| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
1 Department of Radiology, Asan Medical Center, University of Ulsan, 388-1
Poongnap-dong, Songpa-gu, Seoul, 138-736, South Korea.
2 Department of Urology, Asan Medical Center, University of Ulsan, Seoul,
138-736, South Korea.
Received October 30, 2001;
accepted after revision March 5, 2002.
Address correspondence to K.-S. Cho.
Abstract
 Top Abstract IntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
SUBJECTS AND METHODS. Seventy-three consecutive patients who had gross hematuria and whose upper urinary tracts had a normal appearance on single-detector helical CT scans were prospectively evaluated with virtual cystoscopy. Source CT data for virtual cystoscopy were obtained on a multidetector CT scanner with 1.25-mm slice thickness and transferred to a workstation for interactive navigation using volume rendering. Two radiologists independently interpreted the virtual cystoscopic images, and discrepancies were resolved by consensus. All patients also underwent conventional cystoscopy. We assessed the agreement between the findings on virtual and conventional cystoscopy. Using conventional cystoscopy as the gold standard, we evaluated the usefulness of virtual cystoscopy as an aid in identifying bladder lesions and detecting abnormal bladders.
RESULTS. Virtual cystoscopy depicted 60 lesions in the bladders of
43 patients. Fifty-six lesions (in 41 bladders) revealed on virtual cystoscopy
were true-positive findings. Four lesions in two bladders with abnormal
findings were false-positive. On virtual cystoscopy, the radiologists missed
three lesions in two abnormal bladders that were identified on conventional
cystoscopy. On virtual cystoscopy, 15 (88%) of 17 lesions smaller than 0.5 cm
were identified. The agreement between the findings of virtual and
conventional cystoscopy was excellent in the reviewers' identification of
bladder lesions (
= 0.83) and detection of abnormal bladders ( =
0.89). The sensitivity and specificity of virtual cystoscopy were 95% and 87%
for identifying bladder lesions and 95% and 93% for detecting abnormal
bladders.
CONCLUSION. Virtual cystoscopy of the contrast material—filled bladder is useful for the evaluation of the bladder in patients with gross hematuria.
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
CT is usually recommended as a useful radiologic approach for assessing hematuria, but previous reports [1,2,3,4] have shown that CT has low sensitivity for detection of small bladder lesions. For CT to depict a small bladder lesion, optimal imaging conditions, including adequate bladder distention and thin-slice scanning, must be satisfied. Therefore, negative findings on CT warrant performance of conventional cystoscopy in patients with hematuria.
Recent advances in CT hardware and software have led to the development of various forms of virtual reality imaging techniques [3,4,5,6,7,8,9,10,11,12]. Since Vining et al. [3] first described the method, investigators have evaluated the usefulness of virtual cystoscopy for detecting bladder tumors [3,4,5,6,7,8]. Some authors have reported that lesions as small as 3 mm can be detected on virtual cystoscopy [4]. However, previous studies were designed to retrospectively evaluate bladder lesions that had been already confirmed on conventional cystoscopy.
To date, two techniques have been used to obtain the CT source data for reconstructed virtual cystoscopic images, scanning the bladder that has been filled with either air or contrast material [3,4,5,6,7,8]. Most previous studies have chosen to scan the air-filled bladder. However, virtual cystoscopy of the air-filled bladder is inherently invasive because catheterization is required to introduce air into the bladder. On the other hand, filling the bladder with IV contrast material has been easily achieved in many studies. Virtual cystoscopy of the contrast material—filled bladder appears to be more convenient, less invasive, and equally effective as that of the air-filled bladder.
In our study, we sought to evaluate the usefulness of virtual cystoscopy of the contrast material—filled bladder for evaluating patients with gross hematuria. Using a volume-rendering algorithm, we reconstructed virtual cystoscopic images of the contrast material—filled bladders from contrast-enhanced CT scans. We then prospectively examined the bladder in patients with gross hematuria.
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
The patients underwent single-detector helical abdominopelvic CT.
