| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Chest Imaging |
1 Department of Radiology, Shanghai Pneumology Hospital, Shanghai, People's
Republic of China.
2 Department of Radiology, First Affiliated Hospital of Hua-xi Medical Center,
West China University of Medical Sciences, Shichuan, People's Republic of
China.
3 HealthONE Alliance, 5125 Stapleton Dr. N, Denver, CO 80216.
Received June 19, 2003; accepted after revision May 13, 2004.
Address correspondence to Z. Huang
(zheng_huang{at}msn.com).
OBJECTIVE. The aim of our study was to evaluate CT scanning protocols to determine how best to minimize patient exposure to ionizing radiation while maintaining sufficient image quality to detect pulmonary diseases.
SUBJECT AND METHODS. The CT dose index (CTDI) was determined by scanning an acrylic phantom at various tube current–time products (7.5–115 mAs). Image quality was evaluated by comparing the homogeneity and noise level of CT scans obtained in the acrylic phantom with those obtained in a water-equivalent phantom. The CT scans obtained at various milliampere-second settings in patients with nodules or diffuse opacifications were assessed. The relationships between the CTDI and the image quality of the CT scans (noise level and artifacts) were established.
RESULTS. The reduction of a conventional tube current–time product (115 mAs) by 65%, 78%, or 93.5% can decrease the CTDI by 60%, 70%, or 85%, respectively. In correlating the image quality of each CT scan to the milliampere-second settings used to obtain it, we found that homogeneity decreased as milliampere-second settings decreased, whereas the noise level increased as milliampere-second settings decreased. For both 8- and 3-mm slice thicknesses, the homogeneity of CT scans acquired at 7.5 mAs or greater was within the acceptable range (< 4 H). However, the noise level of CT scans remained within the acceptable range (< 0.35%) when 25 mAs or greater was used. The evaluation of the image quality of the patients' CT scans indicated no statistical significance in image quality rating between the scans obtained at 25 mAs and those obtained at 115 mAs (p > 0.01).
CONCLUSION. Low-dose (i.e., 40 or 25 mAs) helical chest CT produced satisfactory image quality and reduced the CTDI, thereby maximally protecting patients from radiation exposure.
Chest X-ray CT can reveal the precise size, location, and other features of a pulmonary lesion and is therefore considered the most accurate and convenient imaging technique to assist in the diagnosis of pulmonary diseases. Increased use of CT scans in recent years raises the potential to add to the radiation burden of the general population [1, 2]. The CT settings routinely used for adults might place children at higher risk of overexposure to ionizing radiation [3–6]. The principle of ALARA (as low as reasonably achievable) suggested by the International Commission of Radiological Protection [7, 8] has become more relevant because of the growing use of CT for diagnostic and interventional procedures. Clearly, the collective efforts of radiologists, physicists, and CT scanner manufacturers are required to reduce unnecessary radiation exposure from CT [9, 10]. Such efforts are particularly needed in developing countries because of recent wide-spread availability of CT scanners, the trend toward mass screening for pulmonary diseases, and the large population base.
Radiation dose is directly proportional to the tube current at a fixed peak tube voltage, scanning time, and slice width [11]. Thus, reduction in tube current or X-ray flux can lower the radiation dose received by patients. Although low- to ultra-low-dose chest CT has been performed at a reduced tube current-time product in some studies, low-dose CT has been used primarily for lung cancer screening and pediatric patients [12–20]. However, this approach has not been implemented in routine practice globally [9]. Some studies have suggested that using a lower milliampere-second setting has some potential disadvantages, such as more incidence of artifacts and false-negative results [20, 21].
The helical CT scanner can acquire images continuously and therefore shorten scanning time, lower radiation exposure, and improve diagnostic accuracy. Several studies have shown that reducing tube current–time product to a range of 50–110 mAs does not substantially compromise the image quality of chest CT scans, mainly because of the intrinsic contrast in the chest and lower pulmonary absorption of radiation [12, 13, 16, 20]. However, these previous studies lacked a precise assessment of image quality at the proposed lower dose. The relationships among the exposure parameters, CT dose index (CTDI), and CT image quality in low-dose CT need to be established. In this study, we assessed the rationale and feasibility of low-dose helical CT by establishing the relationship between CTDI and CT image quality (noise level and artifacts), comparing the image quality of CT scans obtained at lower milliampere-second values (7.5–40 mAs) in phantoms and patients, and evaluating the effect of low-dose CT on the postprocessed image quality. Our objective was to devise a CT protocol that would minimize patient exposure to ionizing radiation while maintaining the sensitivity and specificity of CT for detecting pulmonary diseases.
