The effectiveness of bismuth breast shielding with protocol optimization in CT Thorax examination

Abstract

BACKGROUND:

Numerous techniques had been proposed to reduce radiation exposure in computed tomography (CT) including the use of radiation shielding.

OBJECTIVE:

This study aims to evaluate efficacy of using a bismuth breast shield and optimized scanning parameter to reduce breast absorbed doses from CT thorax examination.

METHODS:

Five protocols comprising the standard CT thorax clinical protocol (CP1) and four modified protocols (CP2 to CP5) were applied in anthropomorphic phantom scans. The phantom was configured as a female by placing a breast component on the chest. The breast component was divided into four quadrants, where 2 thermoluminescence dosimeters (TLD-100) were inserted into each quadrant to measure the absorbed dose. The bismuth shield was placed over the breast component during CP4 and CP5 scans.

RESULTS:

The pattern of absorbed doses in each breast and quadrant were approximately the same for all protocols, where the 4th quadrant > 3rd quadrant > 2nd quadrant > 1st quadrant. The mean absorbed dose value in CP3 was reduced to almost 34% of CP1’s mean absorbed dose. It was reduced even lower to 15% of CP1’s mean absorbed dose when the breast shield was used in CP5.

CONCLUSION:

This study showed that CT radiation exposure on the breast could be reduced by using a bismuth shield and low tube potential protocol without compromising the image quality.

Keywords

Computed tomography bismuth shielding organ equivalent dose scan protocols tube potential breast

1 Introduction

Since the invention of the first scanner in 1972, computed tomography (CT) has developed and widely applied as a medical imaging modality that produces excellent resolution, as well as the capability to perform volumetric imaging by implementing multi-row detectors (MDCT). This advanced technology increases the diagnostic value of CT scans, which has now become a favorite tool among clinicians to study and diagnose diseases. Despite the benefits, its radiation dose exposure can be quite high and hazardous to patients and operating personnel, which is a cause for concern [1, 2].

Several studies have reported the effects of cumulative radiation dose on highly radiosensitive organs, such as thyroid glands, breasts and ovaries [3, 4]. Brenner et al. (2007) determined that patients have high risk of developing cancer after being cumulatively exposed to CT radiation. The risk of radiation-induced cancer among patients generally depends on age at the time of exposure and the absorbed dosage [5].

In 2012, it has been reported that the lifetime probability of developing breast cancer among women in the United States is 13.3%, with more than 200,000 diagnosed with the disease every year [6]. In Malaysia, the incidence of breast cancer is 39.5 cases per 100,000 women [7]. There are several factors that contribute to the development of breast cancer, including radiation exposure, aging and genetic susceptibility [8, 9]. Breast tissue is highly sensitive to ionizing radiation but there is a lack of data regarding the risks in diagnostic radiology studies. Therefore, it is important to understand how the tissue may be affected by diagnostic radiation exposure due to its stochastic risk.

Some of researchers have used Monte Carlo simulations to estimate organ dose exposure on a stylized phantom rather than direct measurements with conventional calculations [10, 11]. This is primarily due to the fact that there is substantial dose variation between CT scanners, especially with different acquisition methods and protocols [12]. Current CT scanners are suited with an automatic tube current modulation (ATCM) system for most clinical protocols, such as thorax and abdomen scans. The ATCM system modulates the tube current during a CT scan, when the CT X-ray tube is rotating, to take advantage of the attenuation properties of the body from different projections, which allow for dose optimization [13, 14]. Moreover, organ-based modulation technology has been introduced in some advanced scanners and it can significantly reduce radiation exposure on the breast. Despite the incorporation of radiation-reducing systems in scanners, protective shields are still vital as they can further reduce the organs’ absorbed dose. The bismuth shield was known many as protective devices used to reduce dose from CT to radiosensitive organs such as breasts, eye lenses, and thyroid. The use of bismuth shields is safer and non-toxic when compared to lead.

