Radiation Dose Savings for Adult Pulmonary Embolus 64-MDCT Using Bismuth Breast Shields, Lower Peak Kilovoltage, and Automatic Tube Current Modulation
Abstract
OBJECTIVE. The purpose of this study was to assess whether radiation dose savings using a lower peak kilovoltage (kVp) setting, bismuth breast shields, and automatic tube current modulation could be achieved while preserving the image quality of MDCT scans obtained to assess for pulmonary embolus (PE).
MATERIALS AND METHODS. CT angiography (CTA) examinations were performed to assess for the presence or absence of pulmonary artery emboli using a 64-MDCT scanner with automatic tube current modulation (noise level = 10 HU), two kVp settings (120 and 140 kVp), and bismuth breast shields. Absorbed organ doses were measured using anthropomorphic phantoms and metal oxide semiconductor field effect transistor (MOSFET) detectors. Image quality was assessed quantitatively as well as qualitatively in various anatomic sites of the thorax.
RESULTS. Using a lower kVp (120 vs 140 kVp) and automatic tube current modulation resulted in a dose savings of 27% to the breast and 47% to the lungs. The use of a lower kVp (120 kVp), automatic tube current modulation, and bismuth shields placed directly on the anterior chest wall reduced absorbed breast and lung doses by 55% and 45%, respectively. Qualitative assessment of the images showed no change in image quality of the lungs and mediastinum when using a lower kVp, bismuth shields, or both.
CONCLUSION. The use of bismuth breast shields together with a lower kVp and automatic tube current modulation will reduce the absorbed radiation dose to the breast and lungs without degradation of image quality to the organs of the thorax for CTA detection of PE.
Introduction
The whole-body effective radiation dose using 16-array or higher MDCT for thoracic vascular imaging can range up to 14-20 mSv [1, 2]. This exposure may result in a 0.2-2.2% increased lifetime relative risk for either breast or lung cancer compared with the nonimaged general population [1]. Modifications of various MDCT parameters—for example, peak kilovoltage (kVp) or tube current (mA) (automatic tube current modulation to modify mA depending on body habitus)—can lower overall radiation exposure while maintaining or even optimizing image quality [3-8]. However, some of these modifications can produce other problems. For example, lowering the kVp may result in overall increased image noise and increased radiation absorbed in the breasts. Automatic tube current modulation, on the other hand, can reduce overall radiation exposure while maintaining image quality at a set noise level. In addition, but not intrinsic to the scanning apparatus, bismuth shields placed on the anterior chest wall can reduce radiation dose to the ventral thorax but may increase image noise [9, 10]. The purpose of this study was to assess the radiation dose savings and image quality for 64-MDCT pulmonary embolus (PE) studies when using these three methods—a lower kVp, automatic tube current modulation, and bismuth breast shields—in various combinations.
Materials and Methods
Dose Measurements
Dose measurements were performed and determined by consensus of one medical physicist and three health physics staff members with a combined 35 years of radiation dosimetry experience. Two anthropomorphic female phantoms of different sizes (models 702-D and 701-D, CIRS) that have been validated for human organ dosimetry measurements were used in combination with metal oxide semiconductor field effect transistor (MOSFET) detectors. Different sizes of phantoms were used to simulate different body habitus.
Phantom 1, the small phantom, had the following specifications: weight, 55 kg; height, 160 cm; and thorax dimensions, 20 × 25 cm. Phantom 2, the larger phantom, had the following specifications: weight, 73 kg; height, 173 cm; and thorax dimensions, 23 × 32 cm. The phantoms are sub divided into contiguous axial sections with the breasts attached separately. Each section contains several 5-mm-diameter through-holes optimized for dosimetry of internal organs. Each hole in the phantoms is labeled with a number that can be referenced to the user's manual for that specific phantom (models 702-D and 701-D handling instructions, CIRS).
For this study, 20 MOSFET detectors (model TN-1002R-D, Thomson-Nielson) were placed as follows: Three detectors were placed in the right lung (anterior, middle, and posterior locations); two detectors were placed in the thoracic esophagus, proximal and distal; one detector each was placed on the skin over the anterior chest wall, the posterior chest wall, and the bone marrow anteriorly in the sternum, posteriorly in the thoracic spine, laterally in a left and right rib, and anteriorly to the bismuth shield. Four detectors were placed in each breast at the 3-, 6-, 9-, and 12-o'clock positions; clock positions are noted as if one is facing the phantom. Preparation of the MOSFET detectors has been described previously [2].
