| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Original Research |
1 Department of Radiology, Asan Medical Center, University of Ulsan, 388-1
Poongnap-dong, Songpa-gu, Seoul 138-736, South Korea.
2 Department of Radiology, Inje University, Sanggyepaik Hospital, 761-1
Sanggye-7 dong, Nowon-gu, Seoul 139-707, South Korea.
Received August 24, 2004;
accepted after revision October 27, 2004.
Address correspondence to J. K. Kim.
Abstract
 Top Abstract IntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
SUBJECTS AND METHODS. Two observers independently evaluated 47 patients (28 men and 19 women; mean age ± SD, 59 ± 16 years) with virtual cystoscopy, multiplanar reconstruction, and source CT images acquired with contrast material-filled bladder using an MDCT scanner (detector array, 4 x 1.25 mm; beam pitch, 0.75). Agreement between the two observers was evaluated for the three reconstruction methods using kappa statistics. Using the conventional cystoscopic findings as a reference, we compared the results of the three reconstruction techniques both by bladder site and by patient using the McNemar test.
RESULTS. The interobserver agreement for the number of positive
sites was excellent for virtual cystoscopy (
= 0.816), fair for
multiplanar reconstruction ( = 0.461), and good for source CT images
= 0.676). For both observers, the sensitivity for lesion detection by
bladder site was significantly greater with virtual cystoscopy (observer 1,
95%; observer 2, 90%) than with multiplanar reconstruction (78% and 60%) and
source CT (68% and 65%) images (p < 0.05), whereas the specificity
by bladder site and the sensitivity and specificity by patient did not differ
with the three methods (p > 0.05). For determining the presence or
absence of lesion at each site, virtual cystoscopy was more accurate than
multiplanar reconstruction and source CT images for both observers (p
< 0.05).
CONCLUSION. Virtual cystoscopy is more accurate than multiplanar reconstruction and source CT images for the detection of lesions in the bladder.
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
Virtual cystoscopy examination requires training for navigation and obtaining qualified source image data during a short period, even though recent software has made navigation easy and possible in a short time. If 2D display of the mucosal surface of the bladder, such as source CT images or multiplanar reformation (multiplanar reconstruction), is comparable to virtual cystoscopy, it would be unnecessary to spend time and efforts to obtain data for 3D display in busy daily practice. In previous studies of CT colonoscopy, multiplanar reconstruction showed accuracy similar to that of virtual endoscopy [6-8]. However, to our knowledge, no study has systemically compared the accuracy for bladder evaluation in virtual cystoscopy versus multiplanar reconstruction or source CT images.
In this study, we compared the diagnostic accuracy of virtual cystoscopy, multiplanar reconstruction, and source CT images in the detection of bladder lesions. Ultimately, we intended to evaluate whether virtual cystoscopy is really necessary for bladder evaluation.
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
Patient Population
Our study population consisted of 47 consecutive patients (28 men and 19
women; age range, 35-72 years; mean age ± SD, 59 ± 16 years) who
visited the urology department due to painless gross hematuria between July
2002 and February 2003 and were prospectively recruited for this study. All
patients showed innumerable RBCs in the urinalyses and thereby underwent MDCT
examinations.
CT Scans Dedicated to Bladder Evaluation
All CT data were obtained on a 4-MDCT scanner (LightSpeed QX/i, GE
Healthcare). CT examination included unenhanced scanning covering the entire
urinary tract (detector array, 4 x 5 mm; beam pitch, 1.5),
contrast-enhanced scanning covering the abdomen and pelvis (detector array, 4
x 5 mm; beam pitch, 1.5) at a scan delay of 100-120 sec, and delayed
scanning dedicated for bladder evaluation; in this study, only delayed CT
images dedicated for bladder evaluation were evaluated. IV contrast material
(Ultravist 300 [iopromide], Schering; or Iopamiro 300 [iopamidol], Bracco) was
administered into an antecubital vein by a power injector at a dose of 2 mL/kg
of body weight at a rate of 3 mL/sec to a maximum of 160 mL.
After the first two scans were obtained, each patient got down from the CT machine and then was asked to leave the CT room during the next patients' examinations. Then, the patient returned to the CT room when he or she felt a desire to void (90-140 min after IV injection of contrast material), and delayed CT was performed for bladder evaluation.