Unenhanced scans with a 3-mm interval and contrast-enhanced scans with a 5-mm
interval were acquired 2 min after IV injection of 120 mL of contrast material
([iopamidol] Iopamiro 300; Bracco, Milano, Italy). Two radiologists reviewed
the CT scans and decided to perform virtual cystoscopy after confirming for
each patient that no urolithiasis was present and that the renal parenchyma,
renal calices, and ureters were normal. Additional CT scans needed for virtual
cystoscopy were obtained when a patient expressed a desire to void (
90-120 min after IV injection of contrast material). Within a week after
performance of virtual cystoscopy, urologists who were unaware of virtual
cystoscopic findings performed conventional cystoscopy in all patients.
Virtual Cystoscopy
Before undergoing virtual cystoscopy, all patients were asked to alternate
taking supine and prone positions four times so that the contrast material and
urine in the bladder could be adequately mixed. Thereafter, scans were
obtained using a multidetector CT (MDCT) scanner (LightSpeed QX/i; General
Electric Medical Systems, Milwaukee, WI) with the patient in a supine
position. Scanning parameters included an 0.8-sec gantry rotation speed,
high-quality scanning mode (pitch, 3), 1.25 mm x 4 detector array,
3.75-mm-per-rotation table speed, 1.25-mm reconstruction interval, 512 x
512 matrix, 250 mA, and 120 kVp. The scanning covered the entire area of the
urinary bladder for 5-6 sec.
The CT data sets were transferred to a workstation (Advantage Windows 3.0, General Electric Medical Systems) for virtual cystoscopy reconstruction using a volume-rendering algorithm. We adjusted the attenuation-coefficient range for voxel categorization to the contrast material in the bladder until normal mucosal surfaces appeared smooth and no noise was seen in the lumen (Fig. 1).
|
Two experienced genitourinary radiologists (who were trained in interactive navigation and interpretation of three-dimensional virtual reality imaging and who had performed at least 20 previous virtual cystoscopic examinations) independently interpreted the virtual cystoscopic images. Discrepancies were resolved by consensus.
We divided the bladder wall into six segments: the anterior, posterior, superior, inferior, right, and left walls. The camera for virtual cystoscopy was placed in the center of the bladder lumen and thereafter was advanced to each of the six locations in turn. When a possible abnormality was discovered, it was fully evaluated from various angles.
Analysis
The virtual and conventional cystoscopic findings for each patient were
documented on separate worksheets. The number, location, morphology, size, and
pathologic findings of the lesions were recorded. The morphology was described
as either polypoid or sessile. The polypoid descriptor indicated a lesion that
protruded into the bladder lumen from a narrow base on the bladder wall,
whereas the sessile descriptor indicated a slightly elevated lesion with an
irregular surface and a broad base on the bladder wall.
We compared the performances of the two reviewers and results of virtual and conventional cystoscopy with regard to identifying bladder lesions and detecting abnormal bladders. To evaluate the agreement, we applied the kappa statistic; a kappa value of less than 0.20 was considered poor, 0.21-0.40 was considered fair, 0.41-0.60 was considered moderate, 0.61-0.80 was considered good, and 0.81-1.00 was considered excellent.
Using conventional cystoscopy as the gold standard, we evaluated the sensitivity, specificity, positive predictive value, negative predictive value, and accuracy for virtual cystoscopy as an aid in identifying bladder lesions and detecting abnormal bladders.
We evaluated the time required for navigating the bladder and the quality of virtual cystoscopic images. The assessment of image quality was based on the degree of bladder distention and the mixing of the contrast material and urine.
The single-detector helical abdominopelvic CT scans were interpreted with regard to identifying bladder lesions and detecting abnormal bladders by a third radiologist who was unaware of the virtual cystoscopic findings.
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
Pathology reports indicated a diagnosis of transitional cell carcinoma in 52 lesions, cystitis in three, adenocarcinoma in three, and lymphoepithelioid carcinoma in one. Among six sessile lesions detected on conventional cystoscopy were one diffuse lesion that was diagnosed as cystitis and two focal lesions that were diagnosed as transitional cell carcinoma.
Of 30 patients with bladders that appeared normal on conventional cystoscopy, 17 patients had normalized RBC in a urinalysis performed 2-3 weeks after conventional cystoscopy without undergoing any specific treatment, 11 patients were found to have primary renal parenchymal disease including IgA nephropathy (n = 7) and minimal glomerular change (n = 4), and two patients were found to have nutcracker syndrome.
The lower limit of the attenuation-coefficient range for voxel categorization on virtual cystoscopy was 100-250 H; the upper limit range was 1250-1400 H. Because the attenuation-coefficient of the bladder lumen was not constant from patient to patient, variable ranges were tried for each case to attain the optimal setting.