Three scans of identical z-axis length were obtained for each of the MDCT protocols listed below. An average dose delivered to each of the 20 detectors was calculated. The MOSFET reader was connected to a laptop computer, and the data were read immediately after each CT scan. The software (model TN-RD-49, Thomson-Nielsen) stored the acquired data in centigrays (cGy) in spreadsheet format (Microsoft Office Excel). The CT dose index volume (CTDIvol) and dose-length product (DLP) values for each scan were recorded as displayed on the operator console of the CT scanner at the time of imaging because these values are separate measures that can be used to estimate radiation dose and CT quality.
MDCT Scanner
A 64-MDCT scanner (LightSpeed VCT, GE Healthcare) was used. After obtaining initial scout views in the anteroposterior and lateral projections, the phantoms were scanned with each of the following protocols: 140 kVp, 120 kVp, 140 kVp with shields, 120 kVp with shields, and 140 kVp with shields and foam pad. Automatic tube current modulation (Smart mA, GE Healthcare) was set with the noise level at 10 HU and a maximum tube current of 700 mA for all protocols at the start of each scan in accordance with our institution's standard clinical PE protocol. Smart mA is a combined angular and longitudinal tube current modulation system with modifications based only on the scout radiograph; therefore, for protocols using the bismuth shields, the shields were placed over the anterior chest of the phantom to cover the breasts after the scout images had been obtained to avoid an increase in the mA calculated by the scanner because of a perceived increase in body density with the bismuth shields in place. The imaging parameters are listed in Table 1.
| PE Protocol | |||||
|---|---|---|---|---|---|
| Parameter | 140 kVp | 140 kVp with Shields | 140 kVp with Shields and Foam Pad | 120 kVp | 120 kVp with Shields |
| Peak kilovoltage (kVp) | 140 | 140 | 140 | 120 | 120 |
| Amperage (mA) | |||||
| Maximum | 690 | 690 | 690 | 686 | 690 |
| Minimum | 515 | 515 | 515 | 578 | 690 |
| Gantry rotation time (s) | 0.6 | 0.6 | 0.6 | 0.6 | 0.6 |
| Longitudinal coverage (cm) | 20 | 20 | 20 | 20 | 20 |
| Beam pitch | 1.375:1 | 1.375:1 | 1.375:1 | 1.375:1 | 1.375:1 |
| Acquisition collimation | 64 × 0.625 | 64 × 0.625 | 64 × 0.625 | 64 × 0.625 | 64 × 0.625 |
| Reconstruction collimation | 1.25 | 1.25 | 1.25 | 1.25 | 1.25 |
| Scanning FOV (cm) | 50 | 50 | 50 | 50 | 50 |
| Small female phantom | |||||
| CTDIvol (mGy) | 31.7 | 31.7 | 31.7 | 23.27 | 25.23 |
| DLP (mGy × cm) | 838.92 | 838.92 | 838.92 | 615.45 | 667.51 |
| Larger female phantom | |||||
| CTDIvol (mGy) | 31.28 | 31.28 | NA | 24.87 | 24.87 |
| DLP (mGy × cm) | 827.44 | 827.44 | NA | 657.76 | 657.76 |
Note—FOV = field of view, CTDIvol = CT dose index volume, DLP = dose-length product, NA = not applicable.
A commercially available bismuth breast shield (ARBO8-4, Dyna Medical) was used. This shield is composed of two 20 × 20 cm squares connected by a strap made of hook-and-loop fasteners (Velcro, Velcro Indus tries). Each square is a thickness of four plies of bismuth. Scanning was performed with the shields placed directly on the breasts for two of the protocols (140 kVp with shields and 120 kVp with shields). In addition, the PE protocol at 140 kVp was performed using the bismuth shields and a foam pad (140 kVp with shields and foam pad) with the small female phantom. This sequence was performed with the shields backed by 1 cm of thick black foam placed directly on the breasts (i.e., with foam pad between shields and breast) to assess whether dose or streak artifact with this configuration might differ; this sequence was performed using the small phantom only because streak artifact should be magnified in a smaller body.
