| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Original Research |
1 Department of Radiology, Medical University of Vienna, General Hospital of
Vienna, Waehringer Guertel 18-20, A-1090 Vienna, Austria.
2 Department of Radiology, University of Amsterdam, Amsterdam, The
Netherlands.
3 Department of Radiology, University Medical Center Utrecht, Utrecht, The
Netherlands.
4 Department of Urology, Medical University of Vienna, Vienna, Austria.
Received December 17, 2006;
accepted after revision March 28, 2007.
Address correspondence to M. Memarsadeghi
(mazda.memarsadeghi{at}meduniwien.ac.at).
Abstract
 Top Abstract IntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
MATERIALS AND METHODS. We included 147 consecutive patients (72 women and 75 men; mean age ± SD, 58 ± 18.1 years) with suspected urinary stone disease who underwent unenhanced MDCT. Scans were obtained with a 4 x 1 mm collimation and were reconstructed with a section thickness of 1.25 and 3 mm. We compared the diagnostic yield of 3-mm axial sections with that of coronal reformations reconstructed from 1.25- and 3-mm axial sections. Imaging data were evaluated in random order by two radiologists. The significance of the difference between the axial sections and coronal multiplanar reformations (MPRs) was tested for the number, size, and location of uroliths and for the presence of alternative diagnoses. The time required for review by both observers was recorded.
RESULTS. We found uroliths in 72 patients. There was no difference between 3-mm axial sections and coronal reformations from 1.25-mm sections with regard to the number of detected stones (n = 264 for both protocols), whereas coronal reformations from 3-mm sections revealed significantly fewer calcifications (n = 255, p = 0.016). Coronal reformations did not improve the localization of calcifications. Review time, however, was significantly shorter for coronal reformations than for axial sections (p = 0.001); however, coronal reformations were less sensitive than axial sections for the detection of additional findings suggestive of alternative diagnoses in 16 (30%) of 53 patients.
CONCLUSION. Coronal reformations from MDCT do not improve urinary stone detection but may reduce evaluation time; however, there is the danger of missing additional findings. Coronal reformations reconstructed from thick (i.e., 3-5 mm) axial sections may result in reduced detection of small stones and should therefore be avoided.
Keywords: emergency radiology • kidney disease • MDCT technique • multiplanar reformations • renal stones • urinary stone disease • urinary system
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
The purposes of this retrospective study were to assess whether coronal reformations improve the diagnostic performance of MDCT in patients with acute flank pain and suspected urinary stone disease; and to determine if performing reformations from 3-mm-thick axial sections used for standard viewing is sufficient or if it is necessary to perform such reformations from thinner sections.
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Image Acquisition
All CT examinations were performed on an MDCT scanner (Somatom Volume Zoom,
Siemens Medical Solutions). Unenhanced scans were obtained with a 4 x 1
mm detector configuration, a 0.5-second rotation time, a tube voltage of 120
kVp, and an effective tube current-time product of 100 mAs. The scan length
was adapted to the length of each patient's abdomen and pelvis with a table
feed per gantry rotation of 6 mm in all patients. For image reconstruction, a
moderately smoothing convolution kernel (B30) and a 512 x 512 matrix
were used.
We reconstructed two sets of transverse (axial) images with a section width of 1.25 and 3 mm, respectively. The reconstruction increment was chosen to be approximately half the section width (0.75 and 1.5 mm, respectively). Coronal reformations were reconstructed using a standard CT workstation (Volume Wizard, Siemens) with the following parameters: section width of 1.5 mm and reconstruction increment of 1.5 mm and section width of 3 mm and reconstruction increment of 3 mm from the 1.25- and 3-mm axial sections, respectively.
We then compared diagnostic yield and evaluation time for transverse 3-mm sections with those for coronal reformations from 1.25- and 3-mm axial sections.
Image Evaluation
Data from three sets of images were independently reviewed by one
fifth-year resident and one board-certified radiologist who are both
experienced in interpreting abdominal CT examinations. Any examination from
any of the image sets that was rated differently by the two observers was
reevaluated by both observers together to reach a consensus opinion. The
observers were asked to separately determine whether pyelocalyceal and
ureteral calcifications were present. They were instructed to document the
location and size of each calcification using standard measurement
devices.
Uroliths were defined as focal calcified structures that were located within the ureter or within the intrarenal pyelocalyceal collecting system. The observers were asked to indicate the location of the calcification as one of the following: the pyelocalyceal collecting system; the proximal, middle, or lower third of the ureter; or the ureterovesical junction zone.
Observers were also asked to record the presence of renal cysts, phleboliths, or any findings suggestive of an alternative diagnosis on the various axial and coronal image data sets.
Three review sessions were performed. During the first session, observers evaluated transverse CT scans of 3-mm section thickness in different random order for patients. During the second and third sessions, observers evaluated the coronal MPRs. The MPRs were randomly intermingled with regard to patient order and section width. The time interval between review sessions was at least 2 weeks. The time required for review by both observers was calculated for two review sessions from overlapping transverse scans of 3-mm section width and nonoverlapping planar coronal reformations of 1.5 mm. Observers evaluated the images on a workstation (Magic View 1000, Siemens) using an interactive mouse-driven cine mode. The size of the largest axis of each stone was measured and recorded on transverse scans of 3-mm section width.
