Abstract
BACKGROUND AND PURPOSE: The combined automatic tube current modulation (ATCM) technique adapts and modulates the x-ray tube current in the x-y-z axis according to the patient's individual anatomy. We compared image quality and radiation dose of the combined ATCM technique with those of a fixed tube current (FTC) technique in craniocervical CT angiography performed with a 64-section multidetector row CT (MDCT) system.
MATERIALS AND METHODS: A retrospective review of craniocervical CT angiograms (CTAs) by using combined ATCM (n = 25) and FTC techniques (n = 25) was performed. Other CTA parameters, such as kilovolt (peak), matrix size, FOV, section thickness, pitch, contrast agent, and contrast injection techniques, were held constant. We recorded objective image noise in the muscles at 2 anatomic levels: radiation exposure doses (CT dose index volume and dose-length product); and subjective image quality parameters, such as vascular delineation of various arterial vessels, visibility of small arterial detail, image artifacts, and certainty of diagnosis. The Mann-Whitney U test was used for statistical analysis.
RESULTS: No significant difference was detected in subjective image quality parameters between the FTC and combined ATCM techniques. Most subjects in both study groups (49/50, 98%) had acceptable subjective artifacts. The objective image noise values at shoulder level did not show a significant difference, but the noise value at the upper neck was higher with the combined ATCM (P < .05) technique. Significant reduction in radiation dose (18% reduction) was noted with the combined ATCM technique (P < .05).
CONCLUSIONS: The combined ATCM technique for craniocervical CTA performed at 64-section MDCT substantially reduced radiation exposure dose but maintained diagnostic image quality.
Recent technical advances have greatly increased the clinical applications of CT, especially since the introduction of the spiral scanning techniques and the subsequent developments in multidetector-row CT (MDCT) technology. Sixty-four-section MDCT allows craniocervical CT angiography (CTA) to be performed with increased coverage, from the aortic arch to the vertex, with improved spatial resolution, shorter acquisition times, and lower doses of contrast material.1 Craniocervical CTA is now routinely used as a noninvasive alternative to conventional angiography for the evaluation and detection of intracranial aneurysms and intracranial and extracranial vascular occlusive disease, particularly in emergency situations (eg, stroke, subarachnoid hemorrhage).2–6
However, the major disadvantage with the increased use of MDCT is associated radiation exposure. Greater use of craniocervical CTA has the potential to increase the radiation burden in general. Induction of cancer is a stochastic event; however, the risk for cancer in humans proceeds in a linear fashion at lower doses without a “safe” threshold, and even the smallest dose has the potential to cause a small increase in risk to humans.7,8 A recent survey showed that CT now accounts for about 11% of all radiology procedures in the United States and constitutes approximately two thirds of the collective medical radiation dose.9 In standard CT of the head and neck, direct radiation effects relative to the eye lenses and the thyroid gland have been documented, likely resulting in cataract formation and the development of thyroid malignancies, respectively.4,10
The automatic tube current modulation (ATCM) technique enables automatic adjustment of the tube current in various planes (x-y or z) based on the size and attenuation of the body area scanned, to achieve constant image quality. Angular or x-y modulation decreases the selected tube current in the projection (in the x-y plane) that causes less attenuation, whereas the z-axis modulation technique adjusts the tube current from section to section, depending on regional body anatomy. A combination of the angular and z-axis modulations involves varying the tube current both during gantry rotation and along the z-axis of the patient. This is one of the most comprehensive approaches to CT dose reduction because the x-ray dose is adjusted according to patient-specific attenuation in all 3 planes.11–16
Despite these pronounced technical advances, little is known about the optimal imaging parameters for MDCT in craniocervical CTA. In this study, we compared image quality and radiation doses of 64-section MDCT craniocervical CTA obtained with the ATCM technique of the combined type with those obtained with a fixed tube current (FTC) technique.
Materials and Methods
Patient and Examination Protocol
The institutional review board approved the study.
We retrospectively reviewed 50 consecutive adult patients who underwent contrast-enhanced craniocervical CTA at our institution between December 2006 and November 2007. All patients had a craniocervical CTA from the aortic arch to the vertex by using a 64-section MDCT scanner (Aquilion 64; Toshiba Medical Systems, Tokyo, Japan).