Image Analysis
Image quality was assessed by repeating the protocol described earlier using a multipurpose chest phantom (n1 Lungman, Kyoto Kagaku Company) that has a detailed anatomic makeup of the unenhanced pulmonary arteries, airways, and lungs. This chest phantom's torso is 43 cm wide and 48 cm tall, has a circumference of 94 cm, and weighs approximately 18 kg. The breast attachments for the small female phantom used for the dosimetry work were attached to the multipurpose chest phantom using tape, and the area between the chest wall and the breasts was filled with a high-fat butter (Plugrá, Keller's Creamery) that does not melt at room temperature to remove any air gaps between the chest wall and the breast attachments that may confound assessment of image quality.
Image quality was assessed using both quantitative and qualitative criteria. For quantitative assessment of change in image noise, individual 300 mm2 round regions of interest were placed over the anterior left chest wall posterior to the breast, the mediastinum at the level of the heart, the left lung anteriorly and posteriorly, and the left posterior chest wall. Attenuation values (in Hounsfield units) were taken in these locations, with the SD of these values reflecting the noise level. These measurements were performed three times in each anatomic location for each protocol and were then averaged.
For qualitative assessment, the images were reviewed independently by three fellowship-trained chest radiologists with 1, 5, and 30 years of experience who were blinded to the protocol and the presence or absence of bismuth shields; the display field of view was decreased to 20 cm to exclude the shields from the field of view. Image quality was assessed in the anterior chest wall, the mediastinum at the level of the heart, the lungs anteriorly and posteriorly, and the posterior chest wall using a rating scale of 1-5. The scoring scale was marked with 5 being excellent image quality with no streak artifact or mottle, 4 being good image quality with minimal streak artifact or mottle in less than 10% of the area being assessed, 3 indicating adequate image quality with scattered streak artifact or mottle in 10-25% of the area, 2 being fair image quality with streak artifact or mottle in 25-50% of the area assessed, and 1 denoting poor image quality with greater than 50% of the assessed area being limited by streak artifact or mottle. Representative images are shown in Figures 1A, 1B, 1C, 1D, 1E, 1F, 1G, 1H, 1I, and 1J.
Results
Radiation Dose
For the small female phantom, the individual absorbed dose for all organs of the thorax was reduced 4-47% when the bismuth breast shields were used for scanning at 140 kVp (Fig. 2). Further dose reduction of 9-67% was achieved when the lower peak kilovoltage (i.e., 120 kVp) was used in conjunction with the bismuth shields and automatic dose modulation (Fig. 3).
For the small female phantom, scanning at 140 kVp using automatic tube current modulation produced breast doses ranging from 3.8 to 6.0 cGy; adding bismuth shields lowered this dose to 2.7-3.5 cGy, an average dose savings of 33%. Changing the scanning parameters in the same way led to lowering the lung dose approximately 15%. Lowering the peak kilovoltage to 120 kVp and scanning with automatic tube current modulation resulted in breast doses ranging from 2.0 to 4.4 cGy, an average dose savings to the breast of 28% (Fig. 4); adding bismuth shields led to a total breast dose savings of 55%. With the lower kVp scanning technique, lung doses ranged from 3.2 to 3.4 cGy, with average dose savings of 47%; adding bismuth shields led to a total lung dose savings of 42%. For the lungs and bone marrow, the dose savings when using the bismuth shields were greatest in the anterior rather than the posterior thorax, 2-50% for the lungs anteriorly versus up to 37% posteriorly, and 23-56% for the bone marrow anteriorly versus up to 36% posteriorly.
For the larger female phantom (Table 2), the average breast dose was 6.2 cGy (range, 2.6-7.4 cGy) when scanning at 140 kVp using automatic tube current modulation, and it was 3.3 cGy (range, 1.5-5.0 cGy) when the bismuth shields were added. Lowering the peak kilovoltage to 120 kVp resulted in an average breast dose of 4.4 cGy (range, 1.6-5.7 cGy); this dose was reduced to 2.0 cGy (range, 0.6-2.3 cGy) with the addition of bismuth shields. For the lungs, the average absorbed dose at 140 kVp using automatic tube current modulation was 6.0 cGy (5.1-6.4 cGy); this dose was lowered to 5.2 cGy (range, 4.8-5.6 cGy) with the addition of bismuth shields. Lowering the peak kilovoltage to 120 kVp resulted in an average lung dose of 4.3 cGy (range, 3.8-5.4 cGy). This dose was lowered to 3.6 cGy (range, 2.4-3.8 cGy) with the addition of bismuth shields. The average dose savings to the breast and the lungs in the larger female phantom were 68% and 40%, respectively, when scanning at 120 kVp using the bismuth shields and automatic tube current modulation as compared with scanning at 140 kVp using automatic tube current modulation without the bismuth shields.
