| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Original Research |
1 Department of Radiology, Massachusetts General Hospital, Boston, MA
02114.
3 Department of Radiology, University College, Cork, Ireland.
Received September 24, 2004;
accepted after revision December 6, 2004.
Address correspondence to M. K. Kalra
(mannudeep_k_kalra{at}yahoo.com).
Abstract
 Top Abstract IntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
MATERIALS AND METHODS. "Extra" images for routine chest CT were defined as those above lung apices (supraapical) and those caudal to the lowermost portion of lung parenchyma (infrapulmonary), including images obtained beyond the adrenal glands (infraadrenal). One hundred and forty-eight consecutive chest CT examinations (70 men, 78 women; age range, 15-90 years) performed September 13-25, 2003, were reviewed to determine the number of supraapical, infrapulmonary, and infraadrenal extra images. All examinations were performed using z-axis modulation (n = 70) or fixed tube current (n = 78). The CT dose index volume and dose-length product (DLP) values for the extra images were calculated. Two radiologists reviewed these extra images for pathologic findings. Student's t test was used to perform the statistical analysis.
RESULTS. One hundred forty-four (97%) examinations had supraapical extra images and 145 (98%) had infrapulmonary extra images. A total of 31 additional findings were observed in extra images. Most clinically important findings were identified in patients with a history of malignancy. With z-axis modulation, the mean DLP for supraapical and infrapulmonary extra images was 39.98 mGy·cm and 132.59 mGy·cm, respectively. With fixed tube current, the mean DLP for supraapical and infrapulmonary extra images was 30.31 mGy·cm and 95.91 mGy·cm, respectively.
CONCLUSION. A substantial number of extra images are acquired during chest CT that do not add clinically important information in patients with nonmalignant indications. The use of z-axis modulation increased radiation dose for the extra images.
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
We hypothesized that "extra" images acquired with chest CT increase radiation dose to patients without adding substantial diagnostic information. Therefore, the purpose of this study was to determine the additional radiation dose associated with scanning beyond the anatomic limits of the thorax in standard chest CT protocol and to assess the effect of z-axis automatic tube current modulation on the additional radiation dose associated with the extra images.
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Consecutive routine chest CT examinations performed at a single institution during a 13-day period from September 13 to 25, 2003, were reviewed to determine the number of extra images acquired during chest CT. Supraapical, infrapulmonary, and infraadrenal extra images were recorded separately for each study. The study cohort was composed of 148 consecutive chest CTs (mean patient age, 57 years; age range, 15-90 years) performed with a z-axis modulation technique (number of examinations, n = 70) or fixed tube current (n = 78). There were 70 men (mean age, 57 years; age range, 23-84 years) and 78 women (mean age, 56 years; range, 15-90 years) in the study cohort.
CT Technique
All CT examinations were performed on 4-MDCT (LightSpeed QX/i, GE
Healthcare) or 16-MDCT (LightSpeed 4.X, GE Healthcare) scanners using manual
selection of fixed tube current (n = 78) or z-axis
modulation technique (n = 70). All examinations with the
z-axis modulation technique (AutomA, GE Healthcare) were performed
using identical scanning parameters (minimum = 75 mA, maximum = 380 mA; 15 H
noise index) as per our routine departmental protocol. The CT studies with the
fixed tube current technique were performed at 200-300 mA. Remaining scanning
parameters for both techniques were identical and included 140 kVp, 0.5-1 sec
gantry rotation time, 5-mm reconstructed slice thickness, standard
reconstruction algorithm, beam pitch of 1.5:1 (table speed/rotation = 15
mm/rotation) with 4-MDCT scanners, and 0.938:1 (table speed/rotation = 18.75
mm/rotation) with 16-MDCT scanners.
Z-Axis Modulation
There is a large variation in X-ray beam attenuation in different regions
of the body due to changes in size, shape, and tissue composition (density).
X-ray quantum noise in the incident beam projections, which depends on the
X-ray beam attenuation at that position, is an important determinant of image
noise at any slice position. Quantitative image noise is an important
characteristic of image quality, which should be maintained at an acceptable
level to acquire diagnostically acceptable image quality at greater radiation
dose efficiency.