Immediately before this CT examination, all patients, with the help of a nurse, were asked to alternately take the supine and prone positions four times on the bed beside the CT machine to obtain adequate mixing of the contrast material and urine in the bladder. Thereafter, CT scans were obtained with the patient in a supine position. Immediately after completing this CT examination, the technicians in the CT room decided whether the contrast material and urine were adequately mixed in the bladder. Adequate mixing was indicated when high attenuation filled the entire bladder lumen homogeneously and there was no fluid-fluid level caused by unopacified urine. Scanning parameters for this examination included a beam pitch of 0.75 (equivalent to a slice pitch of 3 in high-quality mode), a detector array of 4 x 1.25 mm, a table speed of 15 mm per rotation (18.75 mm/sec), a reconstruction interval of 1 mm, an X-ray tube voltage of 120 kVp, and a tube current of 210-240 mA. The scanning covered the entire area of the urinary bladder.
The radiation dose was evaluated using the weighted CT dose index, CTDIw, which was defined by the International Electrotechnical Commission [9]. The value of CTDIw for each scanning technique was shown on the CT console.
Bladder Evaluation
The CT data were transferred to a workstation (Advantage Windows 3.0.5, GE
Healthcare) for reconstructing virtual cystoscopy and multiplanar
reconstruction images and to a PACS (PetaVision, Hyundai Information &
Technology) for the review of source CT images.
To systemically indicate the location of bladder lesion, we identified areas of the bladder wall by evenly dividing the bladder wall into six sites, including anterior, posterior, superior, inferior, right, and left sites. The presence or absence of a bladder lesion at each site was evaluated independently on conventional cystoscopy, virtual cystoscopy, multiplanar reconstruction, and source CT images according to the same location criteria; the results of each examination were then recorded separately. The presence of a bladder lesion (positive finding) was indicated when either a surface irregularity or a polypoid lesion was noted. In addition, color change of the bladder mucosa on cystoscopic findings or focal wall thickening on multiplanar reconstruction or source CT images was regarded as a positive finding.
Urologists performed conventional cystoscopy within a week after CT examination. On the basis of cystoscopy reports, a radiologist recorded the number and location of positive sites and the sum of tumor diameter in each positive site. This study evaluated the accuracy for lesion detection by bladder site but not by lesion, because we suggested that it would be complicated to measure the number and size of lesions and evaluate the detection accuracy by each lesion when many small lesions would be clustered in a focal area. Tumor size at each site was represented as the sum of the greatest diameter of tumors. Thereafter, two radiologists unaware of patient information and cystoscopic findings independently evaluated the bladder using virtual cystoscopy, multiplanar reconstruction, and source CT images. The first observer was a staff radiologist who was highly experienced in virtual cystoscopic navigation because he had examined more than 200 patients, whereas the second observer was a fellow who had performed approximately 40 virtual cystoscopy examinations. Both of these radiologists were experienced in interpreting multiplanar reconstruction images.
Bladder evaluation with the three image reconstruction methods was performed in a random order with an interval of 3 weeks. The observers were forced to complete their interpretation in each session within 10 min.
Two radiologists independently performed virtual cystoscopy examinations on the workstation using a volume-rendering technique. Operators adjusted the attenuation-coefficient range for voxel-categorization to the contrast material in the bladder until the normal mucosal surface appeared smooth and no noise was seen in the lumen. The lower attenuation-coefficient limit varied from 150 to 250 H and the upper limit from 1,250 to 1,400 H. Because attenuation-coefficient of the bladder lumen was not constant between patients, variable ranges were tried in every case to reach the optimal setting.
Then interactive navigation and interpretation of 3D virtual reality imaging were performed. The camera of virtual cystoscopy was located in the center of the bladder lumen and navigated six sites in turn. When any abnormality was identified, it was fully evaluated in various projections.
Multiplanar reconstruction images were also independently reconstructed and interpreted on a workstation by the two radiologists. Multiplanar reconstruction images were reconstructed in the transverse, coronal, and sagittal planes at a slice thickness of 1 mm with an interslice gap. If needed, images in other planes were reconstructed at the discretion of the observers within a given time limit.
Source CT images were also independently reviewed by the two radiologists. During the review, they were allowed to adjust the brightness and contrast level of the images and to use the localized magnification.
Statistical Analysis
For the number of positive sites in each patient, the interobserver
agreement was evaluated in virtual cystoscopy, multiplanar reconstruction, and
source CT images using the kappa statistics; a kappa value of less than 0.20
was considered as poor, 0.21-0.40 as fair, 0.41-0.60 as moderate, 0.61-0.80 as
good, and 0.81-1.00 as excellent.