Patients were divided into 2 groups. In the first group of 25 consecutive subjects (median age, 61 years; range, 32–90 years; 11 men, 14 women), an FTC of 300 mA was used because this was the standard protocol for craniocervical CTA on our scanner. Members of the other group, composed of 25 consecutive subjects (median age, 59 years; range, 18–84 years; 10 men, 15 women) were scanned with the combined ATCM technique. In all except 1 patient, the range of tube currents in the combined ATCM technique was 101–300 mA (the maximum tube current in 1 obese patient was 400 mA). The tube current at the shoulder level was the same (300 mA) for the 2 groups, except for the 1 obese patient with the combined ATCM technique.
The combined ATCM system used was the SUREExposure 3D (Toshiba Medical Systems); it controls and modulates the current in the x, y, and z directions to achieve and maintain a uniform user-selected noise level in the images. For a homogeneous object with a circular cross section, attenuation is constant over all of the projections and all measured values contribute equally. However, because the human body is not perfectly round or uniform in size or attenuation, different milliampere-second (tube current–time product) settings are required to achieve the same image quality in different parts of the body. More current is typically needed when x-rays are passing laterally through the body than when they are passing anteroposteriorly, because patients are shaped approximately elliptically, particularly in the shoulders. The SUREExposure 3D software automatically and rapidly adjusts the current during scanning to adapt to and compensate for changes in the attenuation level. Using data from the anteroposterior and lateral scanograms, the software determines exactly how much current is necessary to maintain a user-defined level of image quality. The software does this in all 3 dimensions (x-, y-, and z-axes). Thus, as the scanning moves from the shoulders to the head, more current is needed to penetrate attenuated and large areas, like the shoulders and head, whereas lower current is adequate in less attenuated and smaller areas like the neck (z-axis modulation). Furthermore, as the tube rotates around the patient, less current is used anteroposteriorly than laterally (x-y−axis modulation) (Fig 1).
Graph illustrating tube current reduction by combined ATCM, which tailors the radiation dose to the individual patient. A higher dose is needed to penetrate attenuated areas like the shoulder and head, whereas a lower dose is adequate for less attenuated areas such as the neck. The overall radiation dose with combined ATCM is less than that with FTC.
Other scanning parameters were held constant, with a tube voltage of 120 kilovolt (peak), a matrix size of 512 × 512, FOV of 28–32 cm, a section thickness of 0.5 mm, a pitch of 1.0, and an isotropic voxel size of 0.5 mm. The acquisition time was 11–16 seconds. CTA images were acquired following intravenous timed injection of contrast agent (iodixanol, Visipaque 270; Amersham Health, Oslo, Norway) by using an autotriggered mechanical injector. The injection rate was 4 mL/s to a total injection volume of 40 mL of contrast agent, followed by an injection of 20 mL of contrast agent at 3 mL/s.
Image processing consisted of standardized axial, coronal, and sagittal multiplanar volume-reformatted maximum intensity projections (MIP; 8-mm thickness, 2-mm reconstruction interval) and 3D volume-rendered reconstructions (performed on an Advantage Windows, Version 4.2 workstation; GE Healthcare, Milwaukee, Wis).
Indications for craniocervical CTA were the detection of aneurysms and possible atherosclerotic disease of the intracranial and extracranial vascular system in patients with a transient ischemic attack or ischemic stroke. Patients with poor-quality CTA due to patient movement and those who had undergone carotid stent placement or coiling or clipping for an aneurysm were excluded from the study. Patients who simultaneously underwent craniocervical CTA with standard head CT or cerebral perfusion were also excluded because we could not separate radiation doses from the craniocervical CTA from those of the combined imaging.
Image Analysis
To compare patient sizes in the groups, we used 2 measurements. The length of scan coverage and the maximum transverse neck diameter at the level of the hyoid bone were recorded because these patient-related factors may affect image quality or the radiation dose.
Objective evaluation of image quality was based on an evaluation of image noise. Image noise was recorded for each examination in the muscles at 2 anatomic levels: the sternocleidomastoid muscle in the upper neck (submandibular gland level, possibly the lowest current site) and the pectoralis major muscle, at shoulder level. For measurement of image noise, circular or ovoid regions of interest with a size of 0.5–1.0 cm2 (0.5 cm2 in the sternocleidomastoid muscle and 1.0 cm2 in the pectoralis major muscle) were placed in a homogeneous region of muscle without fat. SDs of the attenuation in these regions of interest were recorded. The SD, in Hounsfield units, of the attenuation in a particular region of interest can be used as a noise measurement.13
Qualitative image scoring was performed independently by 2 staff neuroradiologists with experience in craniocervical CTA. Scanning parameters were removed at the PACS workstation; thus, the readers were blinded with regard to the scanning parameters used. The readers evaluated subjective image quality at standardized axial, coronal, and sagittal multiplanar volume-reformatted MIP and 3D volume-rendered reconstructions of craniocervical CTA. The readers evaluated and independently scored the image quality parameters of the vascular delineation in various arterial vessels, the visibility of small arterial detail, image artifacts, and the certainty of diagnosis.