| Absorbed Organ Dose (cGy) | % Difference in Organ Dose | ||||
|---|---|---|---|---|---|
| Without Shields | With Shields | ||||
| Peak Kilovoltage Setting | Max | Range | Max | Range | |
| 120 kVp | |||||
| Skin | |||||
| Front, top of shields | NA | NA | 3.85 | 0.731-3.85 | |
| Front, under shield | 4.73 | 1.83-4.73 | 2.53 | 1.23-2.53 | 47 |
| Back | 4.73 | 2.50-4.73 | 4.73 | 2.27-4.73 | 0 |
| L breast | |||||
| 12-o'clock position | 3.97 | 3.03-3.97 | 1.68 | 1.39-1.68 | 58 |
| 3-o'clock position | 5.08 | 2.45-5.68 | 2.50 | 1.00-2.50 | 56 |
| 6-o'clock position | 3.61 | 3.06-3.61 | 2.10 | 1.42-2.10 | 42 |
| 9-o'clock position | 4.55 | 2.30-4.55 | 1.58 | 0.576-1.58 | 65 |
| R breast | |||||
| 12-o'clock position | 3.91 | 1.66-3.91 | 2.16 | 0.938-2.16 | 45 |
| 3-o'clock position | 5.21 | 2.39-5.21 | 1.25 | 0.679-1.25 | 76 |
| 6-o'clock position | 4.44 | 2.03-4.44 | 2.22 | 1.09-2.22 | 50 |
| 9-o'clock position | 4.16 | 2.66-4.16 | 2.25 | 1.19-2.25 | 46 |
| Esophagus, upper thorax | 1.61 | 0.50-1.61 | 1.18 | 0.429-1.18 | 27 |
| Bone marrow in sternum | 4.54 | 2.88-4.54 | 2.83 | 2.08-2.83 | 38 |
| Bone marrow in thoracic spine | 2.62 | 2.00-2.62 | 2.12 | 1.62-2.12 | 19 |
| Bone marrow in L rib | 3.27 | 2.33-3.27 | 2.83 | 2.27-2.83 | 13 |
| Bone marrow in R rib | 4.96 | 2.43-4.96 | 5.09 | 3.13-5.09 | −3 |
| R lung | |||||
| Front | 5.35 | 5.09-5.35 | 4.22 | 3.70-4.22 | 21 |
| Middle | 3.80 | 2.90-3.80 | 2.83 | 2.43-2.83 | 26 |
| Back | 3.81 | 2.93-3.81 | 3.81 | 2.67-3.81 | 0 |
| Esophagus, lower thorax | 2.75 | 2.62-2.75 | 2.44 | 2.22-2.44 | 11 |
| 140 kVp | |||||
| Skin | |||||
| Front, top of shields | NA | NA | 5.83 | 1.87-5.83 | |
| Front, under shields | 5.74 | 2.32-5.74 | 3.10 | 1.65-3.10 | 46 |
| Back | 5.18 | 2.61-5.18 | 4.86 | 2.61-4.86 | 6 |
| L breast | |||||
| 12-o'clock position | 5.81 | 3.65-5.81 | 2.96 | 1.88-2.96 | 49 |
| 3-o'clock position | 7.38 | 2.63-7.38 | 3.08 | 1.88-3.08 | 58 |
| 6-o'clock position | 5.70 | 4.96-5.70 | 2.70 | 1.89-2.70 | 53 |
| 9-o'clock position | 5.28 | 2.79-5.28 | 2.69 | 1.52-2.69 | 49 |
| R breast | |||||
| 12-o'clock position | 6.78 | 2.83-6.78 | 4.96 | 1.65-4.96 | 27 |
| 3-o'clock position | 6.19 | 2.74-6.19 | 3.41 | 1.63-3.41 | 45 |
| 6-o'clock position | 6.42 | 4.96-6.42 | 3.71 | 2.46-3.71 | 42 |
| 9-o'clock position | 5.90 | 4.00-5.90 | 3.24 | 2.95-3.24 | 45 |
| Esophagus, upper thorax | 1.94 | 1.27-1.94 | 1.52 | 0.727-1.52 | 22 |
| Bone marrow in sternum | 6.27 | 3.50-6.27 | 3.46 | 2.27-3.46 | 45 |