Using the conventional cystoscopic findings as a reference, we calculated the sensitivity, specificity, positive predictive value, negative predictive value, and overall accuracy of virtual cystoscopy, multiplanar reconstruction, and source CT images for the detection of lesions by bladder site and by patient. Then the sensitivity and specificity were compared in the three reconstruction methods using the McNemar test.
In addition, the accuracy of virtual cystoscopy, multiplanar reconstruction, and source CT images for determining the presence or absence of a lesion at each bladder site was evaluated. The result of each reconstruction method at each site was classified as "correct" or "incorrect"; the result of reconstruction method was considered to be correct when it was concordant to that of conventional cystoscopy and vice versa. Thereafter, the accuracy was compared in the three reconstruction methods using the McNemar test.
In every statistical analysis, significance was considered to be present when the p value was less than 0.05.
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
Consistency and Radiation Dose of Virtual Cystoscopy
In all patients, the mixing of the contrast material and urine was adequate
on the supine scan and the bladder was adequately distended. Therefore,
virtual cystoscopy examination with contrast material-filled bladder was
available in all patients. The value of CT-DIw for each
bladder scanning was 21.4-24.5 mGy. This value was dependent only on the
variation of tube current (210-240 mA) because the other scanning parameters
were fixed during this study period.
Interobserver Agreement
With virtual cystoscopy, the two observers agreed that 19 patients had
normal bladders, 11 patients had a single positive site, seven patients had
two positive sites, two patients had three positive sites, one patient had
four positive sites, and one patient had six positive sites. Therefore, the
two observers agreed on the number of positive sites in 41 (87%) of the 47
patients.
With multiplanar reconstruction, the two observers agreed on the number of positive sites in 30 (64%) of the 47 patients, including 14 patients without a positive site, 11 with a single positive site, three with two positive sites, one with three positive sites, and one with six positive sites.
With source CT images, the two observers agreed in 22 patients without a positive site, nine with a single positive site, three with two positive sites, two with three positive sites, and one with six positive sites. Therefore, the two observers agreed on the number of positive sites in 37 (79%) of the 47 patients.
The interobserver agreement was excellent for virtual cystoscopy ( =
0.816 ± 0.070 [mean ± SD]), fair for multiplanar reconstruction
( = 0.461 ± 0.102), and good for source CT images ( =
0.676 ± 0.084).
Lesion Detection by Bladder Site
Table 1 shows the results of
virtual cystoscopy, multiplanar reconstruction, and source CT images for
lesion detection by bladder site. For the first observer, the sensitivity for
lesion detection was significantly greater with virtual cystoscopy (95%) than
multiplanar reconstruction (78%) (p = 0.039) and source CT images
(68%) (p = 0.001), whereas it was similar between multiplanar
reconstruction and source CT images (p = 0.344). The specificity was
similar in the comparison pairs of the three reconstruction methods
(p > 0.05). In 11 positive sites with the sum of tumor diameter
less than 0.5 cm, the first observer identified nine sites (82%) with virtual
cystoscopy, seven sites (64%) with multiplanar reconstruction, and six sites
(55%) with source CT images.
|
In the second observer, the sensitivity was also greater in virtual cystoscopy (90%) than in multiplanar reconstruction (65%) (p = 0.006) and source CT images (65%) (p = 0.013), whereas it was similar between multiplanar reconstruction and source CT images (p > 0.05). The specificity was similar in the comparison pairs of the three reconstruction methods (p > 0.05). In 11 positive sites with the sum of the tumor diameter less than 0.5 cm, the second observer identified eight sites (73%) with virtual cystoscopy and six sites (55%) with multiplanar reconstruction and source CT images.
Determination of Each Bladder Site as Positive or Negative
The comparison of the accuracy for determining the presence or absence of a
bladder lesion in three imaging methods is shown in
Table 2 (Figs.