The most important objective when performing a CT scan is to produce high-quality diagnostic images with minimum radiation dose exposure to the body [15]. Hence, the diagnostic value for specific CT examinations may have room for improvement. In this context, patient-specific characteristics can be used to reduce radiation doses and, thereby, decreasing the risk of developing cancer and tissue injury. Presently, the most appropriate technique to reduce radiation exposure in CT scans is to optimize acquisition parameters in accordance with patient habitus and scanner configurations. Most studies used phantoms to evaluate organ dose exposure with various dosimeters to identify the most suitable CT protocol [1 , 16]. The common detectors that are used to measure radiation are thermoluminescence dosimeters (TLD) and metal oxide semiconductor field-effect transistors (MOSFET) [17]. In addition to dose exposure, information on diagnostic quality performance is also important and is usually tested using a quantitative calculation for CT images such as CT numbers and signal-to-noise ratios (SNR).

Optimization techniques involves many factors, including the parameters used, shielding techniques and patient condition. Hence, the aim of this study is to evaluate the efficiency of using a bismuth shield and protocol optimisation to reduce CT radiation exposure to the breasts of a female-configured RANDO phantom fitted with TLD dosimeters.

2 Materials and methods

An Alderson-Rando anthropomorphic phantom (The Phantom Laboratory, Salem, USA) was used as the subject of this study. Two symmetrical external phantom breast components were attached to the top of the phantom’s chest to represent female breasts. The breast components were 10 cm high and 4 cm in diameter, with a small hole at the side to place the dosimeters.

2.1 CT scanning and protocols

CT scans were performed using the Siemens Somatom Definition 64-slice AS+(Siemens Healthineers, Erlangen, Germany) located in Hospital Sultanah Aminah, Johor, Malaysia, with the parameters stated in Table 1. The parameters were based on those used in clinical practice to reduce radiation dosage on patients and produce good images [18]. Protocol 1 (CP1) is a routine for CT thorax examination. Some modifications are made on the slice thickness in Protocol 2 (CP2), while in Protocol 3 (CP3), the tube potential values were reduced from 120 kV to 100 kV. For Protocols 4 (CP4) and 5 (CP5), AttenuRad^® bismuth sheets (Infab Corporation, Camarillo, USA) were placed on the breast components after the scanogram as shown in Fig. 2 For every protocol, the acquisition of CT helical thorax image was from the apex of the lungs to the diaphragm dome.

Fig.1

Anthropomorphic breasts were divided into 4 quadrants starting from the superior aspect of the lateral breast. Each quadrant was named as RL.

Fig.2

A bismuth breast shielding (AttenuRad^®) was placed on the top of the female-configured phantom’s breast component. The shield is applied together with scanning protocols CP4 and CP5.

Table 1

CT acquisition parameters and dose information of scan protocols

Protocols	CP1	CP2	CP3	CP4	CP5
Parameters	Routine CT	Modified slice thickness	Modified kV	Routine (shielding)	Modified (shielding)
Scan length (mm)	38	38	38	38	38
Tube potential (kVp)	120	120	100	120	100
Total mAs	2350	1840	1520	2350	1450
Effective/ref mAs	188	135	158	188	152
Slice thickness (mm)	10	5	10	10	10
Noise index	110	110	110	110	110
No of slice	40	77	40	40	40
Beam width	0.6	1.2	0.6	0.6	0.6
Pitch	1	1	1	1.2	1.2
Dose modulation	On	On	On	On	Off
CTDI_vol (mGy)	12.1	8.9	6.2	12.1	5.4
DLP (mGy.cm)	524	371	270	524	210

2.2 Direct dose measurement using TLD-100

CT radiation exposure was measured directly using thermoluminescence dosimeters (TLD-100) (LiF: Mg, Ti) (Harshaw Chemicals Ltd, Ohio, USA). The TLD-100 was chosen because it had high accuracy and was a widely used as a passive dosimeter.