| Bone marrow in thoracic spine | 3.41 | 2.95-3.41 | 3.64 | 2.36-3.64 | −7 |
| Bone marrow in L rib | 6.70 | 4.95-6.70 | 5.10 | 4.20-5.10 | 24 |
| Bone marrow in R rib | 6.17 | 4.96-6.17 | 4.75 | 3.88-4.75 | 23 |
| R lung | |||||
| Front | 6.42 | 6.21-6.42 | 5.13 | 4.79-5.13 | 20 |
| Middle | 6.43 | 5.29-6.43 | 5.57 | 4.52-5.57 | 13 |
| Back | 5.12 | 3.96-5.12 | 4.81 | 3.77-4.81 | 6 |
| Esophagus, lower thorax | 4.12 | 3.63-4.12 | 3.81 | 3.70-3.81 | 8 |
Note—Max = maximum, NA = not applicable.
CTDIvol and DLP values at the time of scanning are displayed in Table 1. Keeping the noise level constant at 10 HU while lowering the peak kilovoltage from 140 to 120 kVp and using automatic tube current modulation resulted in a decrease in the CTDIvol and DLP. When scanning at a constant peak kilovoltage of 140 or 120 kVp in conjunction with automatic tube current modulation, the CTDIvol and DLP did not vary whether the bismuth breast shields were present or not because these values are determined from the scout radiographs, which were obtained when the shields were not present.
Image Quality
The noise levels from the multipurpose chest phantom for the anterior chest wall, lungs, mediastinum, and posterior chest wall are listed in Table 3. There was no significant variation in the noise levels in the lungs using a lower kVp or bismuth shields. For the mediastinum, the anterior chest wall, and the posterior chest wall, the noise levels increased with a reduction in kVp or when the bismuth shields were used. Overall, the greatest increase in image noise using a constant kVp and automatic tube current modulation with the bismuth breast shields was in the anterior chest wall (42-43%), with less of an increase in image noise in the other structures of the thorax (21-23% in the mediastinum; 6.7-9.3%, anterior lungs; 1.5-2.7%, posterior lungs; and 3.5-14%, posterior chest wall).
| Mean ± SD Noise (HU) | |||||
|---|---|---|---|---|---|
| Area | 140 kVp | 120 kVp | 140 kVp with Shields | 120 kVp with Shields | 140 kVp with Shields and Foam Pad |
| Anterior chestwall | 7.87±0.04 | 8.24±0.02 | 13.47±0.12 | 14.43±0.08 | 13.86±0.09 |
| Mediastinum | 6.78±0.02 | 7.72±0.04 | 8.64±0.03 | 10.04±0.14 | 9.60±0.02 |
| Lung | |||||
| Anterior | 78.91±0.48 | 66.18±0.63 | 73.98±1.40 | 73.01±1.37 | 73.75±2.30 |
| Posterior | 79.46±1.80 | 81.15±1.30 | 77.40±0.31 | 82.41±1.10 | 80.99±0.13 |
| Posterior chestwall | 13.62±0.12 | 13.54±0.20 | 13.16±0.11 | 11.86±0.03 | 13.57±0.06 |
Note—Repeat measurements of 300 mm2 regions of interest, 181 slices.
