This web only provides the extract of this article. If you want to read the figures and tables, please reference the PDF full text on Blackwell Synergy. Thank you.

- Review -

Virtual endoscopy of the urinary tract

George C. Kagadis¹, Dimitrios Siablis², Evangelos N. Liatsikos³, Theodore Petsas², George C. Nikiforidis¹

¹Department of Medical Physics, ²Department of Radiology, ³Department of Urology, School of Medicine, University of Patras, GR 26500 Rion, Greece

Abstract

Technological breakthroughs have advanced the temporal and spatial resolutions of diagnostic imaging, and 3 dimensional (3-D) reconstruction techniques have been introduced into everyday clinical practice. Virtual endoscopy (VE) is a non-invasive technique that amplifies the perception of cross-sectional images in the 3-D space, providing precise spatial relationships of pathological regions and their surrounding structures. A variety of computer algorithms can be used to generate 3-D images, taking advantage of the information inherent in either spiral computed tomography or magnetic resonance imaging (MRI). VE images enable endoluminal navigation through hollow organs, thus simulating conventional endoscopy. Several clinical studies have validated the diagnostic utility of virtual cystoscopy, which has high sensitivity and specificity rates in the detection of bladder tumor. Published experience in the virtual exploration of the renal pelvis, ureter and urethra is encouraging but still scarce. VE is a safe, non-invasive method that could be applied in the long-term follow-up of patients with ureteropelvic junction obstruction, urinary bladder tumors and ureteral and/or urethral strictures. Its principal limitations are the inability to provide biopsy tissue specimens for histopathologic examination and the associated ionizing radiation hazards (unless MRI is used). However, in the case of endoluminal stenosis or obstruction, VE permits virtual endoluminal navigation both cephalad and caudal to the stenotic segment. To conclude, VE provides a less invasive method of evaluating the urinary tract, especially for clinicians who are less familiar with cross-sectional imaging than radiologists. (Asian J Androl 2006 Jan; 8: 31-38)

Keywords: computed tomography; three-dimensional imaging; virtual endoscopy; urethral stricture

Correspondence to: Dr George C. Kagadis, PhD, Department of Medical Physics, School of Medicine, University of Patras, GR 26500 Rion, Greece
Tel/Fax: +30-2610-996-106
E-mail: george.kagadis@med.upatras.gr
Received 2005-02-17 Accepted 2005-06-22
DOI: 10.1111/j.1745-7262.2006.00096.x

1 Introduction

Three-dimensional (3-D) reconstruction techniques appeared in published reports in the mid-1990s, but at that time the imaging techniques could not acquire continuous and complete sets of raw data, leading to pronounced artifacts in the final reconstruction. However, recent technological breakthroughs have advanced the temporal and spatial resolutions of diagnostic imaging, and 3-D reconstruction techniques have been introduced into everyday clinical practice.

Virtual endoscopy (VE) is a non-invasive technique that amplifies the perception of cross-sectional images, acquired by axial computed tomography (CT), in the 3-D space, providing precise spatial relationships of pathological regions and their surrounding structures. The use of appropriate software and relative algorithms produces virtual reality images, enabling endoluminal navigation through any hollow viscous, thus simulating conventional endoscopy. In addition, VE enables the depiction of endoluminal as well as extraluminal adjacent structures in all directions. It may also allow the diagnostic exploration of body regions that are either unaccessible or incompatible with conventional endoscopic procedures. VE has been applied to many hollow anatomical structures, such as trachea, colon, aorta, brain ventricles, nasal cavity and paranasal sinuses [1-8].

Here we reviewed various applications of VE in the urinary tract. A survey of published works on the PubMed database was performed using the following keywords: virtual, three-dimensional, endoscopy, nephroscopy, ureteroscopy, cystoscopy, and urethrosco-py. Seventy-five relevant publications from January 1996 to January 2005 were traced. The selection criteria included: (1) case reports describing the first applications of the technique; (2) novel small series; and (3) prospective clinical studies with well-defined endpoints. Our referred bibliography was confined to 35 articles. The advantages and limitations of VE in each setting were discussed.