Before the dosimeters were cast-off, they had to be calibrated with constant gamma ray exposure from a Cessium-137 (Cs-137) source at the Secondary Standard Dosimetry Laboratory (SSDL) of the Malaysian Nuclear Agency in Bangi, Selangor, Malaysia. The irradiated dosimeters were then read-out using a Harshaw 3500 TLD reader (Harshaw Chemicals Ltd, Ohio, USA) to determine the individual calibration factor as computed by Equation 1. $CF = \frac{Q_{rad}}{(M - M_{background} - C_{background})}$ (1)

CF is the individual correction factor, Q_rad is gamma energy of 8mGy from Cs-137, M is the result of reading from the Harshaw reader, M_background is the reading background of the TLD reader, and C_background is the background reading of the TLD-100 chips used on the environment exposure.

TLD-100 chips were annealed at 400°C for 1 hour, 100°C for 2 hours, and cooled slowly to ambient temperature to remove any residual signal. The chips were packed in a numbered capsule and stored in the dark to avoid radiation contamination. As indicated in Fig. 1 the dosimeters were inserted into each quadrant of the phantom breast components. The 1st and the 2nd quadrant were in the superior and inferior lateral aspect, while the 3rd and the 4th quadrant were in the inferior and superior of the medial aspect. All dosimeter readings were converted into absorbed dose using Equation 2:

$D_{t} = readingTL \times CF$ (2)

D_t is the absorbed dose of the phantom, reading TL is the result of measurement from the readers and CF is the conversion factor obtained from Equation 1. All results were processed using the SPSS Version 17 (IBM SPSS, Armonk, USA). Quantitative data was presented as a mean±standard deviation and analyzed using descriptive statistics. All calculations were performed in accordance with scanning protocols.

2.3 Assessment of image quality

The quality of CT images in each protocol were evaluated on selected slices using the ImageJ software Version 1.3 (National Health Institute, Maryland, USA). For each image, three regions of interest (ROI) ∼100 mm × 100 mm were marked on the lungs, breasts and heart as shown in Fig. 3 The pixels and standard deviation (SD) were converted into Hounsfield units (HU). Therefore, mean SD values for each protocol were analyzed to represent the value of the diagnostic index.

Fig.3

A CT scan image of the phantom with ~100 mm × 100 mm circles indicating the regions of interest (ROI) in the lungs, breast and heart.

3 Results

The dose distribution for each breast quadrant is presented in Fig. 4 As indicated, the pattern of absorbed doses in each breast and quadrant were approximately the same for all protocols, where the 4th quadrant > 3rd quadrant > 2nd quadrant > 1st quadrant. The lowest absorbed dose was in the first quadrant of CP5 on the right breast, with a value of 5.6mGy, while the highest dose was observed in the 4th quadrant of CP1 on the left breast, with a value of 25.0mGy.

Fig.4

Mean absorbed dose values of each quadrant obtained from all test protocols.

The mean absorbed dose of the breast components and for each protocol is shown in Table 2, where the doses were generally different for both breasts in all quadrants. However, the trend was the same as observed in the absorbed dose for both right and left breasts. The mean absorbed dose in the breasts was highest when using the standard routine protocol (CP1), which was 19.8±4.4mGy, slightly higher than CP2 at 13.3±3.1mGy. The parameters for CP4 and CP5 varied in terms of tube potential, where they were set at 120 kV and 100 kV, respectively. Although CP4 and CP5 were basically the same as CP1 and CP3, it was observed that the breast components absorbed a dose that decreased by a factor of 2.2 in CP4 and CP5 compared with CP1. This showed that in addition to the optimised protocol, bismuth shielding could augment the protection against radiation exposure on the breasts.

