| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Original Research |
1 Department of Pediatric Radiology, Indiana University School of Medicine,
Riley Hospital for Children, 702 Barnhill Dr., Rm. 1053, Indianapolis, IN
46202.
2 Department of Radiology, Duke University School of Medicine, Durham, NC
27710.
Received June 13, 2008;
accepted after revision July 21, 2008.
Address correspondence to B. Karmazyn.
Abstract
 Top Abstract IntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
MATERIALS AND METHODS. Forty-five patients 20 years old or younger divided into two groups weighing 50 kg or less and more than 50 kg underwent unenhanced 16-MDCT in the evaluation of acute flank pain. An investigational computer-simulated tube current reduction tool was used to produce additional 80- and 40-mA examination sets (total number of image sets = 135). Three independent blinded readers ranked random images for stones (confidence scale, 1-5, least to most), hydronephrosis, noise-based image quality, and presence of nonrenal lesions.
RESULTS. Compared with the standard tube current used for the original CT scans, there was no significant reduction (p = 0.37) in detection of renal stones at the 80-mA setting (mean dose reduction, 67%; range, 43-81%); and at the 40-mA setting (mean dose reduction, 82%; range, 72-90%), the detection rate was significantly lower (p = 0.05). At the 40-mA setting, there was no significant difference among the children weighing 50 kg or less (p = 0.4). Detection of ureteral stones and hydronephrosis was not significantly different at 80 and 40 mA; however, disease frequency was low, and no definite conclusion can be made.
CONCLUSION. Simulated dose reduction is a useful tool for determining diagnostic thresholds for MDCT detection of renal stones in children. Use of the 80-mA setting for all children and 40 mA for children weighing 50 kg or less does not significantly affect the diagnosis of pediatric renal stones.
Keywords: CT • pediatric imaging • radiation • renal stones
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
A reduction in tube current is one method of lowering radiation exposure during CT [5]. The disadvantage of a decrease in tube current is an increase in noise, which can reduce the radiologist's ability to detect disease [6]. Investigational repeated scanning of children with lower doses is not justifiable. One option is to use a computer to simulate a tube current lower than that used for the original CT examinations [7]. Although this simulation tool has been investigated for systematic reduction in tube current in the adult population with nephroureterolithiasis [8], similar investigations have not been performed for evaluation of nephroureterolithiasis in the more radiation-sensitive pediatric population. The purpose of our study was to determine how tube current reduction affects the diagnostic yield in CT examinations for suspected renal or ureteral stones in children.
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
CT Technique
All studies were performed with a 16-MDCT scanner (LightSpeed, GE
Healthcare). Imaging parameters were detector configuration, 16 x 0.625
mm; section thickness, 5 mm; interval, 2.5 mm; tube voltage, 120 kVp; gantry
rotation time, 0.5 seconds for body weight less than 60 kg, 0.8 seconds for
body weight 60 kg or greater; pitch, 1.375; matrix size, 512 x 512;
standard reconstruction algorithm. The tube current depended on our
institutional weight-adjusted MDCT protocols. For 37 children (82%) the tube
current ranged from 140 to 260 mA (70-208 effective mAs) according to the CT
renal stone protocol. Eight children (18%) underwent routine abdominal studies
with up to 425 mA (340 effective mAs) according to the radiologist's
discretion. No automatic tube current modulation was used.
Image Simulation
Simulated tube current reduction is a proprietary tool that operates with
projectional scan data from diagnostic CT examinations. User-controlled random
gaussian noise is added to the original scan data to simulate a reduction in
tube current. The amount of simulated noise added to the scan data simulates
the increased noise that would occur in scans actually obtained at these tube
currents. This technique has been described in detail elsewhere
[7].
All scan and patient identification was removed from the original CT images. Postprocessing of the original CT raw data was performed. Each CT study was reproduced to simulate tube currents of 80 and 40 mA, yielding three examinations (original and two simulated examinations) for each patient for a total of 135 image sets. Images from the original and simulated examinations were viewed on standard clinical PACS stations (Centricity, GE Healthcare) under clinical reading conditions.
