| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Experimental Studies |
1 From the Department of Radiology, Beth Israel Deaconess Medical Center, Harvard Medical School, 330 Brookline Ave, Boston, MA 02215. Received June 17, 2002; revision requested August 14; final revision received October 29; accepted November 5. Address correspondence to P.M.B. (e-mail: pboisell@caregroup.harvard.edu).
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONMATERIALS AND METHODS RESULTS DISCUSSIONREFERENCES |
|---|
MATERIALS AND METHODS: Five dogs each underwent five consecutive helical CT examinations (one single–detector row CT examination at 240 mA and four multi–detector row CT examinations at 240, 120, 40, and 20 mA), with 0.8-second gantry rotation time. Multiplanar reformation and 3D reconstruction images were created from each of the 25 CT acquisitions. The images were randomized and blindly reviewed with consensus agreement by three radiologists who graded image quality by using a five-point scale. In a separate review, the three radiologists independently used a four-point scale to rank the comparative image quality of the multi–detector row CT 3D images, while blinded to specific tube currents. The radiation doses were measured for each type of scan, and the relative radiation dose length products that were normalized to single–detector row CT values were used to compare radiation doses of the various CT techniques. Statistical analysis was performed with the Wilcoxon signed-rank test and the Friedman analysis of variance test.
RESULTS: Image quality was consistently ranked higher for multi–detector row CT images than for single–detector row CT images (P = .03). Although there were no distinguishable differences between images obtained with multi–detector row CT at 240, 120, or 40 mA, images obtained with 20 mA were given a significantly (P = .04) lower relative rank (mean, 2.4) than those obtained with higher tube currents (mean, 1.4–1.7). Multi–detector row CT radiation doses were 1.64, 0.82, 0.27, and 0.14 (for 240-, 120-, 40-, and 20-mA multi–detector row CT, respectively) relative to the dose for 240-mA single–detector row CT.
CONCLUSION: Multi–detector row CT is superior to single–detector row CT for multiplanar and 3D imaging of the central airways. Substantial dose reductions can be made, while maintaining high image quality.
© RSNA, 2003
Index terms: Animals • Bronchi, CT, 671.1211 • Computed tomography (CT), multi–detector row • Computed tomography (CT), radiation exposure • Trachea, CT, 671.1211
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODS RESULTS DISCUSSIONREFERENCES |
|---|
On the other hand, a relative disadvantage of multi–detector row CT is that it requires a higher radiation dose than does single–detector row CT, particularly when narrow beam collimation is employed (8). For example, when using a four-channel multi–detector row CT scanner, the narrow beam dose inefficiency can result in a 100% dose increase for the narrowest beam collimations (4 x 1 mm or 4 x 1.25 mm) (8). In recent years, there has been growing interest in the use of low-dose CT techniques in the assessment of lung parenchyma in both children and adults (9,10). Considering the inherent high contrast between the airways and adjacent mediastinal structures, we hypothesized that low-dose techniques could also be employed for multi–detector row CT studies of the central airways, without substantial loss of image quality. If image quality can be preserved with a low-dose technique, then the potential enhancement in image quality obtained with multi–detector row CT would not be offset by the cost of additional radiation exposure.
The purpose of our study was to compare in a canine model the quality of multiplanar reformation and 3D reconstruction images of the central airways obtained with single–detector row CT and multi–detector row CT at varied current settings and to compare radiation dose.
|
MATERIALS AND METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODS RESULTS DISCUSSIONREFERENCES |
|---|
Scanning Parameters
Each of five dogs underwent five consecutive helical CT examinations of the thorax. Single–detector row CT examinations were performed with a CTi scanner (GE Medical Systems, Milwaukee, Wis), with 3-mm collimation, 1.5-mm reconstruction interval, pitch of 1.5, 120 kVp, and 240 mA. Four consecutive multi–detector row CT examinations were performed with a QXi scanner (GE Medical Systems) with 240, 120, 40, and 20 mA. Imaging parameters for all multi–detector row CT scans were 2.5-mm collimation, 1.25-mm reconstruction interval, high speed mode, pitch of 1.5, and 120 kVp. Gantry rotation time was 0.8 second for both single–detector row CT and multi–detector row CT scanners, and yielded 192, 96, 32, and 16 mAs for 240, 120, 40, and 20 mA currents, respectively. All scanning was performed during quiet respiration, without suspended breath holding.