2 3-D reconstruction techniques

Spiral computed tomography (SCT) and magnetic resonance imaging (MRI) provide continuous and complete sets of raw data that are transferred to a computer workstation for post-processing and analysis. Once the final 3-D dataset is obtained, a variety of computer algorithms can be used to generate 3-D images, taking ad vantage of the information inherent in either the SCT or MRI scan. The most commonly applied techniques are shaded surface display, maximum or minimum intensity projection and volume rendering [1-5].

VE is one of the most recent innovations in the field of post-processing techniques and provides supplementary information to those already mentioned. The main goal of VE was to develop a non-invasive diagnostic tool that would be easily tolerated by the majority of patients, by producing images similar to those acquired by the conventional endoscopy.

3 Virtual nephroscopy and ureteroscopy

Renal pelvic and ureteral tumors usually present with gross hematuria and pain. The conventional endoscopic examination used for the diagnosis of these tumors is an invasive and technically demanding procedure which has the potential risks of ureteral injury, hematoma, urinoma, ureteral obstruction and fistula [9, 10]. The use of a flexible ureteroscope makes access easier and can minimize patients’ discomfort and complications. However, investigators have explored non-invasive techniques, such as virtual nephroscopy and ureteroscopy, in an effort to overcome the shortcomings of the conventional endoscopic approach.

Published VE studies of the upper urinary tract are still limited. Takebayashi et al. [9, 10] pioneered this field by reporting the usefulness of CT nephroscopy and ureteroscopy in the diagnosis of malignancies of the renal pelvis and the ureters. Delayed SCT was performed after intravenous contrast media and a diuretic agent were given in order to achieve the dilatation of the pelvicaliceal system and homogeneous dense opacification of the ureters. CT nephroscopy and renal axial CT were able to detect 92 % and 83 % of tumors, respectively. They showed a good correlation of CT nephroscopic images with the pathological findings and concluded that CT nephroscopy can help in the preoperative planning of endourological treatment. However, CT nephroscopy could not evaluate tumor infiltration of the surrounding structures, renal parenchyma or other adjacent tissues. In the evaluation of ureteral tumors, CT ureteroscopy clearly depicted ureteral stenosis and allowed proximal and distal evaluation of the ureter to the stenotic lesion. Sensitivity for detecting ureteral tumors using CT ureteroscopy was 81 % and the specificity was 100 %. However, neither tumor infiltration beyond the ureteral wall, nor lesion texture or color could be adequately evaluated [10]. Evaluation of the upper urinary tract with VE may also be performed using non-contrast MR urography datasets. Neri et al. [11] reported that VE of the renal pelvis and calices was able to be performed in all the 26 cases on the site of the urinary obstruction. VE and optimal depiction of the ureter was able to be obtained from the ureteropelvic junction to the site of obstruction if the ureteral diameter was at least 5 mm. However, VE of MR urography datasets was limited by the degree of dilation of the ureter and by the occurrence of artifacts. Artifacts occurred at low ureteral diameters and the ureter was visualized as narrow or occluded. The non-dilated side could be partly explored in almost half of the cases. The advantage of MR urography, however, is that it does not require the administration of iodinated contrast media and that it avoids radiation hazards.

More recent publications have reported increased resolution of VE images of the upper urinary tract with the use of volume rendering algorithms [12, 13]. In these studies, authors reported their experience with the application of VE to evaluate ureteral patency after the treatment of upper urinary tract obstruction with the use of self-expandable metallic stents. VE findings concurred with the excretory urography findings and VE permitted accurate 3-D visualization of the stented area, and of the proximal ureter cephalad and caudal to the stent, from different angles. The main disadvantage reported by the authors was the inability of the method to differentiate structures with similar absorbing characteristics used in the CT acquisition settings, despite its ability to provide information about the presence of intraluminal stenosis. That is, ureteral wall structures were depicted with similar densities regardless of the underlying histopathology (normal urothelium, luminal encrustation, mucosal hyperplasia or tumoral infiltrations). VE does not differentiate the fine detail in the epithelial lining of anatomical structures, which can be visualized with conventional endoscopic procedures. VE is less invasive compared with endoscopy of the upper urinary tract and is probably superior to excretory urography. We have also recently reported virtual navigation within the pelvis and calices with the efficient depiction of any pelvicaliceal anatomic deformities [14]. Because of the dilation provided by the stent, the quality of data acquisition and VE images of the stented ureter were superior. Metal stents, due to their minimal mass density, caused reduced scat tering to the X-rays’ quantum energies. Application of specially modified CT reconstruction protocols helps to overcome artifacts from strut reflections [15].