The stone size was measured using metric software on the Magic View 1000 workstation. The ambient light was subdued and there were no time constraints.
To approximate the clinical situation, the observers were informed about which side on the patient prompted the clinical symptoms.
Standard of Reference
The standard of truth was determined by consensus of the same two observers
who had been involved in the evaluations. For the determination of the final
diagnosis, information from all scans was available, as well as information
about clinical symptoms (e.g., hematuria, pain); clinical history, as shown on
treatment charts; and outcome. The standard of truth was determined after
finishing the radiologic evaluation.
Statistical Analysis
Data were analyzed with the SAS statistical package (SAS I.I., SAS, PROC
FREQ SAS/STAT [version 8], SAS Institute). Results are expressed as absolute
numbers and means ± SDs. The significance of the difference between the
detection rates with axial sections and coronal reformations from 1.25- and
3-mm sections was calculated for calcifications and associated abnormalities
in the various coronal sets of images using the McNemar test at a p
level of < 0.05 for each set separately. The significance of the difference
was calculated separately for transverse reconstructions and coronal
reformations. An analysis of variance for repeated measures was used to
determine the differences in review time between the two sets of images and
both observers.
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Seventeen patients had only one stone, and the remaining 55 patients had more than one stone (Table 1). The stone size ranged from 1 to 12 mm, with a mean size ± SD of 3.1 ± 2.1 mm and a median size of 2 mm. Microscopic hematuria or positive results for dipstick urine analysis were reported in all 72 patients. Surgery was performed to remove the stone in 15 patients with stones ranging in size from 6 to 12 mm (mean ± SD, 7.7 ± 1.8 mm). Spontaneous stone excretion followed by immediate relief of pain was reported in 57 patients.
|
A total number of 264 uroliths (139 stones on the left side and 125 stones on the right) were seen in 72 patients on axial sections with a 3-mm section width. Of these 264 stones, 174 (66%) were located in the pyelocalyceal system and 90 (34%) in the ureter. Of the ureteral stones, 35 were located in the proximal third of the ureter; 13, in the middle third; 15, in the distal third; and 27, in the ureterovesical junction zone.
Only coronal reformations from thin (i.e., 1.25 mm) axial sections revealed the same 264 stones as the 3-mm axial sections. No additional calcifications could be seen on these thin coronal reformations. Coronal reformations from thicker (3 mm) sections showed 255 uroliths in 65 patients. The difference was statistically significant (p = 0.016). The transverse diameter of the missed stones ranged from 1 to 2 mm (mean ± SD, 1.7 ± 0.5 mm; median, 2 mm). The missed stones were nonobstructive. Nine stones were missed in seven patients: two stones each in two patients and one stone in each of five patients. The missed stones were located in the pyelocalyceal system (n = 8) in six patients and in the distal ureter (n = 1) in one patient (Fig. 1A, 1B, 1C). However, the missed stones would not have led to a false-negative interpretation of the examination in any of these patients because of the simultaneous presence of larger stones.
|
|
|
Alternative Diagnoses
Alternative diagnoses other than urolithiasis were found in 53 patients
(36%) on 3-mm axial sections. Alternative diagnoses included diverticulitis
(n = 31), small-bowel obstruction (n = 8), pancreatitis
(n = 6), infrarenal aortic aneurysm (n = 5), and acute
pyelonephritis (n = 3). Diverticulitis was missed on coronal
reformations in 12 of 31 patients (39%), pancreatitis in three of six (50%),
and infrarenal aortic aneurysm in one of five (20%).
Review Time
For both observers, the mean review time for transverse reconstructions
with a 3-mm section width was statistically significantly longer than that for
coronal reformations reconstructed from 1.25-mm axial sections (mean ±
SD, 35.1 ± 15.1 vs 15.0 ± 7.9 seconds, respectively, for the
radiologist; and 66.5 ± 25.5 vs 33.0 ± 13.0 seconds,
respectively, for the resident; p < 0.0001).
The mean time spent by the more experienced observer for reviewing the coronal reformations was significantly shorter than that of the fifth-year resident (p < 0.0001).
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
We used the thinnest collimation on our 4-MDCT scanner, 4 x 1 mm, and reconstructed 1.25-mm-thick sections. This reconstructed section thickness is somewhat thicker than those on 16- or 64-MDCT units if the thinnest possible collimations were used; for example, 16 x 0.75 mm yields 1-mm-thick sections, and 64 x 0.625 mm collimation yields 0.9-mm-thick sections. Such thin collimations, however, are not always advantageous in clinical practice because they induce substantially higher levels of noise, which is especially disturbing in a relatively low-dose setting, such as for diagnosis of urolithiasis. Therefore, frequently the next wider collimation is used (e.g., 32 x 1 mm instead of 64 x 1 mm on Toshiba 64-MDCT units or 32 x 1.25 mm on Philips Medical Systems or GE Healthcare units), and it yields a section thickness of between 1 and 1.5 mm, which is in the same range as our setting. In summary, the reconstructed section thickness of 16- and 64-MDCT units is between 28% lower (0.9 mm) and 20% wider (1.5 mm) than the section width we used for this study. We believe that the differences are small enough to ensure that our results remain valid for 16- and 64-MDCT units as well.