Qualitative assessment revealed excellent or good image quality in the lungs, heart, and mediastinum for all protocols with no variation in score regardless of kVp setting or the presence of bismuth shields (Table 4). Image quality for the posterior chest wall did not vary notably with the different protocols. Using the bismuth shields resulted in a lower quality score for the anterior chest wall that was independent of the kVp setting (Table 4). No significant streak artifact extending into the thorax was noted with any of the protocols (Figs. 1A, 1B, 1C, 1D, 1E, 1F, 1G, 1H, 1I, and 1J).
| Mean Image Quality Score | |||||
|---|---|---|---|---|---|
| Area | 140 kVp | 120 kVp | 140 kVp with Shields | 120 kVp with Shields | 140 kVp with Shields and Foam Pad |
| Anterior chestwall | 4.0 | 4.0 | 2.7 | 3.0 | 3.3 |
| Mediastinum | 4.0 | 4.0 | 4.0 | 4.0 | 4.0 |
| Heart | 4.0 | 4.0 | 4.0 | 4.0 | 4.0 |
| Lung | |||||
| Anterior | 5.0 | 5.0 | 5.0 | 5.0 | 5.0 |
| Posterior | 5.0 | 5.0 | 5.0 | 5.0 | 5.0 |
| Posterior chestwall | 3.3 | 3.7 | 3.0 | 3.0 | 3.3 |
Note—Scoring scale was 5 being excellent image quality with no streak artifact or mottle, 4 being good image quality with minimal streak artifact or mottle in less than 10% of the area being assessed, 3 indicating adequate image quality with scattered streak artifact or mottle in 10-25% of the area, 2 being fair image quality with streak artifact or mottle in 25-50% of the area assessed, and 1 denoting poor image quality with greater than 50% of the assessed area being limited by streak artifact or mottle.
Discussion
Radiation dose savings can be achieved by altering several CT parameters. Lowering the mA results in a linear decrease in dose, whereas lowering the kVp reduces overall dose by approximately the power of 2. Both modes of dose reduction affect image quality throughout the field of view. In contrast, bismuth shields allow selective reduction in radiation dose by filtering out lower-energy photons from the polychromatic x-ray beam. The shields have been promoted as a way to reduce radiation dose preferentially in the superficially located organs such as the eye, thyroid, and breast. The results of our study show that the savings in absorbed dose to the thoracic organs are substantial when using bismuth breast shields in conjunction with a lower kVp and automatic tube current modulation for adult 64-MDCT thoracic vascular protocols. In fact, the reduction of absorbed dose to the breast and lungs seen in our study using breast shields is more impressive than previously reported [9, 10]. Also, we determined, for the first time to our knowledge, the degree of dose savings for adult thoracic vascular MDCT when bismuth shields are used in conjunction with tube current modulation; in addition, we showed that using a lower kVp in conjunction with bismuth breast shields further reduces the radiation dose to the breast and internal thoracic organs.
Using bismuth breast shields resulted in greater dose savings in the anterior thorax than in the posterior thorax. This difference in dose savings is expected because the anteriorly placed shields filter out lower-kVp-energy radiation from the polychromatic beam that would normally be absorbed preferentially by the more anteriorly located soft tissues. Placement of these shields over the anterior thorax should have minimal effect on x-rays entering the thorax from a posterior approach, with backscatter being limited to approximately 3% (Yoshizumi T, unpublished data). One could postulate that doses to the thorax could be reduced further by wrapping shields around the entire thorax or by advocating that manufacturers make shields that filter out lower-energy photons before they impact the patient.
Qualitative assessment of the images showed that the presence of the foam pad reduced streak artifact in the chest wall. This reduction is streak artifact is presumedly because the shields lie flat across the breasts (Figs. 1E and 1F) when the foam is present and are angled between the breasts when the foam is absent (Figs. 1C and 1D).
Although radiation dose reduction is a laudable goal, the techniques mentioned will not be used clinically if adequate image quality for routine use is not maintained. Acceptable image noise levels for MDCT are determined by the structures and pathophysiology being assessed. To our knowledge, no studies have established the acceptable noise level for accurate assessment of the presence or absence of pulmonary artery emboli. However, extrapolation from other thoracic and cardiac MDCT protocols suggests that a noise level of less than 25 HU is acceptable for this purpose [6, 11].
The results of our study indicate that the use of bismuth shields for thoracic MDCT maintains a noise level of less than 25 HU. Furthermore, Prasad et al. [12] and other investigators [13-15] have shown that for the evaluation of lung parenchyma, a 50% reduction in radiation dose to the chest does not compromise evaluation of the lungs or airways. In those studies, the dose reduction was greater (decrease in CT dose index of up to 85% or reduction of milliampere-second setting to 40-140 mAs) than that used in our protocols.