Image Analysis
The images at the three tube current settings were viewed randomly. Each
study was interpreted independently by three pediatric radiologists with 14-30
years of experience, all fellowship trained with a certificate of added
qualification. The reviewers were blinded to all technical parameters.
Observers could use the standard PACS tools to make the adjustments they would
make during routine interpretation. There was no time limit for viewing a
study. When a calculus was believed to be present, the window and level
settings were standardized at width 650 HU and level 150 HU for measuring the
maximum diameter of renal stones with PACS calipers. The reference standard
was the observers' interpretations of the images from the original
examination. Criteria for the presence of stones and hydronephrosis were at
the discretion of the radiologist.
The reviewers rated each image on a modified 5-point Likert scale for the presence of stones in the right kidney, left kidney, right ureter, and left ureter (1, definitely no stone; 2, probably no stone; 3, uncertain finding; 4, probably stone; 5, definitely stone). The data were grouped as depicting stones if there was any rating of 4 or 5. If more than one stone was identified, the readers counted the number of stones present and measured the largest stone. Findings associated with renal obstruction, including hydronephrosis, hydroureter, and soft-tissue stranding in perirenal fat, were rated as present, uncertain, or absent.
The presence or absence of nonurologic abnormalities, such as bowel wall thickening and distention, enlarged lymph nodes 8 mm or more in short axis [9], adnexal mass, and adnexal cyst larger than 5 mm, and visualization of a normal appendix were rated with the same modified 5-point Likert scale. Each reviewer then evaluated the overall quality of the image as adequate, suboptimal, or nondiagnostic.
Statistical Analysis
For each qualitative variable, the mean responses from the three readers
were compared at each tube current level by use of the paired Wilcoxon's
signed rank test. The resulting p values were not adjusted for
multiple comparisons. Reader agreement was measured with Cohen's kappa
statistic. Standard errors of estimated kappa statistics and averages and
differences of kappa statistics were computed by use of the statistical
jackknife. A kappa value less than 0.2 indicated poor agreement; 0.21-0.40,
fair agreement; 0.41-0.60, moderate agreement; 0.61-0.80, good agreement; and
0.81-1.00, excellent agreement. For all tests, statistical significance was
set at p < 0.05. Statistical software was used (SAS version 9.1,
SAS Institute).
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Renal and Ureteral Stones
The median diameter of the renal stones was 3 mm (range, 1-12 mm). In
kidneys with multiple stones, the largest stone was measured. For each tube
current level, the percentages of kidneys rated as having stones are listed in
Table 1. Figures
1A,
1B, and
1C shows a renal stone at the
three tube current settings. In the original image sets, the mean of the three
readers was detection of stones in 32% of kidneys, compared with 29% at 80 mA
(p = 0.37) and 23% at 40 mA (p = 0.05).
Table 2 summarizes the rate of
detection of renal stones in patients weighing 50 kg or less (n = 28)
and patients weighing more than 50 kg. Among patients weighing 50 kg or less,
the mean rating was 33% on the original studies, compared with 32.1% at 80 mA
(p = 1.0) and 29.8% at 40 mA (p = 0.4). Among patients with
a body weight more than 50 kg, the mean detection rate was 29.4% on the
original studies, compared with 17.6% at 80 mA (p = 0.4) and 11.8% at
40 mA (p = 0.06).
|
|
|
|
|
With the original study as the reference standard, the sensitivity and
specificity in the detection of renal stones were 85% and 97% for the 80-mA
image sets and 69% and 97% for the 40-mA sets. Interobserver agreement for the
presence of renal stones was good to excellent for all tube current groups
(
= 0.68-0.87), and no significant differences were found between
readers.
Only eight ureteral stones were detected in our study group. There was no significant difference between tube current groups regarding detection of ureteral stones. The interobserver agreement was low to moderate, ranging from 0.20 to 0.64, but no significant differences were found between readers.
|
|
|
|
|
|
Nonurologic Abnormalities
Only a few patients had non-urinary system abnormalities. On the original
studies, the mean rating for the presence of bowel wall thickening or
distention was 1.9; enlarged lymph nodes, 2.0; and adnexal mass or cyst, 1.3.