Using virtual cystoscopy, the two reviewers initially agreed on the
identification of 28 normal bladders and 57 lesions in 41 abnormal bladders.
By consensus, the reviewers decided that three additional lesions were present
in the bladders of two patients and that two additional bladders were normal.
(The agreement between the two readers was excellent: = 0.90;
p = 0.00; agreement, 96%.) Thus, the two reviewers identified 60
lesions in 43 abnormal bladders and 30 normal bladders using virtual
cystoscopy.
On virtual cystoscopy, 10 patients had multiple lesions: five patients had two lesions and the other patients had three lesions each. As they had on conventional cystoscopy, the reviewers found on virtual cystoscopy three sessile lesions with diffuse involvement of the bladder; which we regarded as one lesion in each patient. The locations of the 57 focal lesions included the posterior wall (n = 23 lesions), inferior wall (n = 13), right wall (n = 10), left wall (n = 5), anterior wall (n = 4), and superior wall (n = 2). The morphology was polypoid (Fig. 2) in 53 lesions and sessile (Fig. 3A,3B) in seven lesions. The diameter of the polypoid lesions ranged from 0.2 to 5.0 cm (2.2 ± 1.6 cm). The diameter of the four localized sessile lesions ranged from 2.5 to 3.5 cm (3.1 ± 0.4 cm).
|
|
|
In identifying bladder lesions, virtual cystoscopy could depict 56 of 59 lesions noted on conventional cystoscopy and generated four false-positive findings of lesions. Among the 17 lesions smaller than 0.5 cm at conventional cystoscopy, 15 (88%) were detected on virtual cystoscopy (Fig. 4A,4B). Three false-negative lesions, confirmed as transitional cell carcinomas, consisted of two polypoid lesions with diameters of 0.2 cm and 0.5 cm, respectively, and one focal sessile lesion with a diameter of 2 cm. This sessile lesion was accompanied by a true-positive lesion. The false-positive lesions included two sessile and two polypoid lesions, two of which were accompanied by other true-positive lesions. Therefore, in determining abnormal bladders, the reviewers' interpretations of virtual cystoscopic data agreed with their interpretations of conventional cystoscopic data for 41 abnormal and 28 normal bladders. Two abnormal bladders were misdiagnosed as normal, and two normal bladders were misinterpreted as abnormal.
|
|
On retrospective evaluation of three false-negative lesions, the reviewers identified the 0.5-cm polypoid lesion and 2.0-cm sessile lesion; the 0.2-cm polypoid lesion was not visualized. On retrospective evaluation of the four false-positive lesions, the misinterpretation of two sessile lesions was thought to have resulted from the inadequate mixing of contrast material and urine. The misinterpretation of one polypoid lesion was thought to have been generated by extrinsic compression by an enlarged prostate. An air bubble that entered the bladder during catheterization for cystography (performed 3 days before virtual cystoscopy) was believed to have resulted in misinterpretation of the other polypoid lesion (Fig. 5A,5B).
|
|
The agreement between virtual and conventional cystoscopic findings was
excellent for both the identification of bladder lesions ( = 0.84;
p = 0.00; agreement, 91%) and determination of abnormal bladders
( = 0.93; p = 0.00; agreement, 96%). Considering conventional
cystoscopy to be the gold standard, we found the following diagnostic values
for the identification of bladder lesions on virtual cystoscopy: sensitivity,
95%; specificity, 87%; positive predictive value, 93%; negative predictive
value, 90%; and accuracy, 93%. For determination of abnormal bladders on
virtual cystoscopy, the sensitivity was 95%; specificity, 93%; positive
predictive value, 95%; negative predictive value, 93%; and accuracy, 95%.
