Comparative analysis of radiation dose and image quality between organ dose modulation and 3D smart mA modulation during head-neck CT angiography

Abstract

OBJECTIVE:

To assess the difference in absorbed organ dose and image quality for head-neck CT angiography using organ dose modulation compared with 3D smart mA modulation in different body mass indices (BMIs) using an adaptive statistical iterative reconstruction (ASiR-V) algorithm.

METHODS:

Three hundred patients underwent head-neck CTA were equally divided into three groups: A (18.5 kg/m²≦BMI < 24.9 kg/m²), B (24.9 kg/m²≦BMI < 29.9 kg/m²) and C (29.9 kg/m²≦BMI≦34.9 kg/m²). The groups were randomly subdivided into two subgroups (n = 50): A1-A2, B1-B2 and C1-C2. The patients in subgroups A1, B1 and C1 underwent organ dose modulation with the ASiR-V algorithm, while other patients underwent 3D smart mA modulation. The signal-to-noise ratio (SNR) and contrast-to-noise ratio (CNR) of all head-neck CT angiography images were calculated. Images were then subjectively evaluated. Mean values of several indices including dose-length product (DLP) were computed. The DLP was converted to the effective dose (ED). SNR, CNR and ED in groups A, B, and C were compared in statistical data analysis.

RESULTS:

SNR, CNR, and subjective image scores show no statistical differences in three groups (P > 0.05). However, there is significant difference of ED values (P < 0.05) . For example, in subgroup A1 mean ED values are 15.30% and 23.66% lower than those in subgroup A2 at thyroid gland and eye lens, respectively. Similar patterns also exist in groups B (B1 vs. B2) and C (C1 vs. C2).

CONCLUSIONS:

Using organ dose modulation and applying the ASiR-V algorithm can more effectively reduce the radiation dose in head-neck CT angiography than using 3D smart mA modulation, while maintaining image quality. Thus, using organ-based dose modulation has the additional benefit of reducing dose to the thyroid gland and eye lens.

Keywords

Organ dose modulation 3D smart mA modulation radiation dose image quality head-neck CT angiography

1. Introduction

With the advent of multi-detector row CT scanners, head-neck CT angiography provides a reliable and noninvasive method for studying the neck and intracranial arteries. Its clinical applications the identification and characterization of aneurysms, steno-occlusive diseases, arteriovenous malformations, vessel dissections, sino-venous thrombosis, idiopathic hemorrhage, and other vascular pathology [1]. However, the amount of radiation exposure to the patient population of head-neck CTA, especial to the most radiation sensitive organs (thyroid gland and eye lens) and its risks have also drawn radiologists’ attention [2 –5]. Therefore, decreasing the patient’s radiation dose from a head-neck CTA examination and optimizing the CT scan parameters and protocols while still maintaining image quality has become a focus of clinical radiology practice.

Some technologies such as the three-dimensional (3D) smart mA modulation allow a reduction of the tube current with no compromise to image quality [6 –8]. These techniques, however effective, lack specificity for targeted dose reduction of radiosensitive organs included in the SFOV, such as thyroid gland and eye lens. Currently, the only available organ dose modulation (ODM) technique is a new technique of radiation dose reduction that modulates tube current by reducing the tube current over a prescribed 120° radial arc over the anterior aspect of the body [9]. The organ dose modulation technique aimed at reducing the absorbed thyroid gland and eye lens dose during head-neck CTA is the use of bismuth shields, which has shown maximal dose reduction to the thyroid gland and eye lens [10]. 3D smart mA modulation for CT has been one of the most important technological developments in CT radiation dose reduction, and it yields excellent image quality while reducing the tube current [11, 12]. However, decreasing the tube current can increase image noise and adversely affect image quality. Adaptive statistical iterative reconstruction (ASiR-V) is an effective and widely used algorithm for reducing noise and maintaining image quality [13 –15].

To our knowledge, no previous study has compared the image quality and radiation dose of organ dose modulation and 3D smart mA modulation by using the ASiR-V algorithm in head-neck CTA. In this study, we compared the image quality and radiation dose between organ dose modulation and 3D smart mA modulation by using the ASiR-V algorithm for head-neck CTA.

2. Materials and methods

2.1. Patient population

From May 2017 to March 2018, 300 patients (143 men, 55.6±21.3 years; 157women, 56.7±23.1 years) with suspected internal carotid artery, vertebral artery, and intracranial arteries diseases had head-neck CTA examination, and the BMI of the patients was from 18.5 to 34.9 kg/m². The patients were divided into three groups (n = 100) according to body mass index (BMI): group A (18.5 kg/m²≦BMI < 24.9 kg/m²), group B (24.9 kg/m²≦BMI < 29.9 kg/m²) and group C (29.9 kg/m² < BMI≦34.9 kg/m²). These three groups were randomly subdivided into two subgroups: A1, A2; B1, B2, and C1, C2, with each subgroup having 50 patients. The patient characteristics are shown in Table 1.

Table 1
Patient Characteristics

Parameter Group A (n = 100) Group B (n = 100) Group C (n = 100) P value

Age (years)* 59.3±15.5 52.3±17.1 56.2±14.7 0.422

Female/ male 55/45 49/51 53/47 0.621

Weight (kg)* 56.1±7.1 75.5±7.3 84.3±7.8 0.436

Height (cm)* 170.5±13.3 166.5±11.8 160.5±103 0.768

Parameter	Group A (n = 100)	Group B (n = 100)	Group C (n = 100)	P value
Age (years)*	59.3±15.5	52.3±17.1	56.2±14.7	0.422
Female/ male	55/45	49/51	53/47	0.621
Weight (kg)*	56.1±7.1	75.5±7.3	84.3±7.8	0.436
Height (cm)*	170.5±13.3	166.5±11.8	160.5±103	0.768

*Data are expressed as mean standard deviation (SD). p < 0.05 was considered statistically significant.