Automatic tube current modulation techniques estimate image noise and adjust tube current to maintain a constant image noise with lower radiation exposure. The z-axis modulation technique automatically adjusts tube current in the scanning direction to the changing attenuation characteristics of the area being scanned to maintain a user-specified noise level. The z-axis modulation technique (AutomA technique) assessed in the present study allows the user to preselect the noise level (noise index) that will be acceptable in the reconstructed images for a given clinical indication. The noise index value is approximately equal to the quantitative image noise in the central region of the image obtained from scanning a uniform water phantom [1]. Selection of a higher noise index results in higher noise in reconstructed images but a lower radiation dose. A 5% increase in the noise index results in increased image noise with about a 10% decrease in associated radiation exposure. Conversely, CT at a lower noise index results in less image noise but a higher radiation dose. Thus, a 5% decrease in noise index typically increases radiation exposure by 10%.
To use this z-axis modulation technique, the technologist selects a noise index (a vendor-specific measure of desired image quality or noise) and a range of acceptable tube currents (minimum and maximum mA limits). The technique determines the tube current based on the patient's localizer radiograph projection data and a set of empirically determined noise prediction coefficients for a reference technique. The reference technique has a 2.5-mm slice thickness at the selected peak kilovoltage and 100 mAs using a standard reconstruction algorithm. The technique then adapts the tube current required to meet the preset noise levels (noise index) from information available in a single localizer radiograph, which includes density, size, and shape information about the patient. Based on this information, it adapts tube current according to the selected noise index and minimum and maximum mA limits. The minimum and maximum limits of 75 and 380 mA used in the present study allowed the technique to use any tube current values from 75 to 380 mA to achieve selected noise level (noise index). Manual selection of tube current for a continuous helical acquisition uses the same single tube current (mA) throughout the scan series. On the other hand, z-axis modulation can change the tube current from one slice position to the next (use lower current for midchest and higher for shoulders and abdomen) depending on the specified noise index and size, shape, and attenuation of the area being scanned. Although users can reduce radiation exposure for smaller patients by selecting a lower tube current value for performing CT with manual selection of fixed tube current, z-axis modulation automatically adapts tube current to patient size without the need for arbitrary selection of tube current values.
Number of Extra Images
For each examination, supraapical extra images were defined as images
obtained beyond the lung apices. All images obtained beyond the lowermost
portion of the lung parenchyma were categorized as infrapulmonary extra
images. The infrapulmonary extra images also included the infraadrenal extra
images, which were obtained beyond the lowermost portion of the adrenal
glands. For supraapical, infrapulmonary, and infraadrenal extra images, a
score of zero was given for those examinations that began at a level below the
lung apices or ended above the lowermost portions of lung parenchyma or
adrenal glands, respectively.
Clinical Indications for Chest CT
The study cohort was classified based on the clinical indications for chest
CT examination provided on the requisition into benign (n = 29),
indeterminate (n = 61), or malignant (n = 58) categories.
The indication was classified as malignant if the study was performed to stage
or follow a patient with a known malignant neoplasm. An indication was
classified as benign if the study was performed to evaluate a known disease
that was not a malignant neoplasm. For example, a patient with an indication
such as "history of scleroderma" or "evaluate for
interstitial lung disease" was classified as a benign indication.
Likewise, an indeterminate indication included patients that could not be
classified as definitely malignant or benign, such as "history of
cough" and "solitary pulmonary nodule." All CT examinations
were reviewed in blinded fashion by two radiologists (one with 5 years of
experience and one with 3 years of experience) and a fourth-year radiology
resident, in consensus, for pathology in the supraapical, infrapulmonary, and
infraadrenal extra images. In addition, electronic medical records were
reviewed to determine clinical importance of each additional finding. Findings
were considered clinically important if they represented new findings (i.e.,
not seen on a previous imaging examination), led to a dedicated imaging
examination, or affected clinical staging of disease or management.
Radiation Dose Associated with Extra Images
Tube current and gantry rotation time were recorded for each extra image in
CT studies performed with manual selection of tube current and z-axis
modulation technique. Tube current-time product (mAs) was calculated by
multiplying tube current by gantry rotation time. The CT dose index volume
(CTDIvol) and dose-length product (DLP) for CT examinations and for
the extra images were estimated using the manufacturer's technical reference
manual [5]. Although
CTDIvol and DLP are not the actual dose to a specific patient, they
form a standardized index of the average dose delivered from the scan series.
The technical reference manual published by the vendor was used to estimate
standard dose values for the 32-cm-diameter body phantom (CTDI100)
with adjustment factors that were used for CT examinations included in our
study [5]. The weighted CTDI
(CTDIw) was estimated by adding one-third of the central
CTDI100 to two-thirds of the peripheral CTDI100.