Subjects and Methods
CT Scanner
All CT scans were obtained on an ultra-high-speed MDCT scanner
(Asteion-Multi, Toshiba) using the basic imaging parameters of 120-kVp tube
voltage, 0.75-sec scanning time, and 360° table rotation, except for
helical scans (135 kVp, pitch of 3).
Phantom Studies
CTDI measurement.—The CTDI was determined in a CTDI
dosimetry phantom (
= 32 cm) with a radiation meter (T6580, Institute of
Metrological Science) equipped with an ionization chamber detector (T6C-LA,
Institute of Metrological Science). This solid acrylic phantom ( = 32
cm; 76-415, Victoreen) has five probe holes (i.e., one in the center and four
at the periphery—left, right, top, and bottom) for hosting the detector.
The CTDI phantom was placed in the center of the radiation field and received
115, 40, 25, or 7.5 mAs, at slice thicknesses of 8 and 3 mm. An average of
three measurements were obtained for each exposure dose.
CTDI was defined as [22]:
 | (1) |
In practice, the CTDI of the chest (CTDIw) was derived
from the CTDI values of center regions (CTDIcenter) and
peripheral regions (CTDIperiphery) according to the
following formula:
 | (2) |
CT Scan Quality Assessment
The image quality of the scans obtained with low-dose CT was determined by
the comparison of CT values of CT scans obtained in a solid and
water-equivalent phantoms (Model 461A, Gammex RMI). The water-equivalent
phantom was placed in the center of the radiation field and received 115, 40,
25, 15, or 7.5 mAs. Four images with slice thicknesses of 8 or 3 mm were
acquired for each milliampere-second setting. Five areas of 100 mm2
each—one from the center and four from four peripheral areas (10 mm away
from the edge of the image)—were selected from the identical location of
each image. The CT value and SEM were calculated. The homogeneity (U)
of CT image (in Hounsfield units) was determined according to the following
equation:
 | (3) |
 | (4) |
Patient CT Scan Quality Assessment
After receiving patient consent and institutional review board approval, we
compared the image quality of CT scans obtained with low doses in 60 patients
with one or more pulmonary nodules (
3 cm) (n = 30 patients) or
diffuse opacifications (n = 30 patients). We obtained multiple chest
CT scans in each patient at 115, 40, 25, 15, or 7.5 mAs with a slice thickness
of 8 or 3 mm. Patient age ranged from 29 to 76 years old (mean, 43.5 years; 48
men and 12 women).
All of the patients' CT scans were displayed on the cathode-ray tube monitors and divided using both mediastinal (level, 50 H; width, 400 H) and lung window settings (level, 500 H; width, 1,200 H), and photographed with a laser camera (Scopix LR-5200, Agfa). The quality of each CT scan was rated by four reviewers in a single-blinded regimen according to the following evaluation criteria: normal image quality (no artifacts), mild artifact (does not affect rendering an accurate diagnosis), and severe artifact (does affect rendering an accurate diagnosis) (Figs. 1A, 1B, and 1C). CT parameters were blocked during the rating session. The ratings for each group of scans obtained at the various milliampere-second settings and slice thicknesses from four reviewers were pooled together and the proportions of image quality level were calculated. The statistical analysis of proportions among the five milliampere-second groups of scans was performed with the chi-square test.
|
|
|
Of 60 patients in our study, 15 (age range, 43–65 years; mean, 56.3 years; nine men and six women) also underwent whole-lung helical CT (135 kVp, 0.75 sec, 3-mm slice thickness, and a pitch of 3) and perfusion scanning. One hundred milliliters of contrast material (Omnipaque 300 [iohexol], Nycomed) was injected through a catheter in the right cephalic vein at a rate of 2 mL/sec. Perfusion scanning was first performed at 190 and 40 mAs so that we could determine the time required for the visualization of the blood vessels and then at 150, 25, and 15 mAs. Postprocessing scans were reconstructed at both 3- and 1-mm intervals. The high-resolution CT scans, reconstructed at 1-mm intervals for both the coronary and sagittal views, were compared with the reconstructed perfusion scans obtained at 150, 25, and 15 mAs. In addition, the image postprocessing, including computer volume radiography, maximum intensity projection, 3D, and CT virtual endoscopy, were performed on a high-performance diagnostic workstation (Alatoview, Toshiba). The effect of the various milliampere-second settings on the image postprocessing was compared. Each radiographic exposure group subjected to different reconstruction methods was rated as described above.