2.2. CT scan protocols

All 300 patients underwent head-neck CTA with the gemstone detector to derive 128 slices per rotation on a high-definition discovery CT750 HD (HDCT, GE Healthcare, Wisconsin, USA). The ODM CT scan protocol was used on the patients in subgroup A1, B1 and C1, respectively. The scan parameters were as follows: helical, 0.5-s tube rotation time, pitch factor of 0.984:1, 32-cm SFOV, 39.37-mm/s table speed, matrix of 512×512, and 120-kVp tube voltage. Open the ODM mA table and the range of ODM include the thyroid gland and eye lens on the topogram image of head-neck. The patients in other subgroups underwent 3D smart mA technique.

The scan parameters were as follows: helical, 0.5-s tube rotation time, pitch factor of 0.984:1, 32-cm SFOV, 39.37-mm/s table speed, matrix of 512×512, and 120-kVp tube voltage. The technique is intended to achieve the desired image quality as specified by the user in the form of a NI with a user-selected minimum and maximum mA range, which extended from 10 to 550 mA, and NI value was 12. For the 3D smart mA technique, tube current was modulated to achieve an image noise of 12HU in images of a uniform object of image slice thickness of 5 mm. All images were restructured with 50% ASiR-V; the 50% ASiR setting implies 50% filtered back projection blending with 50% ASiR in the reconstructed images [16, 17]. The scan range was from the aortic arch to the cranial vault.

For all patients, CT dose index of volume (CTDI_vol), and dose-length product (DLP) was recorded from the scanner. We converted the DLP to the effective dose (ED) in millisieverts (mSv). The ED was calculated by using the formula: ED = DLP×0.0021 or 0.0059, where 0.0021 denotes the conversion factor at head (eye lens), and 0.0059 denotes the conversion factor at neck (thyroid gland) [18, 19]. The mA value of thyroid gland and eye lens was recorded from the mA table of the scanner.

All patients were injected with Iohexol (Omnipaque 350, GE healthcare, USA), a contrast medium, at an injection rate of 4.5 ml/s through the median cubital vein with a total 65 ml contrast medium, followed by 50 ml of saline solution into the median cubital vein with the same flow rate via a 22-gauge catheter. The CT acquisition was triggered using for automated start of the examination with a start trigger of 110 HU within the aortic arch and a start delay of 5 s.

2.3. Data processing and analysis

All the images of the patients were reconstructed by using a dedicated work-station. The thickness of the reconstructed images was 0.625 mm at an interval of 0.625 mm. A 3mm²±0.2 mm² ROI was placed at different spots to measure the mean CT values and image noise, as showed in Fig. 1. These areas included the left and right internal carotid arteries at eye lens slice, brain parenchyma, and the air in the same slice (Fig. 1 A). The left and right common carotid arterieies at thyroid gland slice, the sternocleidomastoid muscle, and the air in the same slice (Fig. 1 B). Measurements were performed three times and the average values were calculated. CT numbers of the left and right sides were averaged for each patient, which were regarded as the final data for statistical analysis. Areas of focal changes devoid of vascular walls, mural calcification, thrombi, medical devices, or any artifacts were carefully avoided. For all measurements, the size, shape, and position of the ROI were kept constant among the image sets by using a copy-and-paste function in the workstation.

Fig. 1.

Example of identical ROI measurements simultaneously placed in axial CT images. The ROI measurements were placed in the left and right internal carotid arteries at eye lens slice, brain parenchyma, and the air in the same slice (A); The left and right common carotid arteries at thyroid gland slice, the sternocleidomastoid muscle, and the air in the same slice (B). The region of interest (ROI) with an area of 3mm²±0.2 mm².

The CNR of all images were computed according to the following formula [17, 20]:

$CNR = ({CT}_{mean target vessel} - {CT}_{sternocleidomastoid muscle / brain parenchyma}) / {SD}_{air},$

where CT_mean _{target vessel} denotes the mean CT value of left and right common carotid arteries or left and right internal carotid arteries, CT_{sternocleidomastoid muscle/brain parenchyma} denotes the mean CT value of sternocleidomastoid muscle or brain parenchyma and SD_air denotes the mean air noise. The SNR was calculated according to the following formula [8, 17]:

$SNR = {CT}_{mean target vessel} / {SD}_{mean target vessel},$

where SD_mean _targetvessel denotes the mean noise generated by the left and right common carotid arteries or left and right internal carotid arteries. CNR and SNR measurements were obtained from the images for the objective evaluation of image quality.

2.4. Subjective evaluation of image quality

All images were interpreted by two independent radiologists with at least 20 years’ experience in head-neck CTA. The two radiologists evaluated the overall diagnostic image quality on a diagnostic PACS workstation with the same brightness and resolution settings of the same viewing monitor over a period of 2 weeks. The overall diagnostic image quality of head-neck CTA was assessed using the following 5-point scale [8 , 21]: 5, excellent image quality; 4, better-than-standard image quality; 3, standard image quality; 2, substandard image quality; and 1, non-diagnostic image quality (noise and artifact affecting image interpretation). Image noise was also evaluated using a 5-point scale: 5, minimal or no noise; 4, standard or less-than- average noise; 3, average noise; 2, above-average noise; and 1, unacceptable noise affecting image interpretation. Artifacts were graded on a 5-point scale: 5, no artifacts; 4, minor artifacts not affecting diagnosis; 3, standard or minor artifacts; 2, artifacts producing substandard image quality; and 1, obvious artifacts affecting diagnosis. The visibility of head-neck CTA was graded on a 5-point scale: 5, sharp; 4, better than average; 3, average; 2, suboptimal; and 1, no visibility.