There was no statistical difference in the number of such abnormalities
detected with 40 mA and 80 mA compared with the original tube current level
(p = 0.39 and p = 0.59, respectively).
The mean visualization of the normal appendix ranged from 23% to 42% (Table 3). The mean detection rate on images obtained at the original CT settings was higher than that on images obtained at 80 and 40 mA (p < 0.01 for each). The decreased detection of the appendix also was significant in the subgroup of patients weighing 50 kg or less in comparisons of either 40- or 80-mA simulated studies with the original studies (p = 0.001 and p = 0.004, respectively). Detection of the appendix was decreased in patients weighing more than 50 kg in comparisons of 40- or 80-mA simulated studies with the original studies. The decreased detection, however, was statistically significant (p = 0.03) only for the 40-mA image sets.
|
Image Quality
The original CT examinations were considered of adequate quality more
frequently than were either the 80- or 40-mA image sets (p < 0.01
for each). There was marked interobserver variability in rating the quality of
the studies (Table 4). The mean
rating for optimal studies increased from 34% to 56% to 67%, respectively,
with the increase in tube current from 40 to 80 mA to the original setting. In
the cases of children weighing 50 kg or less, image quality was significantly
lower between the original and the 40-mA image sets (p < 0.01) but
did not vary between the original and the 80-mA image sets (p = 0.3).
Among patients with higher body weight, image quality was significantly lower
between the original and both the 40- and 80-mA image sets (p <
0.01 for both).
|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
The as-low-as-reasonably-achievable standard asserts that the radiation dose necessary to generate diagnostic CT scans be kept to a minimum. There is a linear relation between tube current and radiation dose. In addition, for a given tube current setting, pediatric doses are much larger than adult doses because a child's thinner torso provides less shielding of organs from radiation exposure. A reduction in the tube current setting proportionately reduces the dose and the risk. However, with decreased tube current, image noise increases, possibly reducing examination quality and diagnostic capability. Therefore, reductions in tube current should be balanced with the increased risk of an incorrect diagnosis [6, 7, 10, 11].
Studies have been conducted with adult subjects to compare routine- and low-dose-radiation CT [12, 13]. This paradigm would be difficult to apply to children because of radiation concerns. The use of a computer-simulated, low tube current (dose) reduction technique eliminates these difficulties. With this technique, there is no need to expose a child to additional radiation. It is possible to simulate multiple tube current reductions. This technique can be used for any CT study, and a prototype of this technology has been tested on children [7]. A study of this simulation tool technique showed that a tube current of 35 mA is acceptable for evaluation of nephrolithiasis in the size spectrum of adults [8]. We used the computer-simulated low-dose technique in our study and thus were able to compare the diagnostic performance of unenhanced MDCT studies at different tube current settings for evaluation of renal and ureteral calculi.
Studies of unenhanced MDCT for evaluation of nephroureterolithiasis in adult populations have shown that unenhanced MDCT at low tube current (reductions of 25-66%) can be used without loss of diagnostic quality [8, 12-16]. It is difficult to extend these results to children because the pediatric population is more heterogeneous in body size and fat distribution. With the increased incidence of obesity among children (5% of our patients weighed > 100 kg), adjusting CT parameters to body size becomes even more challenging.
Our results show that the difference in detection of renal stones was not significant at 80 mA compared with the original settings. Using the 80-mA setting would reduce CT dose an average of 67% (43-81%). In patients weighing 50 kg or less, using the 40-mA setting did not significantly decrease renal stone detection. Using the 40-mA setting in this group of patients would reduce CT dose an average of 82% (range, 71-90%). In patients weighing more than 50 kg, use of the 40-mA setting was accompanied by a marginally significant (p = 0.06) decrease in renal stone detection. This is likely related to the small sample size.