The average time required for navigating the bladder was 3.4 ± 1.2 min, and every navigation took less than 5 min. In 67 (92%) of 73 patients, virtual cystoscopic images were optimal because the bladder distention and mixing of contrast material and urine were adequate. All six suboptimal images were caused by inadequate mixing of contrast material and urine, which resulted in artifacts on virtual images and fluid—fluid levels on source images. In these patients, we applied two sets of attenuation-coefficient range for voxel categorization, one for the supernatant layer and the other for the precipitating layer (Fig. 6A,6B,6C).
|
|
|
The single-dector helical abdominopelvic CT could reveal only 30 (51%) of 59 lesions and 25 (58%) of 43 abnormal bladders identified on conventional cystoscopy. All 17 lesions smaller than 0.5 cm and three focal sessile lesions were missed. All lesions detected on the single-detector helical abdominopelvic CT scans were also identified on virtual cystoscopy (Fig. 7A,7B).
|
|
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
The urinary bladder is a good candidate for virtual endoscopy because of its simple luminal morphology, its relatively small volume, and the absence of involuntary peristalsis. Therefore, a virtual cystoscopic rendering of the bladder takes a short time to navigate and does not require that the operator have great skill. We found that after about 20 experiences of virtual cystoscopic navigation, we could complete an examination in less than 5 min.
The usefulness of virtual cystoscopy for bladder tumor detection has
already been proven
[3,4,5,6,7,8];
however, previous studies have focused solely on known bladder lesions. To our
knowledge, our study is the first prospective trial to evaluate the usefulness
of virtual cystoscopy. In our prospective study, findings on virtual
cystoscopy of the contrast-filled bladder showed excellent agreement with
those on conventional cystoscopy. Reviewers could identify almost all bladder
lesions missed on routine single-detector helical abdominopelvic CT. These
results indicate that virtual cystoscopy is worthy of being part of a
comprehensive CT examination of the urinary tract despite the extra time
(
2 hr), cost, and radiation exposure that it requires. We suggest that use
of virtual cystoscopy may obviate invasive procedures by providing reviewers
with data that allow them to determine whether a patient's bladder has
abnormalities. In addition, virtual cystoscopy can provide information
regarding the lesion size, location, and shape to the surgeons who perform
conventional cystoscopy.
In our study, the detection rate for lesions smaller than 0.5 cm was higher than in a previous study (88% vs 60%) [5]. We suggest that this improvement is due primarily to the methods of CT data acquisition and to the reconstruction algorithm. For CT data acquisition, we used an MDCT scanner that allows 1.25-mm slice thickness, whereas previous studies [3,4,5, 7, 8] had used a single-detector CT scanner to obtain data of 3- to 5-mm slice thickness. With the MDCT scanner, we were able to achieve higher resolution in a shorter scanning time.
For the reconstruction of virtual images, we used a volume-rendering algorithm, whereas previous studies used a surface-rendering algorithm. Volume rendering has an advantage over surface rendering in that volume rendering retains all the raw data [13, 14]. Recently, Hopper et al. [15] compared the mucosal detail on virtual endoscopic images achieved with volume rendering with that achieved with surface rendering; they found that the mucosal detail was better with volume rendering. In the past, surface rendering was frequently used for virtual imaging because it is easy to apply and does not require expensive computer systems. However, surface rendering has limitations; this algorithm is sensitive to artifacts and noise, and the quality of mucosal detail may decrease, depending on the threshold range [15]. Although volume rendering requires more powerful hardware and software than does surface rendering, recent technologic developments allow volume rendering to be applied at commercially available workstations or on personal computer—based three-dimensional image reconstruction systems.
In our study, we used a contrast material—filled bladder for virtual cystoscopy, whereas most previous studies had used the air-filled bladder. Compared with those previous studies, our method has some advantages. First, our method is safer and more comfortable for the patient because bladder catheterization is not necessary. Second, the radiation dose in our method can be halved: CT data are obtained only once, whereas virtual cystoscopy of an air-filled bladder requires two sets of CT data obtained with the patient in supine and prone positions. Third, virtual cystoscopy can be performed as part of routine contrast-enhanced CT, and thus a satisfactory evaluation of the entire urinary tract can be achieved with only one examination.
In our experience, the most important preparation for optimal image quality in the contrast material—filled bladder is the adequate mixing of contrast material and urine, which can be obtained by having the patient make several position changes. Inadequate mixing of contrast material and urine results in a fluid—fluid level visible on the source images and an artifact on virtual cystoscopic images. In this circumstance, interactive navigation may be achieved by applying two sets of the attenuation-coefficient range for voxel categorization, including a set for the supernatant layer and another set for the precipitating layer; nevertheless, suboptimal quality in cases of inadequate mixing is unavoidable. In our study, we asked patients to take alternately supine and prone positions four times just before CT was performed. However, in some patients with a large amount of natural urine in the bladder, more active positional change and evaluation of the bladder on CT before source images are acquired are necessary. In addition, moderate voiding before IV injection of the contrast material may be helpful.