Results
CTDI
The CTDI was determined on a CTDI phantom using various milliampere-second
settings and two slice thicknesses: 8 and 3 mm. An average CTDI was calculated
from three interpretations performed in accordance with equations
1 and
2. CTDI was normalized to the
CTDI value of 115 mAs (Table
1). The CTDI showed a linear correlation with the
milliampere-second settings (Figs.
2 and
3). Compared with the CTDI
obtained at 115 mAs, the reduction of CTDI was 62%, 74%, and 88% at 40, 25,
and 7.5 mAs, respectively, for 8-mm slice thickness. For 3 mm, the reduction
was 61%, 72%, and 86% at 40, 25, and 7.5 mAs, respectively.
|
|
|
Effects of Exposure Dose on the Homogeneity and the Noise Level of CT Scans
The homogeneity and noise level were measured at various CT exposure
parameters and used for quantitative evaluation of the image quality of the CT
scans. The homogeneity and the noise level were obtained by analyzing five
areas of each scan under various exposure parameters according to equations
3 and
4. The results showed that the
homogeneity and the noise level were affected by the milliampere-second
values. The reduction of the exposure dose caused a decrease of homogeneity
and an increase of noise level (Tables
2 and
3). The homogeneity values
derived from the tested milliampere-second values and both slice thicknesses
reached the requirement set by the National Metrological Bureau (< 4 H)
[22]. However, the noise level
of three exposure parameters—7.5 mAs with an 8-mm slice thickness, 7.5
mAs with a 3-mm slice thickness, and 25 mAs with a 3-mm slice
thickness—were equal or greater than the acceptable value set by the
National Metrological Bureau (0.35%)
[22]. Compared with the linear
decline of CTDI, the noise level began to deteriorate rapidly at settings of
less than 25 mAs (graph not shown). These results imply that the image quality
of the CT scans remained at an acceptable level when the exposure dose or the
tube current was 25 mAs or greater.
|
|
Evaluation of Patient CT Scans
Sixty patients with nodules (n = 30) or diffuse opacifications
(n = 30) were selected for this study. CT scans with a slice
thickness of 8 mm or 3 mm were obtained with various milliampere-second
settings and with both mediastinal and lung window settings. The CT image
quality was rated by four experienced reviewers. The results showed that the
percentages of normal-quality images (no artifacts) in the 15-mAs and 7.5-mAs
groups of scans were significantly lower than those of other groups (i.e.,
those acquired at 115, 40, and 25 mAs) (Tables
4 and
5). These findings imply that
the image quality of CT scans obtained at 15 mAs and 7.5 mAs was
compromised.
|
|
A further statistical analysis of the scans obtained at 115, 40, and 25 mAs showed that there was no statistically significant difference in image quality among these groups. For the scans obtained with the mediastinal window settings, the chi-square statistic was 8.43 (0.1 > p > 0.05; df = 4) for the slice thickness of 8 mm and 9.87 (0.1 > p > 0.02; df = 4) for the slice thickness of 3 mm. For scans obtained with the lung window settings, the chi-square statistic was 9.15 (0.1 > p > 0.05; df = 4) for the slice thickness of 8 mm and 8.23 (0.5 > p > 0.1; df = 4) for the slice thickness of 3 mm.
Figures 2 and 3 illustrate the trend of deteriorating image quality of scans obtained at low doses. Compared with the linear decline of CTDI, the image quality of CT scans, indicated as the percentage of normal-quality images, remained stable at 40 and 25 mAs for both the mediastinal and lung window settings. The percentage of normal-quality images acquired with an 8-mm slice thickness exceeded 80%, and the percentage acquired with a 3-mm slice thickness exceeded 75%. The percentage of normal-quality images with the lung window setting was higher than that with the mediastinal window setting. However, image quality of scans obtained with both the lung and mediastinal window settings declined rapidly when the milliampere-second value was reduced to 15 or 7.5 mAs. We believe that it is reasonable to consider that 25 mAs is an inflection point of the percentage curve of normal-quality images. The visual comparison of images also showed that there was no significant difference in image quality between scans obtained at 115 mAs and those obtained at 25 mAs; both were rated as normal-quality images (Figs. 4A and 4B).