The readers were asked to evaluate the vascular delineation of various arterial vessels, such as the common carotid arteries, bifurcation of the carotid arteries, the internal carotid arteries (ICAs), external carotid arteries, intraosseous portion of the ICAs, vertebral arteries, the basilar artery, anterior cerebral arteries, middle cerebral arteries, and posterior cerebral arteries by using a 5-point scale. A score of 5 corresponded to excellent vessel delineation; 4, to good vessel delineation; 3, to adequate vessel delineation; 2, to marginally acceptable vessel delineation; and 1, to unacceptable vessel delineation.
The readers were asked to evaluate the visibility of small arterial detail on the basis of the depiction of small arteries, such as the ophthalmic, the superior cerebellar, and pericallosal arteries with a 5-point scale. A score of 5 corresponded to excellent vessel delineation; 4, good vessel delineation; 3, adequate vessel delineation; 2, marginally acceptable vessel delineation; and 1, unacceptable vessel delineation.
The readers were asked to assess subjective image artifacts on a 5-point scale. Five corresponded to a complete absence of imaging artifacts; 4, to mild artifacts not interfering with diagnostic decision making; 3, to moderate artifacts slightly interfering with diagnostic decision-making; 2, to pronounced artifacts interfering with diagnostic decision-making, though it was still possible to arrive at a diagnosis; and 1, to a situation in which artifacts completely hindered diagnostic decision making.
In addition, the readers were asked to rate their certainty of diagnosis on a 5-point scale. Five corresponded to a full and confident certainty of diagnosis based on the results of CT angiography alone; 4, a good certainty of diagnosis based on the results of CT angiography alone, though additional imaging would increase the certainty of diagnosis; 3, an adequate certainty of diagnosis based on the results of CT angiography alone, though additional imaging would be desirable; 2, a marginal certainty of diagnosis based on the results of CT angiography alone, though additional imaging would be required to establish the diagnosis; and 1, a situation in which the diagnosis was uncertain on the basis of the results of CT angiography alone. A score of ≥3 was considered as an acceptable level of artifacts or as constituting adequate diagnostic acceptability.
Radiation Dose
The fundamental radiation dose parameter in CT is the CT dose index (CTDI). The volume CTDI (CTDIvol), a derivative of the CTDI, can be used to express the average dose delivered to the scan volume in a specific examination. The dose-length product (DLP) provides an indication of the energy imparted to the organs and can be used to assess the overall radiation burden associated with a CT study.
To compare radiation exposure with combined ATCM and FTC techniques, we recorded the DLP, as a CT radiation dose descriptor, for each study. The CTDIvol was calculated from the following equation: DLP = CTDIvol × scan length (cm).
Statistical Analysis
The ages, maximum transverse diameter, and scan lengths between combined ATCM and FTC were compared for study homogeneity. Image quality scores for the delineation of various arterial vessels, visibility of small arterial details, subjective image artifacts, and certainty of diagnosis were compared between combined ATCM and FTC. Furthermore, quantitative image noise values in the muscles at 2 anatomic levels, CTDIvol and DLP, for examinations performed with the different techniques were also compared. The Mann-Whitney U test was used to assess differences between these values (Statistical Package for the Social Sciences, Version 12.0; SPSS, Chicago, Ill). A value of P < .05 indicated a statistically significant difference.
Results
The ages, length of scan coverage, and maximum transverse neck diameters of the subjects between the combined ATCM and FTC were not significantly different (P > .05, Table 1).
Age, scanning length, and maximum transverse neck diameter of subjects for examinations performed using FTC and combined ATCM
The scores of subjective image-quality parameters, such as vascular delineation of various arterial vessels, visibility of small arterial detail, image artifacts, and certainty of diagnosis, for studies performed with FTC and combined ATCM are summarized in Table 2. Again, no significant difference was detected in image quality parameters for examinations performed with the 2 different techniques (Fig 2 and Table 2). All subjects in both study groups had acceptable subjective artifacts or adequate diagnostic acceptability according to both readers, but reader 1 reported a suboptimal study for subjective artifacts in 1 subject with the FTC due to mild motion artifacts and metallic artifacts from the teeth.