Multiplanar Reformation and 3D Reconstruction Techniques
The data from the 25 CT acquisitions were transferred to two workstations. Two-dimensional multiplanar volume reformation images were obtained in the oblique coronal plane along the axis of the trachea with an Advantage workstation (GE Medical Systems), and 3D external images of the central airways were obtained with a Vitrea workstation (Vital Images, Plymouth, Minn). All reconstructed images were generated with standard protocols by using preset algorithms. All images of a similar type (multiplanar reformation or 3D reconstruction) were displayed in the same projection and size on hard copy.
Image Interpretation
The images were analyzed in two ways. First, the images were randomized and blindly reviewed simultaneously by three radiologists (P.M.B., G.D., S.N.G.) who graded image quality with consensus agreement. The multiplanar and 3D images obtained with 240 mA single–detector row CT and multi–detector row CT acquisitions at varying currents were reviewed and graded in two separate groups. The images from each group were randomized so that the radiologists were unaware both of the imaging parameters (ie, type of scanner or current) and of the specific animal from which the images were obtained. All images were reviewed with hard-copy display. For each image, the radiologists rendered separate grades for image quality at the level of the trachea and at the main bronchi. Image quality was graded by using a five-point scale (1 = smooth margins, without ridging or stair-step artifact; 2 = slight rippling of surface edges without frank ridging or stair-step artifact; 3 = mild ridging or stair-step artifact; 4 = moderate ridging or stair-step artifact; 5 = severe ridging or stair-step artifact). To facilitate consistent use of the grading scale, a scoring template with visual examples of each grade of artifact was provided to the radiologists for reference during the grading process.
In a separate review, the three radiologists (P.M.B., G.D., S.N.G.) independently compared quality of the four 3D airway images obtained with multi–detector row CT and differing currents for each of the five animals. For this portion of the study, the radiologists were aware that all four studies were performed on the same animal, but they were blinded to the specific current used. Comparative image quality was ranked with a four-point scale (1 = best quality, 4 = worst quality), and equivalent rankings were allowed.
Statistical Analysis
The Wilcoxon signed-rank test was used to compare the image quality rankings of single–detector row CT and multi–detector row CT images. To compare the rankings of image quality of 3D images obtained with multi–detector row CT at differing currents, we employed the Friedman analysis of variance test, which is recommended for assessment of repeated-measures experiments. All P values less than .05 were considered to indicate a statistically significant difference.
Radiation Dosimetry
The radiation doses were measured by two physicists (J.C., H.K.) for the single–detector row CT and multi–detector row CT scanners by using a calibrated "pencil" ionization chamber (model 6000-100; Victoreen, Cleveland, Ohio) and associated calibrated electronics (NERO 8000; Victoreen). Dosimetry measurements were made with a standard polymethylmethacrylate CT dosimetery "body" phantom (model 76-414, 32-cm diameter; Nuclear Associates, Cleveland, Ohio).
Radiation exposure measurements were collected in roentgens and converted to dose-to-air in milligrays (8.764 mGy = 1 R). The radiation dose-length products (DLPs), which are the value of the dose times the total length of the examination, divided by the pitch of the scan, were calculated for the respective CT scanners by using the "weighted CT dose index" values, which are equal to one-third of the phantom center-dose CT dose index value added to two-thirds of the average phantom peripheral-dose CT dose index value. These calculated DLPs were compared with those obtained by using software (CTDosimetry; Impact, London, England) and UK National Radiological Protection Board Monte Carlo CT data sets (SR250; Radiological Protection Board, Chilton, England). The relative DLPs were used as the comparative measure of radiation dose between the various CT scanners and protocols.