Lowering the mA settings by automatic tube current modulation adjustments can result in radiation dose savings; however, at some point the resultant increase in noise throughout the entire scan might not be accepted in clinical practice, especially for the diagnosis of PE. The advantage of using bismuth breast shields, in contrast, is selective filtering of the radiation beam in the mid aspect of the thorax. Higher noise levels here may be better tolerated because of less-attenuating structures, the lungs.
Similar to results reported by Geleijns et al. [10], our results show that image noise (Table 3) in the field of view of the bismuth shields is increased, but we noted no deleterious effect on image quality (Table 4). In contrast to Geleijns and colleagues, we assessed the dose savings of the bismuth shields in conjunction with automatic tube current modulation and found that the increase in image noise is most noticeable in the anterior chest wall compared with the mediastinum and is greater in the anterior rather than posterior aspect of the lungs.
The CTDIvol and DLP are reduced when a lower kVp is used but are not changed with the use of the bismuth shields. This is in contrast to the change in doses we observed by directly measuring the absorbed organ doses with MOSFET detectors. The reason for this is that CTDIvol and DLP values are determined from scout radiographs, which are obtained when the bismuth shields are not present so as not to increase the perceived density of the phantom by the CT scanner when using automatic tube current modulation. Others have shown that obtaining scout images with shields in place results in the selection of a higher mA setting by the scanner [16]. This difference is important to recognize because of the common use of CTDIvol and DLP for radiation exposure quantification.
Radiation exposure for MDCT examinations can be estimated using the CTDIvol (local slice level) and the DLP (entire study), which are displayed on the CT operator console at the time of examination. Effective dose is estimated by multiplying the DLP by an anatomic-specific conversion factor. Our results support previous work [2] showing that although the CTDIvol and DLP may be used for these purposes, because of underestimation they are not accurate for assessing absorbed organ doses or determining the risk of radiation-induced carcinogenesis [2]. Rather, these values are more appropriate for MDCT protocol optimization.
Our study has some limitations. It shows organ dose savings for two body types but did not evaluate the effect of these modifications for patients with significant subcutaneous fat. Nevertheless, our results are consistent between the two body types, a thin woman and a medium-sized woman. We assessed the image quality as judged by changes in noise around the pulmonary arteries, the mediastinum, and the surrounding soft tissues. We were not able to directly assess the effect of noise on the diagnosis of PE because of limitations of our phantoms. Data about what levels of noise are acceptable for the assessment of this disease are limited. MacKenzie et al. [17] reported a decrease in detection of PEs when mAs was lowered, whereas Tack et al. [18] determined that recognizing the presence or absence of PEs was not influenced by reducing the mAs.
Although we concentrated on the PE protocol, we believe that our results can be extrapolated to other thoracic vascular scanning protocols because, for example, current thoracic aortic imaging protocols have very similar parameters. Our protocol uses a relatively high mA setting, but this established clinical protocol is meant to be useful for patients of all sizes. In comparison with other clinical PE protocols, our CTDIvol values (23-32 mGy) are in range of those reported by others in the current literature (3.4-31 mGy) [3, 4].
In conclusion, the use of bismuth breast shields in conjunction with a lower kVp and automatic tube current modulation for 64-MDCT thoracic vascular protocols reduces the absorbed radiation dose to the thoracic organs while maintaining adequate image quality. These dose-saving strategies have the potential to decrease the likelihood of radiation-induced cancers in thoracic structures. Using bismuth shields and automatic tube current modulation even without lowering the kVp will result in decreased absorbed organ doses and may be especially applicable for large patients when higher radiation dose techniques are needed to maintain image quality. We have integrated the results of this study into our current clinical thoracic vascular MDCT protocols.
Acknowledgments
We thank Ehsan Samei and Nicole T. Ranger for use of the multipurpose chest phantom and Jenny Hoang for assistance in qualitative assessment of the images.
Footnotes
R. C. Nelson is a consultant for GE Healthcare.
Address correspondence to L. M. Hurwitz ([email protected]).
References
1.
Hurwitz LM, Reiman R, Yoshizumi T, et al. Radiation dose from contemporary cardiothoracic multidetector CT protocols with an anthropomorphic phantom: implications for cancer induction. Radiology 2007; 245:742-750
2.