Subsequently, CTDIvol (measured in mGy) was estimated by dividing
the CTDIw by the pitch value. The CTDIvol represents the
average dose within a scan volume (to a standardized CT phantom) and is now
displayed on the user interface of all CT scanners. The DLP (measured in
mGy·cm) was calculated by multiplying the CTDIvol by the
length of the scan volume (in cm). The DLP represents the integrated dose for
the scan series.
Statistical Analysis
All statistical tests were performed with Microsoft Excel software. To
compare trends of extra image acquisition in patients with different clinical
indications, Student's t test was used to determine significant
statistical difference between the numbers of extra images acquired for each
indication category. A p value of less than 0.05 was considered a
statistically significant difference. Mean and median mAs were calculated for
examinations performed with z-axis modulation and fixed tube current
techniques. To compare radiation dose with fixed tube current and
z-axis modulation techniques, statistical differences between mean
mAs, CTDIvol, and DLP for examinations performed with these
techniques were determined using Student's t test.
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
There was a significant statistical difference between the numbers of extra images acquired for different clinical indication categories (p < 0.05). On average, three supraapical extra images (range, 1-18; median, 3.3) were obtained in chest CT examinations included in our study cohort. In patients with benign, indeterminate, and malignant indications, the average number of supraapical extra images was 3.0, 3.4, and 3.6, respectively. Likewise, on average, 11 infrapulmonary extra images (range, 2-27; median, 11.2) were acquired in the CT examinations of the chest. In patients with benign, indeterminate, and malignant indications, the average number of infrapulmonary extra images was 12.1, 11.5, and 11.8, respectively. Of these infrapulmonary extra images, an average of six infraadrenal extra images per CT examination (range, 1-20; median, 6.1) were obtained. In patients with benign, indeterminate, and malignant indications, the average number of infraadrenal extra images was 6.3, 6.1, and 6.4, respectively.
Additional Findings for Extra Images
No additional findings were identified in the supraapical extra images.
Additional findings in infrapulmonary extra images were noted in 45 CT
examinations, which included six adrenal nodules (4.1%, 6/148) and 39
nonadrenal findings (26.4%, 39/148) (Table
1). Nonadrenal findings included low-attenuation lesions in the
liver (n = 15), renal calculi (n = 6), cholelithiasis
(n = 5), fatty liver (n = 3), splenomegaly (n = 2),
focal low-attenuation lesions in the spleen (n = 2), abdominal
lymphadenopathy (n = 2), cirrhosis (n = 1), pancreatic mass
(n = 1), renal mass (n = 1), and indeterminate tubular
soft-tissue structure (n = 1). No additional findings were noted in
extra images acquired in patients younger than 40 years (n = 19).
|
No adrenal lesions were identified in extra images of CT examinations performed for benign indications. However, nonadrenal findings were noted in five examinations. These findings were considered not clinically important and represented incidental findings.
Likewise, in extra images acquired in indeterminate indications, one patient had an adrenal nodule, whereas 19 patients had nonadrenal additional findings. The adrenal nodule was unchanged from a chest CT performed 6 months earlier. Subcentimeter-sized, low-attenuation liver lesions identified in six patients with indeterminate indications for chest CT were not regarded as clinically important, as they were stable for 6 months to 2 years in previous abdominal CT examinations (n = 5) and most likely represented benign lesions such as hepatic cysts or hemangiomas. The low-attenuation tubular soft-tissue structure in the upper abdomen noted in one patient had been stable for several years and was most likely to represent a congenital anomaly of cisterna chyli remnant. CT examinations of three patients with abdominal lymphadenopathy (n = 2) and splenomegaly (n = 1) were misclassified as indeterminate indications because known history of chronic lymphocytic leukemia (n = 2) and metastatic adenocarcinoma (n = 1) was not provided to the radiologists. Among the 19 findings noted in patients with indeterminate indications, only two findings, cirrhosis of the liver in a 70-year-old man and renal mass in a 64-year-old woman (histopathology showed renal cell carcinoma) (Fig. 1) represented new findings.
|
Radiation Dose Associated with Additional Images
Mean tube current-time products for supraapical extra images in
examinations performed with z-axis modulation technique and fixed
tube current were 154 mAs (range, 37-228 mAs; 95% confidence interval, 143-165
mAs) and 135 mAs (range, 90-250 mAs; 95% confidence interval, 130-142 mAs),
respectively (p = 0.0036). A significant statistical difference
(p = 0.0271) was also noted between the mean mAs for infrapulmonary
extra images in CT examinations performed with z-axis modulation
technique (mean, 151 mAs; range, 37-262 mAs; 95% confidence interval, 138-164
mAs) and manual selection of fixed tube current (mean, 135 mAs; range, 90-250
mAs; 95% confidence interval, 129-141 mAs) (Figs.
