (0.35 mmPb) tightly on front
					2700	740	72.6	Two thyroid collars (0.35 mmPb x2) tightly on front and back
		16 × 7	Middle of C5	Adult	2360	695	70.6	Thyroid collar (0.35 mmPb) tightly on front of neck
					2360	695	70.6	Two thyroid collars (0.35 mmPb x2) tightly on front and back
Tsiklakis et al.15	NewTom 9000	N/A	N/A	Adult	320	180	43.8	Shield with lead at front and back
Goren et al.16	i-CAT Platinum	23 × 17	N/A	Adult female	1580	1200	24.1	Thyroid collar (0.50 mmPb)
					1570	900	42.7	Thyroid collar (0.50 mmPb) and leaded glasses (0.75 mmPb)d
		Mandible	N/A		1780	960	46.1	Thyroid collar (0.50 mmPb)
					1950	1190	39.0	Thyroid collar (0.50 mmPb) and leaded glasses (0.75 mmPb)d
					1110	430	61.3	Thyroid collar (0.50 mmPb)
		Maxilla	N/A		470	370	21.3	Thyroid collar (0.50 mmPb)
					590	280	52.5	Thyroid collar (0.50 mmPb) and leaded glasses (0.75 mmPb)d
					1810	170	90.6	Thyroid collar (0.50 mmPb)
Hidalgo et al.17	3D Accuitomo 170	17 × 12	Below chin	Pediatric (10 y.o.)	1620	1050	35.2	Wide thyroid collar (0.25 mmPb)
					1620	950	41.4	Thyroid collar (0.25 mmPb)
					1620	980	39.5	Thyroid collar (0.25 mmPb)
					1620	940	42.0	Thyroid collar (0.50 mmPb)
					1620	1340	17.3	Disposable shield (0.125 mmPb)

CBCT, cone beam CT; FOV, field of view.

^aAbsorbed dose multiplied by radiation weighting factor, without multiplication by tissue weighting factor.

^bmmPb refers to lead-equivalent thickness when lead-free shielding is used.

^cPositive values denote a reduction in dose after shielding, negative values an increase in dose.

^d“Without shielding” refers to “with lead glasses, without thyroid shielding” in this case.

Protection of other organs through thyroid shielding

The use of thyroid shielding can be beneficial for other organs as well. Whereas the oesophagus is only partially exposed during a dental CBCT scan, it has an equal w_T and near-identical relative detriment for incidence (0.023), as well as a higher relative detriment for mortality (0.037), than the thyroid gland.¹ Thyroid shielding may also serve a minor protective effect for fractions of whole-body tissues, primarily the red bone-marrow in the upper cervical spine which is considered radiosensitive even in older individuals.²

Adverse effects of thyroid shielding

Backscattering of secondary radiation exiting the patient

As mentioned above, the majority of the thyroid dose is due to primary radiation and internal scattering. Therefore, one may wonder if thyroid shielding could have an adverse effect by backscattering radiation that is exiting the body. However, the use of high-density X-ray blockers inside a thyroid collar ensures that the majority of attenuation is due to photoelectric absorption (e.g. for lead: absorption represents 95% of total attenuation for 30 keV X-rays, 91% for 50 keV X-rays); in addition, the majority of scatter generated in a high-density material is Rayleigh scattering, which predominantly occurs in the forward direction.¹⁸ Therefore, a net increase of thyroid dose due to the use of shielding is extremely unlikely; the experimental result in Table 1 showing an increased thyroid dose of 9.4% can be attributed to measurement uncertainty.¹³

Metal artefacts

When the thyroid shield lies within the primary X-ray beam, metal artefacts will occur on the reconstructed image, as only a small fraction of X-rays will be able to penetrate the shielding and contribute to image formation. For example, at the typical effective X-ray beam energies of 50 and 60 keV used in CBCT, a 0.5 mm lead sheet attenuates 99 and 94% of X-rays, respectively.¹⁸ However, an important aspect of metal artefacts is that they occur along the axis connecting the X-ray tube and the metal object; in other words, they manifest within the axial plane (and slightly above and below this plane, depending on the beam angle) (Figure 3). The presence of metal artefacts from thyroid shielding can therefore only be considered as an adverse effect when the diagnostic region of interest is at, or slightly above, the axial level of the shielding (e.g. airway evaluation), in which case shielding is best avoided. Sufficient training of the operator, and adherence to a clearly defined referral should result in appropriate detection of cases where shielding is not advised; thus, retakes due to improper use of shielding can be kept to an absolute minimum.