All the images were randomized, and the readers were blinded to the CT scan protocols. The two radiologists pre-assessed the image quality of five cases according to the 5-point rating scales prior to our study; the five cases were randomly selected from a series of head-neck CTA CT scans. The purpose was to train radiologists to be familiar with the evaluation process and criteria so as to reduce inter-observer variability. If case of discrepancy, agreement was reached through open discussions. Subsequently, the case was re-assessed in order to reduce inter-observer variability [22].

2.5. Statistical analysis

All statistical calculations were performed with the software package SPSS 21.0 (SPSS Inc., Chicago, IL) with P < 0.05 indicating a statistically significant difference. The characteristics of the six patient subgroups (age, sex distribution, height and weight), image quality scores (subjective image quality), mA value, CTDI_vol, DLP, and ED were compared using one-way analysis of variance (ANOVA). The SNR and CNR of the images of head-neck CTA were compared using repeated measures ANOVA.

If data were homogeneous between the three groups A, B, and C, the least significant difference (LSD) test was performed. If data were heterogeneous, Dunnett’s T3 test was performed. Inter-observer variability was assessed with k value of concordance to measure the degree of agreement between the two radiologists for various parameters. Agreement was determined according to the value of k:<0, no agreement; 0–0.2, slight agreement; 0.21–0.40, fair agreement; 0.41–0.60, moderate agreement; 0.61–0.80, substantial agreement; and 0.81–0.10, almost perfect agreement.

3. Results

3.1. Patient demographics

As shown in Table 1, there were no differences in patients’ age, sex distribution, weight, and height among the six subgroups (all P > 0.05).

3.2. Quantitative analysis of images

In the quantitative analysis of images, in groups A, B and C, no significant difference was observed among the six subgroups in the mean signal intensity (HU) of the target veins in the head-neck CTA studies with the ODM scan mode and 3D Smart mA modulation (mean HU values of the common carotid arteries: 435.3±123.4 HU, 431.5±126.6 HU, and 433.6±121.1 HU for group A, B and C with ODM, respectively; 441.6±143.2 HU, 440.8±132.3 HU and 442.9±138.9 HU for group A, B and C with 3D smart mA, respectively. Mean HU values of middle cerebral arteries: 443.9±129.8 HU, 435.7±122.5 HU, and 437.9±120.3 HU for group A, B and C with ODM, respectively; 449.7±141.5 HU, 443.3±131.1 HU, and 441.6±128.2 HU for group A, B and C with 3D smart mA, respectively) (P > 0.05) (as shown in Tables 2 and 3, Fig. 2).

Table 2
Objective image quality with different scan mode at thyroid gland in different BMIs

Mean signal intensity (HU) Mean noise (HU) SNR CNR

A1 435.3±123.4 12.2±2.8 35.7±3.9 35.9±3.8

A2 441.6±143.2 11.8±3.1 37.2±4.2 36.6±4.5

P value 0.926 0.916 0.921 0.935

B1 431.5±126.6 13.2±3.9 32.7±3.5 35.1±3.3

B2 440.8±132.3 12.7±3.5 34.7±3.6 36.8±4.1

P value 0.899 0.887 0.917 0.889

C1 433.6±121.1 13.8±3.2 31.4±3.6 34.9±3.7

C2 442.9±138.9 13.5±3.9 32.8±3.9 35.3±3.9

P value 0.923 0.896 0.916 0.931

	Mean signal intensity (HU)	Mean noise (HU)	SNR	CNR
A1	435.3±123.4	12.2±2.8	35.7±3.9	35.9±3.8
A2	441.6±143.2	11.8±3.1	37.2±4.2	36.6±4.5
P value	0.926	0.916	0.921	0.935
B1	431.5±126.6	13.2±3.9	32.7±3.5	35.1±3.3
B2	440.8±132.3	12.7±3.5	34.7±3.6	36.8±4.1
P value	0.899	0.887	0.917	0.889
C1	433.6±121.1	13.8±3.2	31.4±3.6	34.9±3.7
C2	442.9±138.9	13.5±3.9	32.8±3.9	35.3±3.9
P value	0.923	0.896	0.916	0.931

Table 3

Objective image quality with different scan mode at eye lens in different BMIs

	Mean signal intensity (HU)	Mean noise (HU)	SNR	CNR
A1	443.9±129.8	13.1±3.2	33.9±3.3	34.6±3.6
A2	449.7±141.5	12.6±3.3	35.7±4.5	36.9±4.7
P value	0.926	0.916	0.921	0.935
B1	435.7±122.5	12.6±2.5	34.6±3.4	35.9±3.8
B2	443.3±131.1	12.1±3.1	36.6±3.6	37.3±4.1
P value	0.899	0.887	0.917	0.889
C1	437.9±120.3	12.8±3.3	34.2±3.6	35.5±3.3
C2	441.6±128.2	12.3±3.4	35.9±3.8	36.3±3.6
P value	0.887	0.868	0.912	0.923

Fig. 2.