2A, and
2B).
|
|
At our institution, the typical CTDIvol and DLP for chest CT studies performed with fixed tube current using scanning parameters identical to those used in the present study are 20.98 mGy and 604.57 mGy·cm, respectively. The typical CTDIvol and DLP values for chest CT performed with z-axis automatic tube current modulation at our institution are 10.51 mGy and 303.52 mGy·cm, respectively.
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Among the available methods of reducing radiation dose associated with chest CT are limiting imaging to the area of interest and the use of automatic tube current modulation technology. Greess et al. [9] and Tack et al. [10] have shown substantial reduction in the radiation dose associated with chest CT with the use of angular modulation technique. To our knowledge, no study has specifically evaluated the prevalence of scanning beyond the defined anatomic limits of the thorax in chest CT protocol, additional diagnostic information gained, and radiation dose to patients from the extra images in chest CT. In addition, although the effect of angular modulation on chest CT has been evaluated, the effect of z-axis modulation technique on the radiation dose associated with extra images acquired in chest CT has not been reported [7, 9-11].
Our study shows that a substantial number of extra images are acquired in routine chest CT. Most extra images in routine chest CT examinations are acquired below the lowermost portion of lung parenchyma, with an average of 11.2 infrapulmonary extra images and 3.3 supraapical extra images. Whereas extension of scan volume above lung apices can deliver additional radiation dose to the radiosensitive thyroid gland, caudal extension of scan volume below the lowermost portion of lung parenchyma is associated with additional radiation dose to upper abdominal structures. Interestingly, the number of extra images did not vary significantly with the use of manual selection of fixed tube current versus z-axis modulation technique or with the indication for the study (benign, indeterminate, or malignant).
None of the findings in the extra images acquired in patients with benign indications changed the clinical management. Likewise, there were very few clinically important additional findings in patients with indeterminate indications for chest CT. After excluding three findings in studies misclassified into the indeterminate indications subgroup because of unspecified history of malignancy in the requisition, only two examinations (3.4%, 2/58) performed in patients with indeterminate indications were considered important. Both of the examinations classified as indeterminate were performed in patients with lung cancer. On the other hand, 19 patients with malignancy had additional findings (32.8%, 19/58) and of these, nine additional findings (five adrenal nodules, four liver lesions) were clinically important (15.5%, 9/58). As most clinically important findings in the infrapulmonary extra images were identified in patients with a history of malignancy, we recommend that chest CT in these patients should cover the liver and adrenal glands. However, our findings underscore the importance of knowing the history of malignancy in all patients when deciding on the protocol for chest CT.
Indeed, the American College of Radiology (ACR) guidelines do not recommend imaging of the adrenal glands in routine chest CT examinations [8]. However, there is general agreement that CT examination of the chest performed for staging of non-small cell lung cancer should include the adrenal glands and liver as recommended by the ACR Task Force on Appropriateness Criteria by both the Expert Panel on Thoracic Imaging and the Lung Cancer Work Group [13-15]. Our results are in agreement with these recommendations.
Our study also shows that substantial additional radiation dose is delivered to the patients with acquisition of extra images in chest CT. Overall, extra images contributed a 20.9% (126.22/604.57 mGy·cm) increase in DLP in examinations performed with the fixed tube current technique and 56.9% (172.57/303.52 mGy·cm) increase in DLP with the z-axis modulation technique. Most additional radiation dose was associated with the infrapulmonary extra images rather than supraapical extra images, as more images were acquired below the lowermost portion of lung parenchyma compared with the region above lung apices. Therefore, substantial reduction in patient radiation dose can be achieved by limiting the number of images acquired below the lowermost portion of lung parenchyma in routine chest CT.