4 Virtual cystoscopy (VC)

The gold standard method for investigating hematuria and detecting bladder tumors is conventional cystoscopy. Although flexible cystoscopy used for surveillance is very well tolerated by patients, the main drawbacks are the failure to evaluate adjacent structures, a5 % - 15 % risk of urinary tract infection, and patient’s discomfort and anxiety [16, 17]. Although conscious sedation is generally not required, it might sometimes be necessary to relieve pain and discomfort. Iatrogenic injury to the urethra and bladder might also occur [18]. Because of these shortcomings, many investigators proposed the use of VC for bladder malignancies. This technique is based on the use of images acquired mainly from CT scanners. The suggested protocols for bladder distension vary from the use of room air, carbon dioxide and, more recently, intravenously infused contrast agents. The use of contrast media is less invasive, more convenient than, and as effective as, the use of air or carbon dioxide insufflations. Although VE has the advantage of providing both endoluminal and extraluminal information, artifacts on virtual images may occur if inadequate mixing of urine and contrast material takes place, or when a metallic hip prosthesis is present [19].

Vining et al. [20] were the first to perform VC in 1996. After catheterization of the bladder, drainage of urine and insufflation of the bladder with carbon dioxide, CT of the pelvis was carried out in one healthy volunteer and two patients with already identified transitional cell carcinoma of the bladder. The authors succeeded in correctly detecting bladder tumors and established the diagnostic feasibility of VC. Since then, various investigators have reported on the feasibility, safety and accuracy of VE of the bladder and have suggested that VE may be clinically applied in the long-term surveillance of patients with bladder tumors [18, 19, 21, 22]. VC allows the assessment of tumor size, location and morphology and it has been shown to have a 97 % - 100 % sensitivity rate and a 93 % - 100 % positive predictive value [18-20]. The detection rate of VC depends on tumor size. It has been documented to range from 94 % for tumors larger than 1 cm, to 77 % for tumors less than 1 cm. Depending on the location of the tumor, either the supine or prone position is chosen for the CT scan [22].

In 1998, Merckle et al. [23] performed VC with contrast-enhanced CT datasets. Images were acquired in three phases: prior to contrast injection, in the arterial phase during intravenous injection of contrast medium, and in a delayed phase after 30 min. Sedimentation of the contrast medium in the bladder was prevented by the mobilization of the patients. The best visual results were acquired during the delayed phase because of the significant attenuation difference between the bladder lumen and the mucosa. Both conventional and virtual cystoscopy had a 100 % sensitivity rate for tumors greater than 0.5 cm. In 2004, the work of Nambirajan et al. [24] and Yazgan et al. [25] further strengthened the role of VC with the use of contrast media in the investigation of patients with hematuria and/or bladder tumors. However, the radiation dose, potential allergies to contrast medium, the lack of biopsy specimens, and the reduced sensitivity in detecting tumors (0.5 cm - 1.0 cm) in size remain the primary drawbacks of this procedure [23, 26].

One disadvantage common to all VC methods independent of bladder distension protocol is the inability to precisely depict sessile lesions or wall thickening. Schreyer et al. [16] developed an algorithm for color mapping of the thickness of the bladder wall, aiming to ameliorate the accuracy of the detection of subtle masses. To differentiate urine from the bladder wall, contrast medium was given through a catheter. The wall thickness was defined as the shortest distance between a voxel on the inner wall and any voxel on the outer wall. After setting a color scale for the thickness data, the surface could be visualized with different colors depending on the variant wall thickness. The results of this technique were in agreement with those from conventional cystoscopy. Fielding et al. [27] used color mapping of bladder wall thickness in an effort to find a correlation between wall thickness detected by VC and the possible presence of tumors [27]. They demonstrated that when VC images showed a bladder wall thickness of less than 5 mm the possibility of tumor on conventional cysto-scopy was 10 %, whereas areas of bladder with a wall thickness greater than 5 mm had an 80 % possibility of revealing a suspicious region on conventional cystoscopy. The authors suggested that flexible or rigid cystoscopy would be obviated if virtual CT cystoscopy was negative.