Hurwitz LM, Yoshizumi TT, Goodman PC, et al. Effective dose determination using an anthropomorphic phantom and metal oxide semiconductor field effect transistor technology for clinical adult body multidetector array computed tomography protocols. J Comput Assist Tomogr 2007; 31:544-549
3.
Shueller-Weidekamm C, Schaefer-Prokop CM, Weber M, Herold CJ, Prokop M. CT angiography of pulmonary arteries to detect pulmonary embolism: improvement of vascular enhancement with low kilovoltage settings. Radiology 2006; 241:899-907
4.
Heyer CM, Mohr PS, Lemburg SP, Peters SA, Nicolas V. Image quality and radiation exposure at pulmonary CT angiography with 100- or 120-kVp protocol: prospective randomized study. Radiology 2007; 245:577-583
5.
Kalra MK, Maher MM, Toth TL, et al. Techniques and applications of automatic tube current modulation for CT. Radiology 2004; 233:649-657
6.
Singal-Cinqualbre AB, Hennequin R, Abada HT, Chen X, Paul J-F. Low-kilovoltage multi-detector row chest CT in adults: feasibility and effect on image quality and iodine dose. Radiology 2004; 231:169-174
7.
Nakayama Y, Awai K, Funama Y, et al. Lower tube voltage reduces contrast material and radiation doses on 16-MDCT aortography. AJR 2006; 187:1266 ; [web]W490-W497
8.
Abada HT, Larchez C, Daoud B, Sigal-Cinqualbre A, Paul JF. MDCT of the coronary arteries: feasibility of low-dose CT with ECG-pulsed tube current modulation to reduce radiation dose. AJR 2006; 186[6 suppl 2]:S387-S390
9.
Yilmaz MH, Albayram S, Yasar D, et al. Female breast radiation exposure during thorax multidetector computed tomography and the effectiveness of bismuth breast shield to reduce breast radiation dose. J Comput Assist Tomogr 2007; 31:138-142
10.
Geleijns J, Salvadó Artells M, Veldkamp WJ, López Tortosa M, Calzado Cantera A. Quantitative assessment of selective in-plane shielding of tissues in computed tomography through evaluation of absorbed dose and image quality. Eur Radiol 2006; 16:2334-2340
11.
d'Agostino AG, Remy-Jardin M, Khalil C, et al. Low-dose ECG-gated 64-slices helical CT angiography of the chest: evaluation of image quality in 105 patients. Eur Radiol 2006; 16:2137-2146
12.
Prasad SR, Wittram C, Shepard JA, McLoud T, Rhea J. Standard-dose and 50%-reduced-dose chest CT: comparing the effect on image quality. AJR 2002; 179:461-465
13.
Zhu X, Yu J, Huang Z. Low-dose chest CT: optimizing radiation protection for patients. AJR 2004; 183:809-816
14.
Mayo JR, Hartman TE, Lee KS, Primack SL, Vedal S, Müller NL. CT of the chest: minimal tube current required for good image quality with the least radiation dose. AJR 1995; 164:603-607
15.
Ravenel JG, Scalzetti EM, Huda W, Garrisi W. Radiation exposure and image quality in chest CT examinations. AJR 2001; 177:279-284
16.
Coursey C, Frush DP, Yoshizumi T, Toncheva G, Nguyen G, Greenberg SB. Pediatric chest MDCT using tube current modulation: effect on radiation dose with breast shielding. AJR 2008; 190:244; [web]W54-W61
17.
MacKenzie JD, Nazario-Larrieu J, Cai T, et al. Reduced-dose CT: effect on reader evaluation in detection of pulmonary embolism. AJR 2007; 189:1371-1379
18.
Tack D, De Maertelaer V, Petit W, et al. Multidetector row CT pulmonary angiography: comparison of standard-dose and simulated low-dose techniques. Radiology 2005; 236:318-325
Information & Authors
Information
Published In
Copyright
© American Roentgen Ray Society.
History
Submitted: April 14, 2008
Accepted: July 16, 2008
First published: November 23, 2012
Keywords
Authors
Metrics & Citations
Metrics
Citations
Export Citations
To download the citation to this article, select your reference manager software.