Compared with the fixed tube current technique, the z-axis modulation technique increased the radiation dose associated with the acquisition of extra images, with an overall increase of 36.7% ([172.57 - 126.22] / 126.22) in DLP. Likewise, z-axis modulation increased radiation dose by 31.9% ([39.98 - 30.31] / 30.31) for supraapical extra images and by 38.2% ([132.59 - 95.91] / 95.91) for infrapulmonary extra images compared with the fixed tube current technique. As described in the preceding section, the z-axis modulation technique adapts tube current to the regional cross-sectional size, shape, and attenuation. Thus, the increase in radiation dose associated with the z-axis modulation technique for acquisition of extra images with chest CT may be explained on the basis of greater X-ray beam attenuation and image noise at the level of the shoulders and abdomen compared with that within the anatomic limits of the thorax. Despite this increase in radiation dose with acquisition of extra images with the z-axis modulation technique, the overall radiation dose for the entire chest CT performed with z-axis modulation was substantially lower than the overall radiation dose for the fixed tube current technique. Therefore, to achieve maximum radiation dose reduction with automatic tube current modulation, imaging should be confined to the area of interest.
Our study has a few limitations. The major limitation is that CT examinations included in the study cohort were performed on different CT scanners (4-MDCT [n = 57] and 16-MDCT [n = 91] scanners). In addition, we did not assess the effect of patients' weight or regional dimensions on radiation dose associated with z-axis modulation and fixed tube current technique. An important limitation is that we did not assess factors that lead to acquisition of extra images during CT. Several factors could have led to acquisition of extra images, such as erroneous selection of scan volume, a tendency to overshoot the margins to avoid noninclusion of specified scan volume, inappropriate scanning protocols for young patients with benign or indeterminate indications, and lack of interest or awareness on the part of radiologists to avoid inadvertent exposure. Although erroneous selection of scan volume could have led to acquisition of extra images, especially those above the lung apices, infrapulmonary extra images were acquired due to the extended length of our chest CT protocol. Technologists may have acquired extra images due to lack of appropriate training, lack of awareness regarding CT radiation dose risks, lack of stringent guidelines and monitoring of CT studies, and their intention to extend a little beyond defined scanning area to ensure inclusion of defined anatomic limits. The latter possibility may also have resulted from the scanning of uncooperative, noncompliant, or breathless patients. Further studies may be needed to investigate steps that can aid in minimizing acquisition of extra images, which might include determining better anatomic landmarks for defining region of interest, careful training of technologists, and monitoring of CT studies. Future studies must also assess if extra images acquired in our institution with its chest CT protocol are also acquired in other CT centers.
Our study highlights some important considerations in routine chest CT protocols. First, routine chest CT protocols should take into account the indication for the study, as additional images do not provide additional diagnostic information in most patients with nonmalignant indications. Thus, in the absence of a specific request or an indication for simultaneous assessment of the upper abdomen, a routine chest CT should be limited to the anatomic limits of the thorax in most patients with benign and indeterminate indications. Extension of scan length beyond the lowermost portion of lung parenchyma does add important diagnostic information in patients with malignant disease. Second, technologists should realize that unless requested otherwise, scan length should not be exceeded, to ensure that additional radiation dose is not given to the patients undergoing routine chest CT. This may involve retraining of the technologists and explaining and training the patients before scanning, when appropriate, about breath-holding. Scanning with faster state-of-the-art MDCT scanners should not facilitate carelessness in maintaining scan length to the anatomic area of interest. Lastly, z-axis modulation technique may be used for routine chest CT despite increase in radiation dose in the regions of extra images, as overall radiation dose with z-axis modulation is substantially less than with the fixed tube current technique.
In conclusion, a substantial number of extra images are acquired during chest CT. Extra images acquired beyond anatomic limits of the thorax do not contribute additional diagnostic information in patients with benign and indeterminate indications and are associated with increased radiation dose. Although the z-axis modulation technique reduces radiation for images of the thorax compared with the manual selection technique, the use of z-axis modulation results in increased radiation for the extra images acquired outside of the thorax.
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
CiteULike 
Complore 
Connotea 
Del.icio.us 
Digg 
Reddit 
Technorati What's this?
This article has been cited by other articles:
|
|
 |
|
  M.T. Russell, J.R. Fink, F. Rebeles, K. Kanal, M. Ramos, and Y. Anzai Balancing Radiation Dose and Image Quality: Clinical Applications of Neck Volume CT AJNR Am. J. Neuroradiol., April 1, 2008; 29(4): 727 - 731. [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]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