|
|
Various types of image postprocessing were performed on the scans of 15 patients at 3- and 1-mm intervals. We found that the smaller the interval was, the better the image quality of the postprocessed scans was and the less the effect that milliampere-second setting had on the image quality (data not shown). Lower milliampere-second settings, such as 40 or 25 mAs, had little effect on the image quality of reconstructions derived from the images obtained at these settings. Figures 5A and 5B shows representative coronary reconstruction images obtained at 115 and 25 mAs, with 1-mm intervals. Both were rated as normal-quality images. No significant difference in image quality was observed in scans obtained at milliampere-second settings equal to or greater than 25 mAs. Although 3D scans obtained at 25 mAs were slightly coarse compared with those obtained at 115 mAs, the coarseness did not affect the diagnosis (Figs. 6A and 6B). Figures 7A and 7B shows representative CT virtual endoscopic images acquired at 187 and 40 mAs; both were rated as normal-quality images.
|
|
|
|
|
|
Discussion
The radiation dose delivered during CT scanning is related to tube current, voltage, scanning time, slice thickness, scanning volume, and pitch. Although the total scanning time is shortened on new models of CT scanners, the radiation exposure becomes greater mainly because of the higher tube currents and larger scanning volume of the helical CT scanner [1]. Several new technologies used to improve X-ray beam collimation, filter design, or tube current modulation also play a role in reducing the radiation dose of CT scans. However, these complicated procedures often require modification of the scanner.
The radiation dose is directly proportional to the tube current–time product at a fixed peak tube voltage and slice thickness. Therefore, reducing tube current is a simple and practicable means to decrease the radiation exposure. Recently, several studies have shown that it is possible to perform low-dose chest CT (at 10–140 mAs) without severely compromising the image quality needed for pulmonary nodule detection [9, 12, 16, 21, 23–28]. These studies suggest that this simple and effective technique might also be used for the CT examination of other pulmonary abnormalities. In this study, we applied this simple technique and showed that there is a linear correlation between the tube current–time product and absorbed dose at a constant voltage and scanning time with the use of a MDCT scanner (Asteion-Multi) (Table 1). We also explored the rationale of using low-dose chest CT to detect pulmonary nodules and diffuse opacifications and assessed the feasibility of optimizing radiation protection by reducing the CT tube currents.
By obtaining CT scans at various milliampere-second settings and quantitatively evaluating the absorbed dose and the CT image quality, we established the relationship between the CT exposure parameters and the CTDI and the relationship between the CT exposure parameters and the image quality of a CT scan. By comparing the reduction of the CTDI and the deterioration of image quality at reduced milliampere-seconds, we determined the lowest acceptable tube current–time product for low-dose chest CT to decrease the risk of radiation exposure for patients.
The reduction in milliampere-seconds caused a linear decrease in the absorbed dose. The image quality was preserved when the milliampere-second value was 25 mAs or greater, but it started to deteriorate quickly when the value fell below 25 mAs (Figs. 2 and 3). A slight difference was noted between the two tested slice thicknesses in the trend of deterioration of image quality. The thicker the scanning slice, the slower the deterioration rate. Nevertheless, the quality curve of the CT scans obtained from the two slice thicknesses showed that the percentage of the normal-quality image fell quickly at the inflection point of 25 mAs. Statistical analysis indicated no significant difference between the quality of scans obtained at 115, 40, and 25 mAs. Thus, 25 mAs or greater is considered to be an acceptable exposure parameter to ensure satisfactory image quality for chest CT scans. Low-dose CT performed at 25 mAs reduces the CTDI by approximately 70% compared with the CTDI delivered 115 mAs (Table 1).
The percentage of normal-quality images obtained with the lung window setting was higher than the percentage of such images obtained with the mediastinal window setting under the same scanning conditions (Figs. 2 and 3). This finding implies that the variation in the noise level caused by the reduction of milliampere-seconds is lower in the lung window setting, although varying window settings could have improved the images obtained at lower milliampere-seconds. Therefore, a lower milliampere-seconds value could be used to examine lesions at the lung window setting without compromising image quality.
Postprocessing and high-precision image reconstruction can display detailed stereoscopic and highly realistic images, thereby enabling multitechnique diagnosis and improved clinical value of routine CT scans. However, the effect of the reduction in milliampere-seconds values on the quality of reconstruction images has not been reported, to our knowledge. In this study, we examined the image quality of reconstruction images obtained at various milliampere-seconds values and processed on Alatoview, a high-performance diagnostic workstation that is capable of processing and displaying CT images and allowing multitechnique diagnosis through its fusion function. Our preliminary data indicate that obtaining scans at lower milliampere-seconds values, such as 40 or 25 mAs, had little effect on the quality of reconstructed images (maximum-intensity-projection, computer volume radiographic or CT virtual endoscopic images) based on those scans but would reduce the X-ray exposure dose three- to fivefold. Although 3D images obtained at 25 mAs were slightly coarse compared with those obtained at 115 mAs, the coarseness did not affect the accuracy of the diagnosis (Figs. 6A and 6B). It was noted that the clearer the reconstruction image, the less effect the milliampere-seconds values had on image post-processing. This finding suggests that in order to maintain the image quality, image reconstruction should be performed at a smaller interval for maximum-intensity-projection or computer volume radiographic images.