Comparison of image quality. Comparison of image qualities. A to D provide examples of images obtained using ODM with a BMI of 29.32 kg/m² at thyroid gland (A), 3D smart mA modulation with a BMI of 23.57 kg/m² at thyroid gland (B), ODM with a BMI of 28.78 kg/m² at eye lens (C), and 3D smart mA modulation with a BMI of 25.35 kg/m² at eye lens (D).The section thickness is 0.625mmfor images A, B, C, and D. Signal intensity was measured with a ROI tool as CT value of the left and right common carotid arteries and the internal carotid arteries, image noise as standard deviation of the left and right common carotid artery and the internal carotid arteries. Typical mean signal intensity values are noted with 438.31 HU (a) and 437.67 HU (c) for the ODM at the thyroid gland and the eye lens, respectively. 432.93 HU (b) and 438.94 HU (d) for the 3D Smart mA modulation at the thyroid gland and the eye lens, respectively. There was no significant difference in signal intensity values between the two CT scan protocols (p > 0.05).

There were no significant differences in SNR and CNR among groups A, B, and C (SNR: 35.7±3.9, 32.7±3.5, and 31.4±3.6 at thyroid gland with the ODM in group A, B, and C, respectively; 33.9±3.3, 34.6±3.4 and 34.2±3.6 with the ODM at eye lens in group A, B, and C, respectively; 37.2±4.2, 34.7±3.6, and 32.8±3.9 with the 3D smart at thyroid gland in group A, B, and C, respectively; 35.7±4.5, 36.6±3.6, and 35.9±3.8 with the 3D smart at eye lens in group A, B, and C; CNR: 35.9±3.8, 35.1±3.3, and 35.3±3.9 at thyroid gland with the ODM in group A, B, and C, respectively; 34.6±3.6, 35.9±3.8 and 37.3±4.1 with the ODM at eye lens in group A, B, and C, respectively; 36.6±4.5, 36.8±4.1, and 35.3±3.9 with the 3D smart at thyroid gland in group A, B, and C, respectively; 36.9±4.7, 37.3±4.1, and 36.3±3.6 with the 3D smart at eye lens in group A, B, and C, respectively.) (P > 0.05 for all values), as shown in Tables 2 and 3, Figs. 4 and 5.

Fig. 3.

Head-neck CTA performed using the evaluated ODM with a BMI of 33.72 kg/m² (a), 26.53 kg/m² (b), and 21.65 kg/m² (e); 3D Smart mA with a BMI of 31.68 kg/m² (c), 27.31 kg/m² (d), and 22.76 kg/m² (f). The study performed with the 3D Smart mA modulation demonstrated anterior communicating artery aneurysm (white arrow) (f). A similar example is shown in the study obtained with the ODM, where aneurism in the internal carotid artery intracranial segment (white arrow) (e). The left and right internal carotid arteries, left and right vertebral arteries, and third-order or higher branches of the intracranial arteries in all the images were very clear, and there was no significant difference for subjective image quality among the different concentration of contrast at different BMI, and the numbers of head-neck CTA with over all image quality score were above 4.4.

Fig. 4.

Graph showing the comparisons of signal-to-noise ratio (SNR), contrast-to-noise ratio (CNR), and CT dose index of volume (CTDI_vol) between ODM and 3D Smart mA modulation in different body mass index (BMI) values at thyroid gland. No statistical differences in SNR and CNR between the ODM with ASiR-V and the 3D Smart mA modulation in 18.5 kg/m²≤BMI≦34.9 kg/m² (p > 0.05), but there was a significant difference noted in the CTDI_vol value between the ODM and 3D Smart mA modulation. In groups A, B, and C, there was a significant difference in CTDI_vol in subgroups A1, B1, and C1 and was decreased by 29.75%, 32.49%, and 32.16% compared to subgroups A2, B2, and C2 (p > 0.05).

Fig. 5.

Graph showing the comparisons of signal-to-noise ratio (SNR), contrast-to-noise ratio (CNR), and CT dose index of volume (CTDI_vol) between ODM and 3D Smart mA modulation in different body mass index (BMI) values at eye lens. No statistical differences in SNR and CNR between the ODM with ASiR-V and the 3D Smart mA modulation in 18.5 kg/m²≤BMI≦34.9 kg/m² (p > 0.05), but there was a significant difference noted in the CTDI_vol value between the ODM and 3D Smart mA modulation. In groups A, B, and C, there was a significant difference in CTDI_vol in subgroups A1, B1, and C1 and was decreased by 25.37%, 27.13%, and 29.68% compared to subgroups A2, B2, and C2 (p < 0.05).

3.3. Subjective evaluation of image quality

Study results showed that in the subjective evaluation of image quality, there was moderate to substantial inter-observer agreement between the two radiologists for subjective head-neck CTA image quality criteria (k = 0.61–0.8). Overall image quality, noise, artifacts and visibility scores of images scanned at different scan mode are summarized in Table 4. The numbers of head-neck CTA with over all image quality score were 4.52±0.61, 4.51±0.66, and 4.49±0.61 with ODM in different BMIs, respectively. The overall image quality scores in the head-neck CTA studies with 3D smart mA technique were 4.55±0.67, 4.54±0.67, and 4.52±0.67 in different BMIs, respectively. There was no significant statistical difference for all 4 subjective indexes scores among the groups A, B, and C (all P > 0.05; Fig. 3, Table 4).