Table 2

Mean absorbed dose and related ratio of left and right breast components

Protocols	D_t, Absorbed dose (Mean±SD)			Ratio
Protocols	Right	Left	Average	R/L
CP1	18.9±4.4	20.8±4.8	19.8±4.4	0.9
CP2	12.1±2.9	14.6±3.0	13.3±3.1	0.9
CP3	8.6±0.9	8.9±1.6	8.7±1.3	0.9
CP4	9.9±3.5	12.5±3.4	11.2±3.5	0.8
CP5	6.0±2.0	7.6±1.3	6.8±1.8	0.8

The mean radiation exposure in CP5, which utilised a low tube potential and bismuth shielding, was at an even lower value of 6.8mGy compared with 8.7mGy in CP3, which only utilised the reduction of tube potential. The data obtained are broadly consistent with the Karim et al. study where the dose may increases due to pitch factor and for not enabling the ATCM function [10, 13]. The mean absorbed dose value in CP3 was reduced to almost 34% of CP1’s mean absorbed dose when tube potential was lowered from 120 kV to 100 kV. It was reduced even lower to 15% of CP1’s mean absorbed dose when the breast shield was used in CP5. In CP4 protocol, the absorbed dose was reduced to 43% of CP1’s value.

Table 3 shows the quantitative assessment of the image quality based on the mean and standard deviation (SD) of HU values. It had been demonstrated that utilizing low tube potential and application of breast shielding would increase the noise value of images even though they could significantly reduce the absorbed dose.

Table 3

Hounsfield Units (HU) of selected ROI images

Protocols	HU
	Mean	SD (noise)	Min	Max
CP1	5.0	11.5	–30.0	35.0
CP2	3.0	13.7	–36.0	30.0
CP3	10.0	9.2	–21.0	46.0
CP4	11.0	19.8	–73.0	88.0
CP5	12.0	23.0	–67.0	68.0

4 Discussion

Previous studies have reported the advantages of using a bismuth shield in reducing radiation exposure in CT scans. Moreover, some had shown that when used together with optimized scan parameters, the tissue absorbed dose could be reduced by 40% [13 , 20]. Usually, thorax CT scans were not performed to obtain diagnostic information of the breasts, but rather for cancer, lung parenchyma or cardiac phenomena. However, during thorax examination, the breasts would be exposed to a large amount of radiation because they were located inside the primary beam area.

This would explain why the dose values obtained in the 4th quadrant of all protocols was the highest, since this region was situated at the centre of the primary beam compared with the 1st quadrant, which was lateral of the breast. In theory, this was consistent with the Compton Scattering event, in which “thickness of patient” and “centre of beam” would affect the radiation dose received [21]. One of the most important components in a CT scanner was the ATCM system, which reduced the radiation dose by controlling the amount of photons delivered to patients depending on their habitus. In this regard, this study focused on the efficacy of combining the use of bismuth shielding and optimized parameters to reduce the CT radiation dose on the thorax region of patients, specifically the breasts. The current data corresponded to the fact that the radiation dose in diagnostic radiology depended on several aspects, such as tube potential and slice collimation.

In this study, the effectiveness of the optimization process had been established using three key criteria — tube potential, slice-collimation and breast shielding. As indicated in Table 4, adjusting the tube current and potential could result in significant implications in the radiation dose absorbed by the phantom breast components. For example, CP1 (routine protocol) contributed the highest radiation dose to the breast. But CP2 reduced the radiation exposure by a factor of 0.7 compared with CP1. The analysis presented here showed that slice thickness was inversely proportional to slice collimation width, which contributed to lower absorbed dose.

Table 4
Analysis of radiation doses and effective dose reduction

Parameter Protocols

CP1 CP2 CP3 CP4 CP5

CTDI_vol 12.1 8.9 6.2 12.1 5.4

DLP 524 371 270 524 210

D_t, average absorbed dose (mGy) 19.8 13.4 8.8 11.2 6.8

W, organ equivalent dose (mSv) 2.4 1.6 1.1 1.3 0.8

Ratio of D_t / CP1(D_t) 1.0 0.7 0.4 0.6 0.3

Parameter	Protocols
CTDI_vol	12.1	8.9	6.2	12.1	5.4
DLP	524	371	270	524	210
D_t, average absorbed dose (mGy)	19.8	13.4	8.8	11.2	6.8
W, organ equivalent dose (mSv)	2.4	1.6	1.1	1.3	0.8
Ratio of D_t / CP1(D_t)	1.0	0.7	0.4	0.6	0.3