A potential concern with decreased tube current settings is false diagnosis
of renal stones that may alter management. However, with the original studies
as the reference standard, the specificity of both 40- and 80-mA imaging was
very high (97%). Moreover, interobserver agreement was good to excellent for
each tube current setting ( = 0.68-0.87). This finding is supported by
those of previous studies [12,
13] that showed a high
sensitivity of 30 effective mAs MDCT in the detection of
nephroureterolithiasis in adults. Poletti and colleagues
[13], however, found that MDCT
at 30-effective-mAs was limited in the detection of renal stones with a
diameter less than 3 mm. The number of ureteral stones in our study was too
small for statistical evaluation, and there was also poor interobserver
variability ( = 0.20-0.64). We cannot conclude from our results whether
low tube current settings influence detection of ureteral stones.
The numbers of obstructed kidneys were small in our study group. The most common finding that suggested obstruction was hydronephrosis. Whereas soft-tissue stranding in perirenal fat has been found to be sensitive for renal obstruction in adults [17], it was detected in only one patient in our study. It is possible that the small amount of perirenal fat in children, including obese children, makes detection of tissue stranding in the fat more challenging.
Although we detected few substantial non-urinary tract abnormalities, another indication for CT in these patients is to rule out appendicitis. A decrease in the ability of radiologists to confirm the presence of a normal appendix with low-dose CT would argue against its introduction. We found that the ability to confirm the presence of a normal appendix was significantly higher with the original tube current settings. Thus a reduced tube current level may be satisfactory for urinary tract stone detection but not for ruling out appendicitis, and the relative importance of these tasks must be considered before a low-dose protocol is introduced. However, the utility of low-dose imaging in the evaluation of appendicitis itself was not evaluated because there was no case of acute appendicitis in our study group.
When readers evaluated the quality of CT images, there were significantly fewer optimal studies at 80 and 40 mA than at the original tube current settings. Only 34% of the studies were rated technically adequate at 40 mA, compared with 67% at the original settings. This discrepancy was greater in patients with a large or very large body habitus. However, in the subgroup of children weighing 50 kg or less, image quality did not vary significantly between the original and the 80-mA image sets (p = 0.3). This finding is supported by those of other authors [15], who suggested not performing low-dose CT examinations on adults with large body habitus.
Our study had limitations. We did not have enough patients with ureteral stones and renal obstruction to evaluate differences in detection rates between the different tube current settings. In addition, it is conceivable that the few patients with hydronephrosis and ureteral stones were remembered by the readers, thus skewing the data. The studies were evaluated only in the axial plane and not on coronal or sagittal multiplanar reconstructions, use of which may increase detection of nephroureterolithiasis. We evaluated only one parameter that can reduce radiation exposure (tube current), yet there are other parameters, such as kilovoltage, that can influence radiation exposure and rate of detection of nephroureterolithiasis. Dose does not necessarily translate equally to tube current reduction. Although dose reduction was implied in this study of simulated tube current reduction, the actual dose differences were not measured or estimated. We did not specify how to evaluate the quality of the images, which may explain the high interobserver variability.
Our study showed the value of computer-simulated technique as a tool for evaluation of low tube current (low radiation dose) CT images without use of additional CT studies on pediatric patients. We found it possible to decrease the tube current setting to 80 mA for all children and to 40 mA for children weighing 50 kg or less without significant change in detection of renal stones in all children. Changing the tube current setting to 80 mA would decrease radiation exposure 67%, and changing it to 40 mA would decrease the exposure 82%. However, at this reduced dose, there was significantly decreased detection of normal appendixes, and fewer studies were graded optimal at subjective evaluation. In addition, we did not have enough cases to evaluate the change in rate of detection of ureteral stones and renal obstruction. Therefore, a prospective clinical trial is needed for evaluation of tube current reduction in acute renal colic. Our results are a scientific rationale for tube current reduction.
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
CiteULike 
Complore 
Connotea 
Del.icio.us 
Digg 
Reddit 
Technorati What's this?
This article has been cited by other articles:
|
|
 |
|
  A. C. Persaud, M. D. Stevenson, D. R. McMahon, and N. C. Christopher Pediatric Urolithiasis: Clinical Predictors in the Emergency Department Pediatrics, September 1, 2009; 124(3): 888 - 894. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