Table 4
Subjective Image Quality with different scan mode at different BMIs for R2

Over all image quality Noises Artifacts Visibility sharpness

18.5 kg/m²≤BMI < 24.9 kg/m²

A1 4.52±0.61 4.49±0.61 4.47±0.66 4.52±0.63

A2 4.55±0.67 4.53±0.65 4.50±0.69 4.55±0.69

P value 0.751 0.676 0.836 0.821

24.9 kg/m²≤BMI < 29.9 kg/m²

B1 4.51±0.66 4.46±0.59 4.49±0.59 4.51±0.68

B2 4.54±0.67 4.47±0.64 4.48±0.65 4.53±0.66

P value 0.816 0.725 0.819 0.851

29.9 kg/m²≤BMI≦34.9 kg/m²

C1 4.49±0.61 4.48±0.62 4.51±0.63 4.49±0.61

C2 4.52±0.67 4.51±0.67 4.52±0.66 4.51±0.68

P value 0.797 0.787 0.885 0.873

	Over all image quality	Noises	Artifacts	Visibility sharpness
18.5 kg/m²≤BMI < 24.9 kg/m²
A1	4.52±0.61	4.49±0.61	4.47±0.66	4.52±0.63
A2	4.55±0.67	4.53±0.65	4.50±0.69	4.55±0.69
P value	0.751	0.676	0.836	0.821
24.9 kg/m²≤BMI < 29.9 kg/m²
B1	4.51±0.66	4.46±0.59	4.49±0.59	4.51±0.68
B2	4.54±0.67	4.47±0.64	4.48±0.65	4.53±0.66
P value	0.816	0.725	0.819	0.851
29.9 kg/m²≤BMI≦34.9 kg/m²
C1	4.49±0.61	4.48±0.62	4.51±0.63	4.49±0.61
C2	4.52±0.67	4.51±0.67	4.52±0.66	4.51±0.68
P value	0.797	0.787	0.885	0.873

3.4. Radiation dose

Radiation dose and mA value descriptors for the six subgroups are shown in Tables 5 and 6. In group A, a significant difference in mA value, CTDI_vol, DLP, and ED was observed between the two subgroups (P < 0.05). The mean ED was 3.21±0.56 mSv for ODM and 3.79±0.81 mSv for 3D smart mA modulation at thyroid gland, showing a reduction of 15.30% (P < 0.05); The mean ED was 0.71±0.16 mSv for ODM and 0.93±0.19 mSv for 3D smart mA modulation at eye lens, showing a reduction of 23.66% (P < 0.05).

Table 5
Tube currents in different direction and radiation dose with ODM and 3D smart mA at thyroid in different BMIs

mA radiation dose

A L P R CTDI_vol DLP ED

18.5 kg/m²≤BMI < 24.9 kg/m²

A1 478.25±66.58 508.89±38.22 551.37±32.51 502.25±37.71 11.12±1.14 553.81±89.53 3.21±0.56

A2 598.56±73.32 605.76±57.36 598.56±73.32 605.76±57.36 15.83±1.38 627.21±117.43 3.79±0.81

P value <0.01 <0.01 <0.01 <0.01 <0.05 <0.05 <0.05

24.9 kg/m²≤BMI < 29.9 kg/m²

B1 441.68±68.22 535.76±45.29 535.78±47.76 533.67±43.23 11.26±1.31 564.04±96.67 3.25±0.61

B2 621.68±76.89 615.45±66.63 621.68±76.89 615.45±66.63 16.68±1.65 641.21±117.65 3.89±0.68

P value <0.01 <0.01 <0.01 <0.01 <0.05 <0.05 <0.05

29.9 kg/m²≤BMI≤34.9 kg/m²

C1 458.69±87.72 538.36±58.78 543.37±92.78 542.76±62.21 11.52±1.37 569.30±101.23 3.31±0.69

C2 646.41±93.68 635.92±69.47 646.41±93.68 635.92±69.47 16.98±1.69 665.23±119.32 3.98±0.72

P value <0.01 <0.01 <0.01 <0.01 <0.05 <0.01 <0.05

	mA	radiation dose
	A	L	P	R	CTDI_vol	DLP	ED
18.5 kg/m²≤BMI < 24.9 kg/m²
A1	478.25±66.58	508.89±38.22	551.37±32.51	502.25±37.71	11.12±1.14	553.81±89.53	3.21±0.56
A2	598.56±73.32	605.76±57.36	598.56±73.32	605.76±57.36	15.83±1.38	627.21±117.43	3.79±0.81
P value	<0.01	<0.01	<0.01	<0.01	<0.05	<0.05	<0.05
24.9 kg/m²≤BMI < 29.9 kg/m²
B1	441.68±68.22	535.76±45.29	535.78±47.76	533.67±43.23	11.26±1.31	564.04±96.67	3.25±0.61
B2	621.68±76.89	615.45±66.63	621.68±76.89	615.45±66.63	16.68±1.65	641.21±117.65	3.89±0.68
P value	<0.01	<0.01	<0.01	<0.01	<0.05	<0.05	<0.05
29.9 kg/m²≤BMI≤34.9 kg/m²
C1	458.69±87.72	538.36±58.78	543.37±92.78	542.76±62.21	11.52±1.37	569.30±101.23	3.31±0.69
C2	646.41±93.68	635.92±69.47	646.41±93.68	635.92±69.47	16.98±1.69	665.23±119.32	3.98±0.72
P value	<0.01	<0.01	<0.01	<0.01	<0.05	<0.01	<0.05