The scores for image quality parameters for examinations performed using FTC and combined ATCM
Evaluation of craniocervical CTA with combined ATCM (A–C) and FTC (D–F). Transverse images through the shoulder level and upper neck and 3D reconstructions of CTA show that the subjective image quality of the 2 techniques is comparable.
We found no significant difference in the objective image noise values at the shoulder level between the study groups (P = .077). However, mean noise measurements at the upper neck were significantly greater with combined ATCM (P = .003, Table 3). Noise measurements at the shoulder level were much greater than those at the upper neck for both the combined ATCM and FTC groups.
Objective image noise for examinations performed using FTC and combined ATCM
A significant reduction in radiation dose (18% reduction for both CTDIvol and DLP) was noted with the combined ATCM compared with the FTC (P = .001 for CTDIvol and P = .002 for DLP, Table 3).
Discussion
With the rapid development of MDCT technology and increasing concern over the associated radiation dose, optimization of scanning techniques to maintain diagnostic image quality at the lowest possible radiation dose has become very important.17–20
Scanning parameters that affect CT radiation dose include x-ray beam energy (tube potential, kilovolt [peak]), tube current–time product (milliampere-second), pitch, section thickness, and scanning volume. Tube current is an important determinant of radiation dose and image quality in x-ray–based examinations. The radiation dose is linearly related to the current-time (milliampere-second) value when all other factors are held constant. However, a reduction of radiation dose will increase image noise, which may compromise image quality to a variable extent.2,12,17,19,21,22
Previous studies on CT of the head and neck, chest, abdomen, and pediatric pelvis have suggested the possibility of reducing tube currents without jeopardizing imaging quality,17,19,23–28 but the tube current in those studies was adjusted manually.
Such manual adjustment of tube current, based on patient weight or dimensions, can aid in establishing an appropriate balance between image noise and radiation exposure. However, these adjustments do not guarantee constant image quality throughout the examination.
Recent advances in CT technology, including implementation of combined ATCM, allow reduced radiation exposure during CT examinations. ATCM techniques allow maintenance of constant image quality with a reduced radiation exposure level because ATCM responds rapidly to large variations in beam attenuation. ATCM is based on the principle that x-ray attenuation and quantum image noise are determined by the size of the object and its tissue attenuation. Thus, the technique aims to modulate the tube current on the basis of regional body anatomy for adjustment of x-ray quantum noise to maintain constant image noise, with improved dose efficiency.
Compared with the chest and abdomen, craniocervical CTA by using a 64-detector MDCT spans a body region with wide variations in body size and attenuation patterns. With an FTC technique, a low-dose CT may result in suboptimal-quality images at the shoulders, whereas a higher tube current may result in greater exposure in the neck region. Because ATCM automatically adjusts the tube current on the basis of size and attenuation of the body region, it can automatically increase the tube current for the shoulder region and head and reduce it for the neck.
In previous studies, investigators have reported a substantial reduction in radiation dose with the ATCM technique. One study reported a 33% mean tube current–time product reduction, with similar artifacts and diagnostic acceptability, for abdominal and pelvic CT by using z-axis tube current modulation compared with an FTC technique.29 A radiation dose reduction of 56%–77% has been reported for urinary tract stone CT studies by using z-axis modulation.20 Kalra et al16 reported an 18%–26% radiation dose reduction for a chest CT study by using z-axis modulation. Namasivayam et al11 compared the radiation dose and image quality of z-axis tube current modulation with those of FTC for MDCT evaluations of the neck. They reported a significant radiation dose reduction in the z-axis tube current modulation technique, with similar subjective artifacts and diagnostic acceptability.
Rizzo et al13 reported that the use of a combined modulation technique resulted in a substantial reduction in the radiation dose, with acceptable image artifacts and diagnostic acceptability, compared with using a constant tube current, in scans of the abdomen and pelvis. Graser et al30 compared the radiation dose and image quality of combined ATCM with angular tube current modulation alone in CT colonography. They showed that combined ATCM resulted in a significant reduction in radiation exposure in CT colonography, without loss of image quality. Implementation of the x-y-z axis dose modulation (combined dose modulation) technique for neuroradiology CT examinations also revealed substantial dose reduction while maintaining image quality, compared with no dose modulation or z-axis dose modulation only.31 However, this study could not provide the magnitude of dose reduction on 64-section MDCT scanning because the comparison studies were performed on 16-section MDCT.