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODS RESULTS DISCUSSIONREFERENCES |
|---|
|
|
|
|
|
|
Radiation Dosimetry
Control measures of the radiation output for all of the scanners were linear with milliampere seconds. Calculated DLPs were compared with those obtained by using software and CT data sets and showed excellent agreement (average of measured DLP to software DLP was 0.99). Table 2 lists the relative DLP results, which were normalized to the single–detector row CT value for comparison purposes. With the exception of the highest setting (240 mA) used for multi–detector row CT, all of the other settings resulted in lower radiation doses than single–detector row CT. For example, the 20-mA multi–detector row CT technique resulted in a relative radiation dose of 0.14 compared with the 240-mA single–detector row CT technique, whereas the 240-mA multi–detector row CT technique resulted in a relative radiation dose of 1.64.
|
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODS RESULTS DISCUSSIONREFERENCES |
|---|
Similar to our study, a previous laboratory investigation by Fleischmann et al (4) demonstrated that stair-step artifacts on volume-rendered images of an acrylic rod were significantly smaller with multi–detector row CT than with single–detector row CT. Unlike images of an acrylic rod, our images of a living animal model were subject to the effects of respiratory and cardiac motion, both of which negatively affect image quality (7). With these effects in mind, the faster speed of scanning of multi–detector row CT likely had a major effect on the improvement in image quality. With all other factors being equal, four-channel multi–detector row CT is four times faster than single–detector row CT (7). Fast scanning has practical implications for imaging patients with central airway disorders, many of whom have difficulty maintaining a prolonged breath hold (11).
Interestingly, image quality was consistently better at the level of the trachea than at the main bronchi with both single–detector row CT and multi–detector CT. This difference likely reflects the greater susceptibility of the main bronchi to imaging artifacts from respiratory and cardiac motion. Similarly, artifacts from cardiac motion on 3D central airway images of the main bronchi have been described in a recent article by Hoppe et al (12), who studied 20 patients with central airway disease who underwent multi–detector row CT. Recent advances in multi–detector row CT, including faster gantry rotation time and eight- and 16-channel scanner designs, will likely further reduce motion artifacts (13).
Although the technical parameters that were employed for the single–detector row CT in this study were nearly identical to those for multi–detector CT, a slightly thinner section thickness was used for multi–detector CT than for single–detector CT (2.5 vs 3.0 mm, respectively). When one considers the effect that helical CT has on section thickness, however, this difference is only 0.25 mm (3.2-mm vs 3.45-mm effective section thickness). Given the small size of this difference, it is assumed that this factor had only a minimal influence on the differences of image quality observed in this study.
While the quality of images obtained with multi–detector row CT is clearly superior, a relative limitation of this technology is its higher radiation dose compared with that of single–detector row CT, particularly when narrow beam collimation is employed (8). Our work demonstrates that substantial reductions in radiation dose can be obtained with multi–detector row CT while still maintaining high image quality. With all other factors held constant, patient radiation dose is directly proportional to x-ray tube current (8,14); therefore, a decrease in tube current from 240 to 120 mA results in a twofold decrease in dose, whereas a decrease from 240 to 40 mA results in a sixfold decrease. Notably, there were no distinguishable differences between the multi–detector row CT images obtained at settings of 40, 120 and 240 mA (corresponding to 32, 96, and 196 mAs, respectively).
The increase in radiation dose that was observed when multi–detector row CT was performed at the same current as single–detector row CT underscores the importance of carefully considering parameters such as tube current when using this new technology (8). Notably, the 120-, 40-, and 20-mA multi–detector row CT examinations resulted in relative radiation doses of 0.82, 0.27, and 0.14, respectively, compared with those of single–detector row CT. Thus, by modifying the tube current, one can use multi–detector row CT to obtain the benefit of high-quality central airway images, without the penalty of a higher radiation dose.