Table 6

Tube currents in different direction and radiation dose with ODM and 3D smart mA at eye lens in different BMIs

	mA				radiation dose
	A	L	P	R	CTDI_vol	DLP	ED
18.5 kg/m²≤BMI < 24.9 kg/m²
A1	291.33±49.56	312.57±48.51	376.37±76.78	319.36±51.78	8.18±0.94	348.41±59.86	0.71±0.16
A2	436.32±72.23	391.45±52.71	436.32±72.23	391.45±52.71	10.96±0.99	423.78±78.21	0.93±0.19
P value	<0.01	<0.01	<0.01	<0.01	<0.05	<0.05	<0.05
24.9 kg/m²≤BMI < 29.9 kg/m²
B1	293.21±44.16	316.79±47.67	381.52±78.31	317.57±53.18	8.46±0.89	349.30±73.98	0.73±0.19
B2	443.67±71.45	417.68±50.82	443.67±71.45	417.68±50.82	11.61±1.15	425.63±80.94	0.95±0.21
P value	<0.01	<0.01	<0.01	<0.01	<0.05	<0.01	<0.05
29.9 kg/m²≤BMI≤34.9 kg/m²
C1	291.68±46.89	315.71±41.93	383.75±71.62	321.61±56.62	8.53±0.95	350.25±81.53	0.74±0.23
C2	459.67±76.82	433.59±58.81	459.67±76.82	433.59±58.81	12.13±1.22	426.62±85.52	0.97±0.26
P value	<0.01	<0.01	<0.01	<0.01	<0.05	<0.01	<0.05

In group B, a significant difference in mA value, CTDI_vol, DLP, and ED was observed between the two subgroups (P < 0.05). The mean ED value was 3.25±0.61 mSv for ODM and 3.89±0.68 mSv for 3D smart mA modulation at thyroid gland, showing a reduction of 16.45% (P < 0.05); The mean ED was 0.73±0.19 mSv for ODM and 0.95±0.21 mSv for 3D smart mA modulation at eye lens, showing a reduction of 23.16% (P < 0.05). In group C, the mA value, CTDI_vol, DLP and ED were significantly lower in the C1 subgroup than in C2 subgroup. Specifically, The mean ED value was 3.31±0.69 mSv for ODM and 3.98±0.72 mSv for 3D smart mA modulation at thyroid gland, showing a reduction of 16.83% (P < 0.05); The mean ED was 0.74±0.23 mSv for ODM and 0.97±0.26 mSv for 3D smart mA modulation at eye lens, showing a reduction of 23.71% (P < 0.05).

4. Discussion

Head-neck CTA has long scan range, and the scanning range contains areas sensitive to radiation such as the thyroid gland and eye lens. Higher radiation doses will cause potential radiation damage to these organs. With the focus on radiological protection, how to reduce the radiation dose has become the focus of the imaging industry and the imaging professionals. Organ dose modulation, 3D smart mA modulation, and ASiR are now widely applied in routine clinical practice [20 , 24]. In this study, we evaluated and compared the radiation dose and image quality between organ dose modulation and 3D smart mA modulation at different BMIs, as well as organ dose modulation combining ASiR-V in head-neck CTA. Organ dose modulation is a new technique for MSCT scanning that reduces the tube current over a predefined region of the body and is of specific interest in head-neck CTA imaging to reduce absorbed thyroid gland and eye lens radiation dose. This technique decreases the tube current of the scan over an anterior 120° arc of the patient while it relatively increases the tube current over the remaining 240° to prevent the deterioration of image quality at the center. In head-neck CTA examination, this technique aims to take into account the anatomical location of thyroid gland and eye lens to reduce the tube current in a predefined angle volume of the anterior rotation.

However, reducing the tube current will inevitably increase the image noise, thereby degrading the image quality. The ASiR technique is a reconstruction approach that improves the image quality; it has been successfully used to lower the radiation dose [7 , 20–24]. It can be also reduce the image noise induced by the lower tube current. Ma et al. [20] and Zhang et al. [2] have reported the use of a reduced radiation dose with the ASiR technique in head-neck CTA imaging. ASiR-V technique is the third generation of ASiR algorithm for enabling CT images with adequate image quality and diagnostic value at a significantly lower radiation dose. It is a reconstruction technique that enables reduction in image noise (standard deviation), streak artifact at low signal condition and improvement in low-contrast detectability, while preserving the structure details in the image, it contains more advanced noise modeling and object modeling, and also added some physics modeling, which offers further promise toward the aim of acquiring low-dose CT exams at lower radiation dose to patients, while preserving or even enhancing diagnostic value. In our study, we found that the ASiR-V technique can reduce the image noise and improve the image quality in head-neck CTA performed using organ dose modulation technique.

Organ dose modulation is a new radiation dose reduction technique for MDCT scanners that was originally intended for head-neck CTA for dose reduction to the eye lens and thyroid gland, respectively. To the best of our knowledge, organ dose modulation has not been evaluated for radiation dose reduction for head-neck CTA examinations. Our results demonstrated that organ dose modulation combined with the ASiR-V technique reduces the radiation dose by an average of 19.48%, 19.81%, and 20.27%, compared to 3D smart mA modulation for head-neck CTA in different BMI group (18.5 kg/m²≦BMI < 24.9 kg/m², 24.9 kg/m²≦BMI < 29.9 kg/m², and 29.9 kg/m²≦BMI≦34.9 kg/m²), but the image quality showed no statistically significant difference between the two scan protocols at different BMIs, respectively. Both the radiologists also thought that the overall image quality with organ dose modulation technique was identical to that of 3D smart mA technique.

Prior studies have shown that a maximum 28% decrease in organ dose using organ dose modulation compared with tube current modulation for neck CT scan [25]. However, the results of this study show that organ dose modulation combined with the ASiR-V technique reduces the radiation dose by an average of 16.19% and 23.51%, compared to 3D smart mA modulation at thyroid gland and eye lens for head-neck CTA in different BMI group (18.5 kg/m²≦BMI≦34.9 kg/m²), the difference was that prior studies focused on the differences in imaging quality and radiation dose in neck CT scan by auto mA modulation. Auto mA modulation only uses tube current modulation software to modulate the tube current in the x-axes on the basis of body habitus.