To our knowledge, no previous report has included comparisons of image quality and radiation doses on craniocervical CTA at 2 different tube current settings, the FTC and a combined ATCM. Our study showed that the combined modulation technique for craniocervical CTA provided similarly acceptable levels of depiction of the craniocervical vessels and diagnostic acceptability, as well as a reduction in radiation dose, compared with the FTC technique. In our results, objective noise at the shoulder level was greater than that in the upper neck for both the combined ATCM and FTC groups. This was due to the wide and higher beam-attenuating shoulders being in the scanning field. Because the shoulders were not the primary area of interest, we believe that better image quality of the shoulders need not be obtained by using a higher tube current. We saw no significant difference in the objective image noise values at shoulder level between the study groups, but objective noise in the upper neck was significantly higher with combined ATCM. The tube current in the shoulder level was the same (300 mA) for the 2 groups, but the combined ATCM technique automatically lowered the tube current for the neck, whereas the FTC had a constant tube current. We believe that this caused an increase in objective noise in the upper neck with the combined ATCM.
Our study has following limitations: First, different patients were assessed with combined ATCM and FTC modulation techniques because radiation concerns prohibited evaluating the same patients by using >1 method. However, the scanning length, maximum transverse neck diameter, and ages of the patients who underwent craniocervical CTA with the 2 different techniques were not significantly different. Second, our study was targeted at the efficacy of combined ATCM for craniocervical CTA in showing depictions of vascular structures. Its effects on diagnostic information, such as vascular stenosis or aneurysms, remain to be evaluated. Third, we neither included a comparison of the mean tube current–time product nor estimated the effective dose for combined ATCM and FTC techniques. However, currently, CTDIvol and DLP are the standard parameters used to describe CT-associated radiation doses; thus, we considered these useful in assessing the radiation dose.32,33 Last, the cross-sectional dimensions and weight (or body mass index) of the patients may provide better criteria for optimizing scanning parameters without losing relevant diagnostic information because the distance of the pathway traversed by the x-ray beam determines its attenuation.
In this study, we did not correlate the objective image noise, subjective image quality, and radiation exposure with patient weight or cross-sectional dimensions. We believe that such correlations should be assessed in a larger population study. Lowering the tube potential (kilovolt [peak]) can effectively reduce the radiation dose at MDCT examinations without substantially decreasing image quality.34–37 Other scanning parameters such as pitch, section thickness, and scanning volume can also affect radiation dose. Therefore, in addition to the combined ACTM technique, there may still be some room to reduce the radiation dose further though kilovolt (peak), pitch, section thickness, and scanning time optimization.
Recently, GE Healthcare introduced a new CT reconstruction technique, the Adaptive Statistical Iterative Reconstruction (ASIR) algorithm. The ASIR algorithm uses statistical noise profiles in an iterative manner to ramp image clarity and suppress noise compared with traditional filtered back-projection (FBP) techniques. The ASIR algorithm may significantly improve the reconstructed image quality of certain examinations that usually are regarded as very limited for proper interpretations due to image noise such as large patient and low-dose examinations compared with traditional FBP. In other words, use of the ASIR can provide comparable image quality even with a significantly reduced radiation dosage. However, the ASIR algorithm needs substantial computational processes and may require longer reconstruction time than the FBP technique. Therefore, although the ASIR algorithm has theoretic benefits of less noisy reconstructed images and the potential for significant radiation dose reduction on CT examinations, further clinical study is needed to determine its efficacy.
In conclusion, combined ATCM for craniocervical CTA resulted in a substantial dose reduction (18%), with similar diagnostic acceptability and subjective image artifacts compared with the FTC technique, though a small increase was observed in objective image noise in the upper neck with combined ATCM.
Acknowledgments
We thank Dr Dae Sub Choi and Toronto Western Hospital CT technologists, including Behruz Karasfi, Christina Sinclair, and Mike Verrall, for assistance in data collection and valuable advice.
References
- Received February 5, 2009.
- Accepted after revision April 17, 2009.
- Copyright © American Society of Neuroradiology