Previous reports have pointed out the difficulty of detecting sessile lesions as a potential limitation of virtual cystoscopy [5,6,7,8]. In our study, we identified five of the six sessile lesions, although two false-positive lesions were also identified. Sessile lesions usually have an irregular surface that must be minutely depicted on virtual cystoscopy. Various factors influence the detection of sessile lesions, including the method used to acquire the CT data, interactive navigational skill of the operator, attenuation-coefficient ranges used for voxel categorization, and degree of bladder distention. We suggest that adequate bladder distention and attenuation-coefficient range for voxel categorization are the more important factors for revealing sessile lesions. Insufficient distention of the bladder may cause the mucosa to wrinkle, resulting in pseudolesions. The optimal attenuation-coefficient value that differentiates the bladder wall from contrast material in the lumen varies according to the concentration of contrast material in the lumen. A higher attenuation-coefficient range for voxel categorization will assign a minute surface irregularity on the mucosa into the contrast material and vice versa [14, 15]. Therefore, to set the optimal attenuation-coefficient range for voxel categorization, various trials must be performed for every patient.
Limitations of virtual cystoscopy of a contrast material—filled bladder include contraindication of the modality for patients who cannot tolerate IV contrast material injection. Additionally, in patients who cannot easily change position, the image quality of virtual cystoscopy is inevitably suboptimal because of inadequate mixing of the contrast material and urine. Third, using MDCT with a thin slice thickness may increase the radiation dose despite a shorter scanning time. Finally, possible scheduling problems may arise in a busy CT practice because of the repeated patient positioning and scanning required.
In addition, our study design has a potential limitation in patient selection. We did not include patients who had undergone prior bladder surgery, radiation, or biopsy, all of which may complicate virtual cystoscopic analysis.
In conclusion, virtual cystoscopy of the contrast material—filled bladder after CT is a promising imaging modality for bladder evaluation in patients with gross hematuria. The adequate mixing of contrast material and urine, attenuation-coefficient range used for voxel categorization, and degree of bladder distention are the key factors for obtaining a satisfactory evaluation.
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
CiteULike 
Complore 
Connotea 
Del.icio.us 
Digg 
Reddit 
Technorati What's this?
This article has been cited by other articles:
|
|
 |
|
  M. E. Abou-El-Ghar, A. El-Assmy, H. F. Refaie, and T. El-Diasty Bladder Cancer: Diagnosis with Diffusion-weighted MR Imaging in Patients with Gross Hematuria Radiology, May 1, 2009; 251(2): 415 - 421. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
G. Francica, F. Scarano, C. A. Sadow, and S. G. Silverman Bladder Cancer Detection with CT Urography Radiology, March 1, 2009; 250(3): 956 - 957. [Full Text] [PDF]
|
|
|
|
 |
|
 G. Francica, S. A. Bellini, F. Scarano, A. Miragliuolo, F. A. De Marino, and M. Maniscalco Correlation of Transabdominal Sonographic and Cystoscopic Findings in the Diagnosis of Focal Abnormalities of the Urinary Bladder Wall: A Prospective Study J. Ultrasound Med., June 1, 2008; 27(6): 887 - 894. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
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]
|
|
|
|
|
|
E. Kocakoc, A. Kiris, I. Orhan, A. K. Poyraz, H. Artas, and F. Firdolas Detection of Bladder Tumors With 3-Dimensional Sonography and Virtual Sonographic Cystoscopy J. Ultrasound Med., January 1, 2008; 27(1): 45 - 53. [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]
|
|
|
|
|
|
M. Jinzaki, A. Tanimoto, H. Shinmoto, Y. Horiguchi, K. Sato, S. Kuribayashi, and S. G. Silverman Detection of Bladder Tumors with Dynamic Contrast-Enhanced MDCT Am. J. Roentgenol., April 1, 2007; 188(4): 913 - 918. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
J. K. Kim, S.-Y. Park, H. S. Kim, S. H. Kim, and K.-S. Cho Comparison of Virtual Cystoscopy, Multiplanar Reformation, and Source CT Images with Contrast Material-Filled Bladder for Detecting Lesions Am. J. Roentgenol., September 1, 2005; 185(3): 689 - 696. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