Figure 3. .

Metal artefacts due to thyroid shielding. The SK150 adult anthropomorphic phantom (The Phantom Laboratory) was scanned using the NewTom Vgi Evo CBCT (Quantitative Radiology, Cefla Dental Group, Verona, Italy), using an FOV of 15 × 12 cm, with and without a lead-free thyroid collar (7350 NL, 0.50 mmPb-equivalent, Scanflex Medical, Täby, Sweden). Axial slices show (a) severe artefacts at the very bottom of the field of view, (b) posterior artefacts at the lower border of the mandible, with no perceivable effect on diagnostic image quality in the anterior region, (c) no perceivable artefacts at the level of the mental foramen. A lateral maximum intensity projection shows (d) that the visualization of hard and soft tissues in the anterior region is not impaired. CBCT, cone beam CT; FOV, field of view.

Automatic exposure control

Whereas automatic exposure control (AEC) is still uncommon as a feature of dental CBCT, it was introduced in a few early-generation CBCT units, in which the mAs was automatically determined based on the attenuation of scout images acquired during patient positioning (i.e. SafeBeam^™, Quantitative Radiology, Cefla Dental Group). While AEC can be a highly efficacious tool for patient dose optimization, its use in combination with patient shielding does pose some risk. As illustrated in Figure 4, the presence of high-density metal in the FOV could artificially inflate the mAs determined by the AEC system. Therefore, it can be recommended that, for CBCT units using this type of AEC, the thyroid shielding is attached after acquisition of the scout images.

Figure 4. .

AEC used in CBCT can lead to an inadvertent increase in tube output if thyroid shielding is attached before scout acquisition. Top: lateral scout image for a 15 × 12 cm FOV, in which the presence of thyroid shielding resulted in an automatic increase of the mAs by 21%. Bottom: for a 10 × 5 cm FOV with a caudal position, the shielding takes up over 20% of the lateral scout, leading to an increase of the mAs by 57% by the AEC system. Phantom and CBCT unit are the same as in Figure 3. AEC, automatic exposure control; CBCT, cone beam CT

A second type of AEC is tube current modulation (TCM) during the scan. This can be done by varying the relative tube output in a predetermined manner (which is already applied in CBCT, albeit not commonly) or by using real-time feedback from the detector to adjust the mA in accordance with differences in attenuation during image acquisition (which, to the best of the authors’ knowledge, is not yet applied in dental CBCT). Whereas a predetermined TCM is unaffected by the presence of thyroid shielding during the scan, real-time TCM could result in an excessive tube output when the thyroid shielding is present in the primary beam, especially for anteroposterior (AP) and PA projection angles.

Recommendations

Based on the current status of radiobiological evidence of stochastic effects to the thyroid, as well as dosimetric evidence regarding the effect of thyroid shielding on the equivalent dose to the thyroid and neighbouring organs, the following recommendations can be made:

• Thyroid shielding should be routinely used for children, regardless of the position of the FOV, and can be beneficial for thyroid protection of adults up to the age of 50.

• A collar with a lead-equivalent thickness of at least 0.25 mm, tightly fitted to the neck below the chin, is preferred.

• Thyroid shielding should not be used when visualization of tissue below, or slightly above, the axial level of the top of the shielding is needed.

• For equipment using automatic exposure control based on scout images, thyroid shielding should be positioned only after acquisition of the scout image.

• Should any CBCT equipment employ tube current modulation based on real-time detector feedback, the use of thyroid shielding should be avoided unless it is positioned outside of the primary beam with absolute certainty.