The mean absorbed dose of CP3 was reduced by a factor of 0.9, while for both CP4 and CP5, the absorbed doses were reduced by 0.6 and 0.3, respectively. The shielding properties of bismuth could be attributed to its higher atomic number, which had the capacity to absorb most of the photon energy produced in a CT scan. The tissue absorbed dose and the corresponding scan protocols were correlated with image quality index as shown in Table 3. The results in Table 3 indicated that even though bismuth shielding of the breast did affect the quality of the thorax scan, but the image produced was comparable with lowering the tube potential alone in CP3.

The reduced dose exposure justified the use of a bismuth shield as a safety device during CT scans as the image quality was still good and could be interpreted by clinicians. It is necessary to calculate the percentage of dose reduction to the breast during the optimization process and set the basis of the protocols used based on current findings. A CT scan should be performed only when it was justified. Therefore, radiology personnel, especially technologists, should be aware of the good imaging practices before making changes to scan protocols. The implications of these protocols might be exclusive for certain patients, such as younger women. However, the effects of radiation exposure could become cumulative as these women underwent multiple CT examinations throughout their lifetime.

This study was limited by the use of an anthropomorphic phantom. A patient-specific model should be considered for accurate measurement as it could account for different body sizes, which could affect the amount of radiation exposure. A simulation study, such as the Monte Carlo technique, should also be carried out to provide an accurate patient-specific organ dose for comparison. Nevertheless, this study offered a first step towards enhancing our understanding of how CT optimization techniques with bismuth shielding worked and could be applied in routine practice.

5 Conclusion

This study showed that CT breast exposure doses can be reduced by using a bismuth shield and optimising scan parameters by modifying the tube potential. The bismuth shield reduced the breast absorbed doses by up to 14% while preserving the image quality. The study also developed recommendations on how to optimise scanning protocols to reduce the absorbed dose based on direct phantom measurements.

Footnotes

Acknowledgments

The authors like to express sincere thanks to the staff of the Radiology Department in Hospital Sultanah Aminah and Izwan Che Pi from Hospital Sultan Ismail, both based in Johor Baru, Johor, Malaysia. This research is funded by Geran Putra GP/IPM/9619800 of Universiti Putra Malaysia.

References

Karim

M.K.A.

, Hashim

, Sabarudin

, Bradley

D.A.

, Bahruddin

N.A.

, Evaluating organ dose and radiation risk of routine CT examinations in Johor Malaysia, Sains Malaysiana 45 (2016), 567–573.

Rehani

M.M.

, ICRP and IAEA actions on radiation protection in computed tomography, Ann ICRP 41 (2012), 154–160.

Wang

, Duan

, Christner

J.A.

, Leng

, Yu

, McCollough

C.H.

, Radiation dose reduction to the breast in thoracic CT: Comparison of bismuth shielding, organ-based tube current modulation, and use of a globally decreased tube current, Med Phys 38 (2011), 6084–6092.

Parker

M.S.

, Chung

J.K.

, Fatouros

P.P.

, Hoots

J.A.

, Kelleher

N.M.

, Benedict

S.H.

, Reduction of radiation dose to the female breast: Preliminary data with a custom-designed tungsten-antimony composite breast shield, J Appl Res 6 (2006), 230–239.

Balonov

M.I.

, Shrimpton

P.C.

, Effective dose and risks from medical X-ray procedures, Ann ICRP 41 (2012), 129–141.

Pearce

M.S.

, Salotti

J.A.

, Little

M.P.

, McHugh

, Lee

, Kim

K.P.

, Howe

N.L.

, Ronckers

C.M.

, Rajaraman

, Craft

A.W.

, Parker

, De

A.B.

, González, Radiation exposure from CT scans in childhood and subsequent risk of leukaemia and brain tumours: A retrospective cohort study, Lancet 380 (2012), 499–505.

Lim

, Chye

, Yahaya

, Registry

N.C.