To our knowledge, the role of coronal reformations performed with MDCT in the evaluation of patients with acute flank pain has not been previously reported. Our results showed that there was no difference in the number of uroliths detected on 3-mm axial sections and coronal reformations reconstructed from 1.25-mm axial sections. Using thicker (i.e., 3 mm) axial sections to reconstruct coronal reformations, however, led to a significantly reduced sensitivity for the detection of uroliths. Only small stones of between 1 and 2 mm in diameter were missed, most of which were located in the pyelocalyceal system (8/9). The missed stones were nonobstructive.
The ability to acquire isotropic data sets with MDCT has stimulated interest in viewing the abdomen in planes other than the transverse plane. However, the usefulness of MPRs in the evaluation of the abdomen is still controversial. For example, Wong et al. [34] used 4-MDCT with MPRs to evaluate the liver and focused on the relationship of masses to portal veins, hepatic veins, bile ducts, and the vena cava. Caoili and Paulson [35] found MPRs helpful in the evaluation of small-bowel obstruction, particularly for the identification of the point of transition from dilated to decompressed bowel. Similarly, coronal reformations from scans obtained along the length of the pancreatic duct and common bile duct have been shown to be a useful adjunct to routine transverse scans in the local staging of pancreatic carcinoma and in the diagnosis of suspected biliary tract disease [32]. As opposed to these findings, Schmidt et al. [31] reported significantly lower sensitivity and specificity with coronal views in the diagnosis of small-bowel disorders with MDCT enteroclysis.
In our study, we compared the diagnostic performance using 3-mm transverse sections with that using coronal reformations from 1.25- and 3-mm axial sections. All images were evaluated in a cine mode on a soft-copy display. We did not find additional diagnostic information on the coronal reformations compared with the transverse sections. In this regard, we agree with Katz and coauthors [29] that MPRs do not provide more information than transverse sections. We did not specifically quantify observer confidence; other authors [36], however, have reported that coronal reformations may help to increase the confidence of observers regarding urinary stone detection.
In our study, we found that observer performance with MPRs from 3-mm axial sections was significantly worse for the detection of stones: When such coronal reformations were evaluated, calcifications were missed in seven patients. There is only one study to date that specifically addresses the diagnostic value of MPRs for the detection of ureteral calculi [26]. The authors of that study concluded that reformatting single-detector CT data produced images similar in orientation to excretory urograms. However, single-detector techniques yield reformatted images of sub-optimum quality.
As suggested by Schmidt et al. [31], we found diagnostic value to be significantly lower for coronal reformations with regard to diagnoses other than urinary stone disease, although the review time was significantly shorter for coronal reformations than for transverse sections for both observers. Longer review time for transverse sections may potentially be explained by detecting more ancillary findings on transverse sections. In our study, diverticulitis was missed in 12 of 31 patients (39%), acute pancreatitis in three of six (50%), and infrarenal aortic aneurysm in one of five (20%) by both observers. These results may be explained by the fact that many radiologists still are more accustomed to primarily evaluating axial sections for abdominal diseases and that coronal sections are frequently used as an adjunct to further characterize disease processes that have already been detected on axial sections. As a consequence, most pathologic signs, such as bowel wall thickening and intraperitoneal fluid or edema, can be more easily detected on axial than on coronal views [31]. However, it is likely that missing findings suggestive of alternative diagnoses on coronal images is a learning curve issue.
In our institution, we now review CT studies for stones on axial sections only. Coronal reformations are still reconstructed for another reason: Urologists have become accustomed to the better anatomic orientation on coronal images and now use those images to gain a better understanding of stone location.
A limitation of our study is the fact that there was no absolute standard of truth. We defined the diagnosis of urolithiasis as unequivocal evidence of calcification as determined by the consensus of two observers using all available data including axial and MPR data sets. However, as pointed out by previous authors, CT has become the reference method and there is no other imaging method that is more sensitive for the detection of calcifications. Another potential limitation is that we did not compare a combined interpretation of a transverse reconstruction and coronal reformations versus the interpretation of only the transverse reconstruction. Finally, we did not measure stones on coronal reformations as Nadler et al. [37] suggested because we could not provide a sufficient gold standard for such measurements.
In conclusion, we found that coronal reformations did not improve the detection of uroliths compared with overlapping axial sections of 3 mm in width and that a significant number of small calculi may be missed if coronal reformations are directly reconstructed from such 3-mm axial sections. Coronal reformations should, therefore, be reconstructed from thin axial sections (1.25 mm width). Although evaluation of coronal reformations speeds up the evaluation of suspected urolithiasis, using coronal reformations alone and not in conjunction with axial sections may increase the risk of missing additional diagnoses, especially if the interpreting radiologists are not yet accustomed to primary evaluation of coronal reformations.
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
CiteULike 
Complore 
Connotea 
Del.icio.us 
Digg 
Reddit 
Technorati What's this?
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