As indicated, one of the major disadvantages of CT VE is the radiation dose. With the aim to minimize the hazard of ionizing radiation, several researchers com pared VC at regular (240 mAs) versus reduced (43 mAs-70 mAs) milliampere settings [28, 29]. They demonstrated an almost equivalent rate of sensitivity (94 % - 100 %) and specificity (100 %) regardless of the reduced milliampere settings. They also succeeded in minimizing the effective dose to less than 0.5 mSv [29], making the VE technique appropriate for long-term follow-up studies. In the earlier work of Homer et al. [30] the average effective dose of CT urography was 4.95 mSv compared with 1.48 mSv used in intravenous urography.

As far as complications are concerned, Song et al. [17] were the first to report a bleeding complication without any clinical sequela during VC, related to catheter removal. This is the only complication traced by our review, which indicates the high safety profile of VE.

The published experience of VC also includes MRI-based studies. A high consistency of MRI-based VC in the depiction of bladder tumors has been reported [31]. A comparison of MRI and CT cystoscopy with axial CT images and conventional cystoscopy for the detection of bladder tumors validated that the findings at MRI cystoscopy concurred with those of conventional cysto-scopy [32]. When compared with axial images and CT cystoscopy, MRI cystoscopy did not reveal any significant difference in the detection of polyps that were larger than 1 cm. However, MR cystoscopy showed decreased sensitivity and specificity in the detection of polyps smaller than 1 cm and the entire process turned out to be expensive and protracted.

Frank et al. [33] were the first to introduce 3-D CT-based endoscopy of a neobladder. Fifty-four patients undergone bilateral ureteroileal anastomosis were examined with an electron beam CT scanner. The visualization of the pouch, nipple, afferent ileal limb and ureters was feasible, whereas conventional cystoscopy could not reach these structures. Therefore, VC could be used in the evaluation of patients with bladder substitutions and unusual urinary tract symptoms [34].

5 Virtual urethroscopy

Yekeler et al. [35] recently reported an evaluation of urethral strictures with contrast-enhanced 3-D MR voiding urethrography. They carried out gadolinium-enhanced MRI of the bladder and urethra during voiding in both five healthy volunteers and 18 male patients with urethral disease. The authors evaluated the visualization of both normal anatomy and the presence of strictures along the prostatic, membranous, bulbous and penile segments of the male urethra. All the pathological findings detected in the virtual reconstructed images were identical to the ones revealed by conventional urethroscopy. Three-dimensional MR urethrography was superior in the depiction of membranous urethral strictures and in the imaging of strictures of the distal urethra that could not be documented by traditional retrograde urethrography. The technique proved to be excellent in visualizing normal urethral anatomy and promising in evaluating the entire male urethra. However, it should be emphasized that overestimation or underestimation of urethral stricture length may occur.

6 VE limitations

We should stress that the application of VE in the urinary tract presents certain limitations. There are difficulties in depicting small and flat lesions or mucosal thickening, and it cannot provide biopsy tissue specimens for histopathologic examination. In addition, the bladder must be sufficiently dilated and analysis of both the axial and virtual images acquired is usually essential for optimal evaluation. Unless MRI is used, CT-based VE bears the risk of radiation. Of note, VE images of the ureter cannot be reconstructed if renal insufficiency or high-grade tumor obstruction hinders contrast excretion into the upper urinary tract. However, in cases of endoluminal stenosis or obstruction, VE may permit virtual endoluminal navigation both cephalad and caudal to the stenotic segment. The major advantages and principal drawbacks of our survey of the published reports are summarized in Table 1.