, Second Report of the National Cancer Registry Cancer Incidence in Malaysia, Natl Cancer Regist (2003), 1–141.

Hurwitz

L.M.

, Yoshizumi

T.T.

, Reiman

R.E.

, Paulson

E.K.

, Frush

D.P.

, Nguyen

G.T.

, Toncheva

G.I.

, Goodman

P.C.

, Radiation dose to the female breast from 16-MDCT body protocols, Am J Roentgenol 186 (2006), 1718–1722.

Margolies

, Salvatore

, Eber

, Jacobi

, Lee

I.-J.

, Liang

, Tang

, Xu

, Zhao

, Kale

, Wisnivesky

, Henschke

C.I.

, Yankelevitz

, The general radiologist’s role in breast cancer risk assessment: Breast density measurement on chest CT, Clin Imaging 39 (2015), 979–982.

10.

Karim

M.K.A.

, Hashim

, Bakar

K.A.

, Bradley

D.A.

, Ang

W.C.

, Bahrudin

N.A.

, Mhareb

M.H.A.

, Estimation of radiation cancer risk in CT-KUB, Radiat Phys Chem 137 (2017), 130–134.

11.

Bhatia

, Tisseur

, Letang

J.M.

, Convolution-based scatter correction using kernels combining measurements and Monte Carlo simulations, J Xray Sci Technol 25 (2017), 613–628.

12.

Hashim

, Karim

M.K.A.

, Bakar

K.A.

, Sabarudin

, Chin

A.W.

, Saripan

M.I.

, Bradley

D.A.

, Evaluation of organ doses and specific k effective dose of 64-slice CT thorax examination using an adult anthropomorphic phantom, Radiat Phys Chem 126 (2016), 14–20.

13.

Ang

W.C.

, Hashim

, Khalis

, Karim

, Ashiqin

, Salehhon

, Adaptive iterative dose reduction (AIDR) 3D in low dose CT abdomen-pelvic: Effects on image quality and radiation exposure, J Phys Conf Ser 012006 (2017), 1–2.

14.

W.a

Kalender

, S.

Buchenau

, P.

Deak

, M.

Kellermeier

, O.

Langner

, M.

van Straten

, S.

Vollmar

and S.

Wilharm

, Technical approaches to the optimisation of CT, Phys Medica 24 (2008), 71–79.

15.

Dougeni

, Faulkner

, Panayiotakis

, A review of patient dose and optimisation methods in adult and paediatric CT scanning, Eur J Radiol 81 (2012), 665–683.

16.

Musa

, Hashim

, Ghoshal

S.K.

, Bradley

D.A.

, Ahmad

N.E.

, Karim

M.K.A.

, Hashim

, Kadir

A.B.

, General Radiographic Attributes of Optically Stimulated Luminescence Dosimeters: A Basic Insight, Radiat Phys Chem 147 (2018), 1–6.

17.

Salehhon

, Hashim

, Karim

M.K.A.

, Ang

W.C.

, Musa

, Bahruddin

N.A.

, 128 Slice Computed Tomography Dose Profile Measurement Using Thermoluminescent Dosimeter, J Phys Conf Ser 851 (2017), 012002.

18.

Kalender

, Computed tomography: Fundamentals, system technology, image quality, applications, 2011.

19.

Karim

M.K.A.

, Hashim

, Bradley

D.A.

, Bakar

K.A.

, Haron

M.R.

, Kayun

, Radiation doses from computed tomography practice in Johor Bahru, Malaysia, Radiat Phys Chem 121 (2016), 69–74.

20.

Sun

, Lin

, Tyan

, Ng

K.-H.

, Optimization of chest radiographic imaging parameters: A comparison of image quality and entrance skin dose for digital chest radiography systems, Clin Imaging 36 (2012), 279–286.

21.

Huda

, Scalzetti

E.M.

, Levin

, Technique Factors and Image Quality as Functions of Patient Weight at Abdominal CT 1, Radiology 217 (2000), 430–435.