1A,
1B,
1C,
1D,
2A,
2B,
2C,
3A,
3B,
3C,
4A, and
4B). In the first observer,
virtual cystoscopy was correct at 270 (96%) of 282 sites; multiplanar
reconstruction images, at 258 sites (91%); and source CT images, at 253 sites
(90%). The McNemar test showed that virtual cystoscopy was more accurate than
multiplanar reconstruction (p = 0.004) and source CT (p <
0.0001) images, whereas multiplanar reconstruction and source CT images showed
similar accuracy (p = 0.359).
|
|
|
|
|
|
|
|
|
|
|
|
|
In the second observer, virtual cystoscopy was correct at 266 (94%) of 282 sites; multiplanar reconstruction images, at 252 (89%) of 282 sites; and source CT images, at 252 sites (89%). The McNemar test showed that virtual cystoscopy was more accurate than multiplanar reconstruction (p = 0.007) and source CT images (p = 0.007), whereas multiplanar reconstruction and source CT images showed similar accuracy (p = 1.00).
Lesion Detection Results by Patient
Table 3 shows the results of
virtual cystoscopy, multiplanar reconstruction, and source CT images for
lesion detection by patient. In the first observer, the sensitivity and
specificity for identifying patients with a bladder lesion were 96% and 86%
for virtual cystoscopy, 96% and 86% for multiplanar reconstruction, and 80%
and 82% for source CT images, respectively. Statistical analysis showed no
significant difference in any of the sensitivity and specificity in the
comparison pairs of the three reconstruction methods (p >
0.05).
|
In the second observer, the sensitivity and specificity were 92% and 86% for virtual cystoscopy, 80% and 64% for multiplanar reconstruction, and 76% and 82% for source CT images, respectively. Neither the sensitivity nor the specificity was significantly different among the three reconstruction methods (p > 0.05).
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
To our knowledge, most previous studies assessing the accuracy of virtual cystoscopy evaluated the results by lesion [1-4, 10-12]. In contrast, this study evaluated the results by bladder site because the prior methods may be complicated when multiple small tumors are clustered in a focal area. Therefore, for estimating lesion size, we asked observers to assess the sum of tumor diameter in each site, instead of measuring the size of each tumor. We did not perform systemic statistical comparison of the accuracy of the three reconstruction methods according to tumor size because of the small number (n = 11) of sites with the sum of tumor diameter less than 0.5 cm. However, our results showed that observers detected more sites with the sum of the tumor diameter of less than 0.5 cm on virtual cystoscopy than multiplanar reconstruction and source CT images.
Virtual cystoscopy has advantages over multiplanar reconstruction and source CT images. First, virtual cystoscopy is dedicated for evaluating the mucosal surface of the bladder and therefore can detect superficial lesions missed by multiplanar reconstruction or source CT images. Second, virtual cystoscopy can allow the operator to navigate the mucosal surface in various projections. Third, interactive navigation is provided by virtual cystoscopy, and the operator can therefore make decisions more confidently.
In contrast to virtual cystoscopy, which navigates the mucosal surface, bladder evaluation on multiplanar reconstruction and source CT images is mainly based on identifying abnormal wall thickening. It is notable that this study showed similar accuracy for bladder lesion detection on source CT images and multiplanar reconstruction images, despite the advantage that multiplanar reconstruction images allow 3D evaluation with the combination of information from various planes. Our results suggest that visual perception of abnormal bladder wall thickening is limited for lesion detection, although recent studies have shown that computerized measurement of bladder wall thickness provided a relatively high sensitivity (80-89%) [10-12].
The most important advantages of virtual cystoscopy over conventional cystoscopy are that it is less invasive and is more comfortable for patients. Several complications can be caused by conventional cystoscopy, including infections, bladder perforation, scarring, and stricture of the urethra [13-16]. In contrast, few complications have been associated with virtual cystoscopy examinations [1-5]. To our knowledge, there was only one case of a complication related to catheter removal at virtual cystoscopy with air-filled bladder [4].
Despite those benefits of virtual cystoscopy, whether virtual cystoscopy can replace the role of conventional cystoscopy is still questionable. First, it is evident that conventional cystoscopy is superior to virtual cystoscopy because conventional cystoscopy can detect a lesion missed on virtual cystoscopy on the basis of its ability to show mucosal color changes. Second, the performance of virtual cystoscopy may be limited in patients who have bladder outlet obstruction and have undergone previous bladder surgery, radiation, or biopsy, all of which may distort the bladder surface and can cause pseudolesions. Furthermore, an enlarged prostate that indents the bladder bases would likely cause a false-positive diagnosis at virtual cystoscopy. Third, compared with multiplanar reconstruction and source CT images, virtual cystoscopy may be limited in evaluating the depth of a lesion. Last, virtual cystoscopy may be potentially criticized because of increased radiation dose, although radiation dose may not be such an important issue for virtual cystoscopy because the scanning area is limited to the bladder. Future investigations are necessary for a reasonable modification of the imaging protocol, such as a faster scan with a wider beam width, to reduce the radiation hazard.