7 Conclusion

To summarize, VE of the urinary tract is a promising diagnostic technique that could be applied repeatedly in the long-term follow-up of patients with ureteropelvic junction obstruction, urinary bladder tumors and ureteral and/or urethral strictures (Figures 1-3). VE provides a rapid analysis of axial raw data and a more perceptive evaluation of hollow organs, especially for clinicians who are less familiar with cross-sectional imaging than radiologists. Reconstruction of VE images may initially detect tumors and suspicious tissue regions and contribute to the selection of patients who should undergo a more thorough evaluation. The technique can also be used in patients who may be poor candidates for conventional endoscopy, such as those with severe urethral strictures or marked prostatic hypertrophy, or in patients with neobladder where cystoscopy is more complicated. The whole procedure of image processing and the production of VE images depends on the experience of the operator and is usually performed in less than 20 min by the majority of specialists. Nevertheless, comparative studies with larger groups of patients are deemed necessary with a view to further validating the diagnostic value and promoting the clinical utility of VE of the urinary tract.

References

1 Kagadis GC, Patrinou V, Kalogeropoulou CP, Karnabatidis D, Petsas T, Nikiforidis GC, et al. Virtual endoscopy in the diagnosis of an adult double tracheal bronchi case. Eur J Radiol 2001; 40: 50-3.

2 Robb RA. Virtual endoscopy: development and evaluation using the Visible Human datasets. Comput Med Imaging Graph 2000; 24: 133-51.

3 Assimos DG, Vining DJ. Virtual endoscopy. J Endourol 2001; 15: 47-51.

4 Hopper KD, Iyriboz AT, Wise SW, Neuman JD, Mauger DT, Kasales CJ. Mucosal detail at CT virtual reality: surface versus volume rendering. Radiology 2000; 214: 517-22.

5 Stenzl A, Frank R, Eder R, Recheis W, Knapp R, zur Nedden D, et al. 3-dimensional computerized tomography and virtual reality endoscopy of the reconstructed lower urinary tract. J Urol 1998; 159: 741-6.

6 De Nicola M, Salvolini L, Salvolini U. Virtual endoscopy of nasal cavity and paranasal sinuses. Eur J Radiol 1997; 24: 175-80.

7 Kimura F, Shen Y, Date S, Azemoto S, Mochizuki T. Thoracic aortic aneurysm and aortic dissection: new endoscopic mode for three dimensional CT display of the aorta. Radiology 1996; 198: 573-8.

8 Jolesz FA, Lorensen WE, Shinmoto H, Atsumi H, Nakajima S, Kavanaugh P, et al. Interactive virtual endoscopy. Am J Roentgenol 1997; 169: 1229-35.

9 Takebayashi S, Hosaka M, Takase K, Kubota N, Kishiba T, Matsubara S. Computerized tomography nephroscopic images of renal pelvic carcinoma. J Urol 1999; 162: 315-8.

10 Takebayashi S, Hosaka M, Kubota Y, Noguchi K, Fukuda M, Ishibashi Y, et al. Computerized tomographic ureteroscopy for diagnosing ureteral tumors. J Urol 2000; 163: 42-6.

11 Neri E, Boraschi P, Caramella D, Battolla L, Gigoni R, Armillotta N, et al. MR virtual endoscopy of the upper urinary tract. Am J Roentgenol 2000; 175: 1697-702.

12 Barbalias GA, Liatsikos EN, Kagadis GC, Karnabatidis D, Kalogeropoulou C, Nikiforidis GC, et al. Ureteropelvic junction obstruction: an innovative approach combining metallic stenting and virtual endoscopy. J Urol 2002; 168: 2383-6.

13 Siablis D, Kagadis GC, Liatsikos EN, Kalogeropoulou C, Petsas T, Karnabatidis D, et al. Ureteral metallic stent: application of virtual endoscopy for ureteral patency control. Int Urol Nephrol 2003; 35: 327-30.

14 Liatsikos EN, Siablis D, Kagadis GC, Karnabatidis D, Petsas T, Kalogeropoulou C, et al. Virtual endoscopy: navigation within pelvicaliceal system. J Endourol 2005; 19: 37-40.

15 Willoteaux S, Negawi Z, Lions C, Gaxotte V, Beregi JP. Observations from multidetector CT imaging of different types of renal artery stents. J Endovasc Ther 2004; 11: 560-9.

16 Schreyer AG, Fielding JR, Warfield SK, Lee JH, Loughlin KR, Dumanli H, et al. Virtual CT cystoscopy: color mapping of bladder wall thickness. Invest Radiol 2000; 35: 331-4.