Although our general consensus is that 25 mAs is an acceptable and reasonable low-scanning parameter for the Asteion-Multi helical CT scanner, scanning parameters need to be adjusted for different CT systems. In addition, biologic variations in patients and in lesion locations affect image quality more with low-dose scanning than with high-dose scanning. In general, the dose needs to be increased for obese patients. The dose also needs to be increased for examining the upper lobe due to the false shadow caused by the scapula [27, 29].
As suggested by previous studies [16, 23, 26, 29], the low-dose CT scan has several advantages in screening examinations for high-risk populations, such as lung cancer, tuberculosis, and fibrosis. Low-dose CT can significantly reduce the patient's radiation exposure and is superior to conventional chest radiography for discovering and identifying lung diseases. The exposure dose received during acquisition of three or four low-dose chest CT scans is equivalent to that received during acquisition of one routine chest CT scan. The low-dose CT scan is also suitable for special populations such as pediatric patients and pregnant women because of the decreased risk of long-term adverse effects. The reduction of the radiation dose also lowers the risk of damage to CT tubes and detectors, which subsequently reduces the operational cost and prolongs the lifetime of the cathode-ray tubes and detectors.
In conclusion, our results suggest that it is feasible to follow the ALARA principle and to optimize the radiation protection for patients during chest CT by reducing the radiation dose, thereby lowering the patient's risk of radiation exposure while ensuring an adequate image quality.
Acknowledgments
We thank Ken Dole and Jill Beckers for critically reviewing this manuscript.
References
CiteULike 
Complore 
Connotea 
Del.icio.us 
Digg 
Reddit 
Technorati What's this?
This article has been cited by other articles:
|
|
 |
|
  L. M. Hurwitz, T. T. Yoshizumi, P. C. Goodman, R. C. Nelson, G. Toncheva, G. B. Nguyen, C. Lowry, and C. Anderson-Evans Radiation Dose Savings for Adult Pulmonary Embolus 64-MDCT Using Bismuth Breast Shields, Lower Peak Kilovoltage, and Automatic Tube Current Modulation Am. J. Roentgenol., January 1, 2009; 192(1): 244 - 253. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
T. Kubo, P.-J. P. Lin, W. Stiller, M. Takahashi, H.-U. Kauczor, Y. Ohno, and H. Hatabu Radiation Dose Reduction in Chest CT: A Review Am. J. Roentgenol., February 1, 2008; 190(2): 335 - 343. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
J. D. MacKenzie, J. Nazario-Larrieu, T. Cai, M. S. Ledbetter, M. A. Duran-Mendicuti, P. F. Judy, and F. J. Rybicki Reduced-Dose CT: Effect on Reader Evaluation in Detection of Pulmonary Embolism Am. J. Roentgenol., December 1, 2007; 189(6): 1371 - 1379. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 W Mazrani, K McHugh, and P J Marsden The radiation burden of radiological investigations Arch. Dis. Child., December 1, 2007; 92(12): 1127 - 1131. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 M. J. Gollub, R. Hong, D. M. Sarasohn, and T. Akhurst Limitations of CT During PET/CT J. Nucl. Med., October 1, 2007; 48(10): 1583 - 1591. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 A. Madani, V. De Maertelaer, J. Zanen, and P. A. Gevenois Pulmonary Emphysema: Radiation Dose and Section Thickness at Multidetector CT Quantification--Comparison with Macroscopic and Microscopic Morphometry Radiology, April 1, 2007; 243(1): 250 - 257. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
H. Kuehl, P. Veit, S. J. Rosenbaum, A. Bockisch, and G. Antoch Can PET/CT Replace Separate Diagnostic CT for Cancer Imaging? Optimizing CT Protocols for Imaging Cancers of the Chest and Abdomen J. Nucl. Med., January 1, 2007; 48(1_suppl): 45S - 57S. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
= CTDI, 
=
mediastinal window setting (level, 50 H; width, 400 H), 
= lung window
setting (level, 500 H; width, 1200 H), r2 = 0.997.