CT radiation dose has recently received considerable attention in the pediatric population (14–17). In response to growing concerns about excessive radiation exposure from CT in children, the recommended tube current for pediatric helical CT has progressively decreased over the past several years (15). Despite this decrease, however, results of a 2001 study by Paterson et al (16) suggested that standard adult tube current settings are still commonly used with pediatric patients. With regard to tube current recommendations for multi–detector row CT of pediatric patients, a dose of 56 mAs has recently been suggested for a patient of equivalent weight to the animal model used in our study (17). Our results suggest that milliamperes-second and, hence, radiation dose can be reduced by approximately half of this value for central airway studies, while still maintaining high image quality. Our findings also demonstrate that there is no improvement in image quality with use of higher doses that are typical for adult patients.
The benefit of using an animal model for this study is that it allowed us to assess the effect of differing CT parameters on image quality and radiation dose in a way that would not be possible in a human model because of radiation exposure concerns. The animal model employed in this study is an appropriate model for the pediatric population in terms of body weight, because it is similar to that of a child of approximately 10 years of age. Moreover, our method of imaging the animals during quiet breathing closely simulates the practice of imaging small children, who are unable to cooperate with suspended breath holding. Although central airway studies are ideally performed during a suspended breath hold, the results of this study show that multi–detector row CT provides high-quality central airway images, even during quiet breathing.
It should be noted that our study focused on image quality of the normal central airways in an animal model. Although our results show that low-dose techniques provide high-quality images of normal airways, this study did not assess the ability of such techniques to depict or characterize airway abnormalities. Thus, future studies in both animals and humans are necessary to determine whether low-dose techniques are equally efficacious as standard-dose techniques for detecting and characterizing central airway abnormalities.
In conclusion, multi–detector row CT is superior to single–detector row CT for multiplanar and 3D central airway imaging and allows for substantial radiation dose reductions, while maintaining high image quality. Because dose reduction techniques for multi–detector row CT more than offset the difference in dose between single–detector row CT and multi–detector row CT, we believe that dose-modified multi–detector row CT will prove to be the clear study of choice for central airway imaging.
Practical application: In an animal model, we have shown that multi–detector row CT is superior to single–detector row CT for multiplanar and 3D imaging of the central airways. Moreover, we have demonstrated that substantial dose reductions can be achieved with multi–detector row CT while maintaining high image quality. Thus, low-dose multi–detector CT has potential future applications in the imaging of the central airways, particularly in the pediatric population.
|
FOOTNOTES |
|---|
3 Current address: Department of Radiology, Massachusetts General Hospital, 55 Fruit St, Boston, MA 02114.
Abbreviations: DLP = dose-length products, 3D = three-dimensional
Author contributions: Guarantors of integrity of entire study, P.M.B., S.N.G.; study concepts, P.M.B., G.D., S.N.G.; study design, P.M.B., S.N.G.; literature research, P.M.B.; experimental studies, P.M.B., G.D., S.N.G., H.K., J.C., D.W.; data acquisition, P.M.B., G.D., S.N.G., H.K., J.C., D.W.; data analysis/interpretation, P.M.B., S.N.G., E.H., J.C.; statistical analysis, E.H.; manuscript preparation, P.M.B.; manuscript definition of intellectual content, P.M.B., S.N.G.; manuscript editing, revision/review, and final version approval, all authors.
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODS RESULTS DISCUSSIONREFERENCES |
|---|
This article has been cited by other articles:
|
|
 |
|
  K. Yamada, T. Taniura, S. Tanabe, M. Yamaguchi, S. Azemoto, and E. R. Wisner The Use of Multi-detector Row Computed Tomography (MDCT) as an Alternative to Specimen Preparation for Anatomical Instruction J Vet Med Educ, January 1, 2007; 34(2): 143 - 150. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 K. A. Carden, P. M. Boiselle, D. A. Waltz, and A. Ernst Tracheomalacia and Tracheobronchomalacia in Children and Adults: An In-depth Review Chest, March 1, 2005; 127(3): 984 - 1005. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| RADIOLOGY | RADIOGRAPHICS | RSNA JOURNALS ONLINE |