Various methods and techniques can be used to obtain source image data for virtual cystoscopy [3-5, 7, 16-19]. CT of an air-filled bladder was initially introduced; thereafter, MRI of a urine-filled bladder and CT of a contrast material-filled bladder have been applied. Virtual cystoscopy with contrast material-filled bladder in this study is safer and more comfortable for patients because bladder catheterization is not necessary. In addition, the radiation dose for our method can be halved because CT data are obtained only once, whereas virtual cystoscopy with air-filled bladder requires two sets of CT data obtained with the patient in both the supine and prone positions. Virtual cystoscopy can be performed as part of a routine contrast-enhanced CT examination; thus, a satisfactory evaluation of the entire urinary tract can be achieved with only one examination. In contrast, virtual cystoscopy with contrast material-filled bladder may be limited by a risk of contrast-induced reaction and nephrotoxicity. In addition, our method may be criticized for possibly disrupting the schedule for a CT room, because additional CT required full contrast filling of the bladder at variable times of 90-140 min after completing the routine scanning. For resolving this problem, various methods to reduce examination time would be tried, such as administration of diuretics or saline to accelerate urine production.
In our study, the time spent on each reconstruction method was restricted to 10 min per patient for several reasons. First, we considered the equality of the interpretation time among the three reconstruction methods because observers are likely to spend more time for virtual cystoscopy than for either multiplanar reconstruction or source CT images. Such additional time may cause a bias that increases the diagnostic accuracy and confidence level. Second, we attempted to differentiate observer performance according to level of experience by not allowing the second observer (fellow) to overcome her shortage of experience by spending more time than the first observer (staff radiologist). Lastly, we urged the observers to interpret images in a restricted time to make the task as similar as possible to that of clinical practice.
There are limitations in this study. Our study may be biased in patient selection because we collected patients with a high risk of having bladder tumors because they had gross hematuria. Therefore, observers might evaluate the bladder with a high suspicion of bladder tumor. This bias can increase the sensitivity for bladder lesion detection.
Another potential limitation of this study is that involuntary memorization of some patients with characteristic findings cannot be neglected, although bladder evaluation was performed in a random order with 3-week time intervals.
Finally, there may an argument against the advantage of virtual cystoscopy over multiplanar reconstruction and source CT images. If the intended goal of virtual cystoscopy was to avoid unnecessary cystoscopy, the specificity is more important than sensitivity. Our data showed no significant differences in the specificities and negative predictive values among the three methods by either observer. Nevertheless, we believe that the advantage of virtual cystoscopy still exists over multiplanar reformation and source CT image because virtual cystoscopy provides higher interobserver agreement and higher sensitivity and current software allows easy usability of virtual cystoscopy in a short time.
In summary, virtual cystoscopy is superior to multiplanar reconstruction and source CT images for lesion detection in the contrast material-filled bladder. Given the merits of this advanced technology, virtual cystoscopy may contribute to initial screening for patients with hematuria and more accurate patient selection for invasive cystoscopy.
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
CiteULike 
Complore 
Connotea 
Del.icio.us 
Digg 
Reddit 
Technorati What's this?
This article has been cited by other articles:
|
|
 |
|
  S. Kim, L. L. Wang, J. P. Heiken, C. L. Siegel, C. F. Hildebolt, and K. T. Bae Opacification of Urinary Bladder and Ureter at CT Urography: Effect of a Log-rolling Procedure and Postvoiding Residual Bladder Urine Volume Radiology, June 1, 2008; 247(3): 747 - 753. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 C. Tsampoulas, A. C. Tsili, D. Giannakis, Y. Alamanos, N. Sofikitis, and S. C. Efremidis 16-MDCT Cystoscopy in the Evaluation of Neoplasms of the Urinary Bladder Am. J. Roentgenol., March 1, 2008; 190(3): 729 - 735. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
S. B. Park, J. K. Kim, H. J. Lee, H. J. Choi, and K.-S. Cho Hematuria: Portal Venous Phase Multi Detector Row CT of the Bladder A Prospective Study Radiology, December 1, 2007; 245(3): 798 - 805. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