17 Song JH, Francis IR, Platt JF, Cohan Rh, Mohsin J, Kielb SJ, et al. Bladder tumor detection at virtual cystoscopy. Radiology 2001; 218: 95-100.

18 Fenlon HM, Bell T, Ahari H, Hussain S. Virtual cystoscopy: early clinical experience. Radiology 1997; 205: 272-5.

19 Kim JK, Ahn JH, Park T, Ahn HJ, Kim CS, Cho KS. Virtual cystoscopy of the contrast material-filled bladder in patients with gross hematuria. Am J Roentgenol 2002; 179: 763-8.

20 Vining DJ, Zagoria RJ, Liu K, Stelts D. CT cystoscopy: an innovation in bladder imaging. AJR Am J Roentgenol 1996; 166: 409-10.

21 Hussain S, Loeffler JA, Babayan RK, Fenlon HM. Thin-section helical computed tomography of the bladder: initial clinical experience with virtual reality imaging. Urology 1997; 50: 685-9.

22 Narumi Y, Kumatani T, Sawai K, Kuriyama K, Kuroba C, Takahashi S, et al. The bladder and bladder tumors: imaging with three-dimensional display of helical CT data. Am J Roentgenol 1996; 167: 1134-5.

23 Merkle EM, Wunderlich A, Aschoff AJ, Rilinger N, Gorich J, Bachor R, et al. Virtual cystoscopy based on helical CT scan datasets: perspectives and limitations. Br J Radiol 1998; 71: 262-7.

24 Nambirajan T, Sohaib SA, Muller-Pollard C, Reznek R, Chinegwundoh FI. Virtual cystoscopy from computed tomography: a pilot study. BJU Int 2004; 94: 828-31.

25 Yazgan C, Fitoz S, Atasoy C, Turkolmez K, Yagci C, Akyar S. Virtual cystoscopy in the evaluation of bladder tumors. Clin Imaging 2004; 28: 138-42.

26 Kawai N, Mimura T, Nagata D, Tozawa K, Kohri K. Intravenous urography - virtual cystoscopy is a better preliminary examination than air virtual cystoscopy. BJU Int 2004; 94: 832-6.

27 Fielding JR, Hoyte LX, Okon SA, Schreyer A, Lee J, Zou KH, et al. Tumor detection by virtual cystoscopy with color mapping of bladder wall thickness. J Urol 2002; 167: 559-62.

28 Bernhardt TM, Rapp-Bernhardt U. Virtual cystoscopy of the bladder based on CT and MRI data. Abdom Imaging 2001; 26: 325-32.

29 Tsili AC, Tsampoulas C, Chatziparaskevas N, Silakos A, Kalef-Ezra J, Sofikitis N, et al. Computed tomographic virtual cystoscopy for the detection of urinary bladder neoplasms. Eur Urol 2004; 46: 579-85.

30 Homer JA, Davies-Payne DL, Peddinti BS. Randomized prospective comparison of non-contrast enhanced helical computed tomography and intravenous urography in the diagnosis of acute ureteric colic. Australas Radiol 2001; 45: 285-90.

31 Lämmle M, Beer A, Settles M, Hanning C, Schwaibold H, Drews C. Reliability of MR imaging-based virtual cystoscopy in the diagnosis of cancer of the urinary bladder. Am J Roentgenol 2002; 178: 1483-8.

32 Bernhardt TM, Schmidl H, Philipp C, Allhoff EP, Rapp-Bernhardt U. Diagnostic potential of virtual cystoscopy of the bladder: MRI vs CT. Preliminary report. Eur Radiol 2003; 13: 305-12.

33 Frank R, Stenzl A, Frede T, Eder R, Recheis W, Knapp R, et al. Three-dimensional computed tomography of the reconstructed lower urinary tract: technique and findings. Eur Radiol 1998; 8: 657-63.

34 Stenzl A, Kölle D, Eder R, Stöger A, Frank R, Bartsch G. Virtual reality of the lower urinary tract in women. Int Urogynecol J Pelvic Floor Dysfunct 1999; 10: 248-53.

35 Yekeler E, Suleyman E, Tunaci A, Tunaci M, Balci NC, Onem K, et al. Contrast-enhanced 3D MR voiding urethrography: preliminary results. Magn Reson Imaging 2004; 22: 1193-9.