This study focused on comparing the image quality and radiation dose of organ dose modulation and 3D smart mA modulation by using the ASiR-V in head-neck CTA, the goal of 3D smart mA is to make reconstructed images contain X-ray quantum noise at a same level desired by the user, and independent of patient size and/or anatomy. It is an effective way to reduce the radiation dose, and it can modulate the x- and y-axes to the z-axis modulation of auto mA with modulating 4 times per rotation by the patients’ shape and habitus in all three dimensions to reduce the radiation dose, so it can produce less radiation dose than auto mA modulation technique.

Another important finding was that the mA attenuation in anterior was more than in posterior using organ dose modulation technique in head-neck CTA, so the dose reduction was larger in anterior organs than in posterior organs. That is because organ dose modulation provides a mode to reduce X-ray tube current (mA) in anterior direction of the patient where the most radiation-sensitive organs (thyroid gland and eye lens) are located while maintaining overall diagnostic image quality by modulating X-ray tube current according to the X-ray tube angle. The mean mA value was 459.54 and 292.07 for the anterior of thyroid gland and eye lens, 543.51 and 380.55 for the posterior, a reduction of 15.46% and 23.28%, respectively. That why ODM can modulate tube current along the z-axis, also modulates the X-ray beam in the x- and y-plane output for a constant 120° of the gantry rotation, and centrally located organs have a constant radiation exposure while superficially located organ (thyroid gland and eye lens) experience reduced exposure within the prescribed 120° radial arc.

We also found that the radiation dose in the eye lens decreased more than in the thyroid gland. The mean ED with ODM was 16.11% lower than that with 3D smart at thyroid gland, and was 23.58% lower than that at eye lens. The reason for the results may be that the X-ray attenuation of the eye lens is greater than that of the neck. We also have another found that the mean CT values of the common carotid arteries and middle cerebral arteries, sternocleidomastoid muscle, and brain parenchyma were affected by the measurement area size and location of the measurement area, especially in the common carotid arteries and middle cerebral arteries, and brain parenchyma. The larger the measurement area size, the greater the measurement error. Therefore, the measurement area size should not exceed 4mm².

Our study has several limitations. First, the patients in all of the groups were not scanned consecutively; nevertheless, there were no significant differences in patient characteristics (age, sex distribution, height, and weight) among the groups. Second, we did not evaluate any parameters reflecting image quality because ODM was not switched on in the patients in our study. Third, we did not study the differences in iodine intake between ODM scan mode and 3D smart mA modulation scan mode. Fourth, ASiR-V is a third-generation reconstruction algorithm, the most recent Veo is an advanced algorithm that can reduce radiation dose more efficiently. However, our CT scanner was not equipped with the Veo.

In conclusion, this study demonstrated that the radiation dose was significantly lower for ODM combined with the ASiR-V technique than for 3D smart mA modulation, and the use of organ-based tube current modulation also reduced the radiation dose in the thyroid gland and eye lens without inducing image quality deterioration in head-neck CTA. Thus, the recommended imaging scanning mode is OMD combined with the ASiR-V for head-neck CTA examination with BMIs (18.5 kg/m²≦BMI≦34.9 kg/m²).

Compliance with Ethical Standards:

Funding: This study was funded by national training program of innovation and entrepreneurship for undergraduates of China (grant number: 201510075026), Hebei province program of training and basic project of clinical medicine of China (grant number: 361007), and the affiliated hospital of Hebei university outstanding youth foundation (grant numbers: 2015Q002 and 2015Q017).

Conflict of Interest: we declare that we do not have any commercial or associative interest that represents a conflict of interest in connection with the work submitted.

Ethical approval: All procedures performed in our study involving human participants were in accordance with the ethical standards of the institutional and/or national research committee and with the 1964 Helsinki declaration and its later amendments or comparable ethical standards.

Informed consent: Informed consent was obtained from all individual participants included in the study. All patients signed informed consent.

This article does not contain any studies with animals performed by any of the authors.

This prospective study also received institutional board approval from the affiliated hospital of Hebei University and each participant provided informed consent.

References

Hirai ,

Korogi ,

Takahashi , et al., CT angiography in the assessment of intracranial vessels. In: Multidetector-Row CT Angiography. Berlin, Germany: Springer-Verlag; 2005.

W.L.

Zhang ,

Li ,

Zhang , et al., CT angiography of the head-and-neck vessels acquired with low tube voltage, low iodine, and iterative image reconstruction: Clinical evaluation of radiation dose and image quality, PLoS One 8 (2013), e81486.

T.H.

Wu ,

S.C.

Hung ,

J.Y.

Sun , et al., How far can the radiation dose be lowered in head CT with iterative reconstruction? Analysis of imaging quality and diagnostic accuracy, Eur Radiol 23 (2013), 2612–2621.

E.S.

Cho ,

T.S.

Chung ,

D.K.

Oh , et al., Cerebral computed tomography angiography using a low tube voltage (80kVp) and a moderate concentration of iodine contrast material a quantitative and qualitative comparison with conventional computed tomography angiography, Investig Radiol 47 (2012), 142–147.

Zhao ,

Wu ,

Zuo , et al., CT pulmonary angiography using different noise index with iterative reconstruction algorithm and dual energy CT imaging using different body mass indices: Image quality and radiation dose, J Xray Sci Technol 25 (2017), 79–91.

Zhao ,

Zuo ,

Suo , et al., A comparison of the image quality and radiation dose using 100-kVp combination of different noise index and 120-kVp in CT pulmonary angiography, Journal of Computer Assisted Tomography 40 (2016), 784–790.

Zhao ,

Zuo ,

Suo , et al., CT pulmonary angiography using automatic tube current modulation combination with different noise index with iterative reconstruction algorithm in different body mass index: Image quality and radiation dose, Academic Radiology 23 (2016), 1513–1520.

Chen ,

Zhang ,

Schoepf , et al., Radiation dose and image quality of 70 kVp cerebral CT angiography with optimized sinogram-affirmed iterative reconstruction: Comparison with 120 kVp cerebral CT angiography, Eur Radiol 25 (2015), 1453–1463.

M.P.

Lungren ,

T.T.

Yoshizumi ,

S.M.

Brady , et al., Radiation dose estimations to the thorax using organ-based dose modulation, Am J Roentgenol 199 (2012), W65–W73.

10.

L.M.

Hurwitz ,

T.T.

Yoshizumi ,

P.C.

Goodman , et al., Radiation dose savings for adult pulmonary embolus 64-MDCT using bismuth breast shields, lower peak kilovoltage, and automatic tube current modulation, Am J Roentgenol 192 (2009), 244–253.

11.

Apfaltrer ,

Sudarski ,

Schneider , et al., Value of monoenergetic low-kV dual energy CT datasets for improved image quality of CT pulmonary angiography, European Journal of Radiology 83 (2014), 322–328.

12.

R.D.

MacDougall ,

P.L.

Kleinman and

M.J.

Callahan , Size-based protocol optimization using automatic tube current modulation and automatic kV selection in computed tomography, J Appl Clin MedPhys 17 (2016), 328–341.

13.

Ghetti ,

Ortenzia and

Serreli , CT iterative reconstruction in image space: A phantom study. Phys Med 28 (2012), 161–165.

14.

A.K.

Hara ,

R.G.

Paden ,

A.C.

Silva , et al., Iterative reconstruction technique for reducing body radiation dose at CT: Feasibility study, Am J Roentgenol 193 (2009), 764–771.

15.

Wang ,

Zhang ,

Xue , et al., Combined use of iterative reconstruction and monochromatic imaging in spinal fusion CT images, Acta Radiol 58 (2017), 62–69.

16.

Kwon ,

Cho ,

Oh ,

Kim , et al., The adaptive statistical iterative reconstruction-V technique for radiation dose reduction in abdominal CT: Comparison with the adaptive statistical iterative reconstruction technique, Br J Radiol 88 (2015), doi:10.1259/bjr.20150463

17.

Cai ,

C.H.

Hu ,

X.M.

Wang , et al., Applied research of "quadri-low" combined with automatic tube current modulation and iterative model reconstruction technology in head and neck CT angiography, Zhonghua Yi Xue Za Zhi 98 (2018), 30–35.

18.

International Commission on Radiological Protection. 2007 recommendations of the International Commission on Radiological Protection (publication no. 103), Ann ICRP 37 (2007), 2–4.

19.

J.A.

Christner ,

J.M.

Kofler and

C.H.

McCollough , Estimating Effective Dose for CT Using Dose-Length Product Compared With Using Organ Doses: Consequences of Adopting International Commission on Radiological Protection Publication 103 or Dual-Energy Scanning, Am J Roentgenol 194 (2010), 881–889.

20.

Ma ,

Yu ,

Duan , et al., Subtraction CT angiography in head and neck with low radiation and contrast dose dual-energy spectral CT using rapid kV-switching technique, Br J Radiol 91 (2017), 1–7.

21.

Cai ,

Hu ,

Hu , et al., Feasibility study of iterative model reconstruction combined with low tube voltage, low iodine load, and low iodine delivery rate in craniocervical CT angiography, Clinical Radiology 73 (2018), 217.e1–217.e6.

22.

F.F.

Behrendt ,

Schmidt ,

Plumhans , et al., Image fusion in dual energy computed tomography: Effect on contrast enhancement, signal-to-noise ratio and image quality in computed tomography angiography, Invest Radiol 44 (2009), 1–6.

23.

Fanous ,

Kashani ,

Jimenez , et al., Image quality and radiation dose of pulmonary CT angiography performed using 100 and 120 kVp, Am J Roentgenol 199 (2012), 990–996.

24.

Y.X.

Zhao ,

H.N.

Suo ,

Z.W.

Zuo , et al., A comparison of the image quality and radiation dose with routine CT and the latest Gemstone Spectral Imaging combination of different scanning protocols in CT angiography of the kidney, Journal of Computer Assisted Tomography 40(6) (2016), 632–640.

25.

J.K.

Hoang ,

T.T.

Yoshizumi ,

K.R.

Choudhury , et al., Organ-based dose current modulation and thyroid shields: Techniques of radiation dose reduction for neck CT, Am J Roentgenol 198 (2012), 1132–1138.

Comparative analysis of radiation dose and image quality between organ dose modulation and 3D smart mA modulation during head-neck CT angiography

Abstract

OBJECTIVE:

METHODS:

RESULTS:

CONCLUSIONS:

Keywords

1. Introduction

2. Materials and methods

2.1. Patient population

Table 1 Patient Characteristics Parameter Group A (n = 100) Group B (n = 100) Group C (n = 100) P value Age (years)* 59.3±15.5 52.3±17.1 56.2±14.7 0.422 Female/ male 55/45 49/51 53/47 0.621 Weight (kg)* 56.1±7.1 75.5±7.3 84.3±7.8 0.436 Height (cm)* 170.5±13.3 166.5±11.8 160.5±103 0.768

2.3. Data processing and analysis

2.5. Statistical analysis

3. Results

3.1. Patient demographics

3.2. Quantitative analysis of images

Compliance with Ethical Standards:

References