University of Texas at Arlington
University of Texas at Dallas
Modeling of Magnetic Tools for Use with Superparamagnetic Particles for Magnetic Stone Extraction
Introduction: Complete stone removal is important in upper tract stone surgery. Unfortunately, even with the latest technological advances, current methods only achieve 50–80% complete clearance of upper tracts stones at the time of primary treatment. Our group has explored the novel use of peptide coated iron oxide superparamagnetic microparticles that bind to calcium stones allowing for extraction of these stones with magnetic tools. We have achieved binding of the particles to stone and pick up of the coated stones with magnets as a proof of concept. This study focused on the analytical and experimental work carried out to predict capture thresholds for feasible magnetic tool sizes and kidney stone magnetization, aimed at understanding the theoretical limits of this technology.
Methods: Magnetostatics equations were applied to a simplified, one-dimensional scenario of a spherical target coated with a variable amount of superparamagnetic particles, placed under the influence of a magnetic field aimed at vertical attraction of the target. Equations were parameterized in terms of (a) target size, ranging from 0.5 mm to 3 mm to represent stone sizes of interest, (b) effective emu per surface area delivered by the particle binding chemistry, and (c) distance to the field source. The field is generated by a fixed, axially-magnetized, high-energy rare earth barrel taking up the maximum diameter of the ureteroscope when backloaded and tethered through the instrument port (2.54 mm diameter x 12.7 mm long). Actual magnetometer data for coated stones was numerically fit and used to synthesize magnetization curves for varying surface-loading chemistries (emu/mm2). The estimated magnetic dipole was then multiplied by the spatial derivative of the magnetic field projected by the tool to predict the net force acting upon the stone.
Results:Figure 1 depicts a distance-to-capture performance envelope, consisting of a family of curves for various stone sizes tracing the predicted attractive range as a function of available magnetization. The range of emu per unit area resulting from microparticle loading is order-of-magnitude representative of current research standards and commercially available magnetophoretic particles. Experimental testing for known magnetic loading corroborates the low single-digit centimeter range for magnetic stone capture.
Conclusion: The use of paramagnetic particles in combination with magnetic tools to extract stones continues to be a promising area of research. Enhanced superparamagnetic micro-particle binding chemistry stands to extend this range, albeit along an asymptotic trendline.
Outstanding Paper Award
SchweinsbergerGino R.CilipChristopher M.1FriedNathaniel M.12
Department of Optical Science and Engineering, University of North Carolina at Charlotte, NC
Department of Urology, Johns Hopkins Medical Institutions, Baltimore, MD
Computer Simulations of Thermal Damage to the Human Vas Deferens During Noninvasive Laser Vasectomy
Introduction: Successful noninvasive laser thermal coagulation of the canine vas deferens, in vivo, has been reported previously both with and without the use of optical clearing agents (to make the scrotal skin more transparent). However, there are significant differences between the tissue properties of canine and human skin. Since the matrix of laser and cooling parameters is too large to completely explore in experiments, it is important to use computer simulations to predict the optimal set of treatment parameters. In this study, a standard three-stage (optical, thermal, and tissue damage) computer model was used to explore the feasibility of noninvasive laser vasectomy in humans.
Methods: Part 1 (optical) of the model consists of a Monte Carlo model of light transport in tissue to determine the spatial distribution of absorbed photons in the major tissue layers (epidermis, dermis, and vas). The scrotal fold anatomy in the vas ring clamp makes it necessary to include skin layers on the top and bottom of the model. An index of refraction, absorption and scattering coefficient, and anisotropy factor (direction of scattering) was assigned to define the optical properties of each tissue layer. Part 2 (thermal) of the model consisted of using the MC results as the spatial heat source for the heat transfer model. Each tissue layer was assigned a value for the density, specific heat, and thermal conductivity. Part 3 (tissue damage) of the model used the temperature-time data from Part 2 as the input in a standard Arrhenius integral model for irreversible thermal tissue denaturation. In this model, tissue damage is treated as a first order rate process (based on chemical reaction kinetics) that is exponentially dependent on temperature and linearly dependent on time. Tissue damage is defined by a single parameter (Ω), where a value of Ω >1 results in irreversible tissue damage. Multiple treatment parameters were studied, including laser output power (5–9 W), cryogen pulse duration (60–100 ms), cryogen cooling rate (0.5–1.0 Hz), and increase in optical transmission (0–50 %) through skin due to application of optical clearing agents. The parameters were chosen based on those used successfully in a canine model.
Results: After application of an optical clearing agent to increase skin transmission by 50%, an average laser power of 6 W, cryogen pulse duration of 60 ms, and cryogen cooling rate of 1 Hz, resulted in vas wall temperatures of approximately 58 oC, sufficient for thermal coagulation (Ω >1), while 1 mm of the skin surface remained at a safe temperature of approximately 45 oC with Ω <1 (Figure 1). A critical temperature above 52 oC would result in undesirable thermal coagulation of skin.
Map of accumulated thermal damage after 60 sec of laser irradiation and 6 W of average laser power. The vas wall is thermally coagulated (Ω >1) while the skin remains undamaged (Ω <1).
Conclusion: A computer model including Monte Carlo, heat transfer, and thermal damage simulations, indicates that it is possible to noninvasively thermally coagulate the human vas deferens without adverse side-effects (e.g., scrotal skin burns) if an optical clearing agent is applied to the skin prior to the procedure.
WangHuiKangWeiZhuHuiMacLennanGregoryRollinsAndrewCase Western Reserve University, Cleveland, OH
Three-Dimensional Imaging of Ureter with Endoscopic Optical Coherence Tomography
Introduction and Objectives: To verify the ability of identifying the layered structures of ureteral wall and imaging a segment of ureter in three dimensions with a high speed endoscopic optical coherence tomography (EOCT).
Methods: We imaged a porcine ureter ex vivo using a spectral domain EOCT with an optimally designed circumferential scanning fiber catheter. The images were correlated with the histological images to identify corresponding structures. Three-dimensional images and enface images at different radical depths were reconstructed from the multiple cross-sectional images to show the layered structure of a segment of the ureter from different views.
Results: EOCT images can clearly reveal all layers of the ureteral wall, as shown in the histological images. Especially with the optimally designed fiber catheter, the light beam was well centered during the rotation and pull back, which allowed constant acquisition of high fidelity images and unambiguous identification of the smooth muscle layers in all images. With significantly improved imaging speed, a segment of ureter (20 mm) can be imaged in a short period of time.
Conclusions: With its capability to image all layers of the ureteral wall, EOCT offers the potential to stage urothelial cancers that have infiltrated the muscular wall (stage T2). This information will be complimentary to the diagnostic information obtained through ureteroscopic biopsy and CT urogram.
LiuJen-Jane13PanYing13VolkmerJens-Peter2MachKatherine E.13WeissmanIrving2JensenKristin3LiaoJoseph C.12
Department of Urology, Stanford University
Institute for Stem Cell Biology and Regenerative Medicine, Stanford University
Veterans Affairs Palo Alto Health Care System, Palo Alto, USA
Targeted Imaging of Bladder Cancer with Cancer-Specific Molecular Contrast Agents
Introduction: Current diagnosis of bladder cancer is by white light cystoscopy, which has suboptimal specificity for differentiating non-papillary cancer from inflammation. Probe-based confocal laser endomicroscopy (CLE) provides dynamic in vivo imaging of the endoluminal tracts with micron-scale resolution. Real-time analysis of confocal images remains challenging with non-specific contrast agents such as intravesical fluorescein. Tumor imaging specificity may be enhanced by coupling CLE with fluorescently labeled tumor-specific antibodies or peptides. We report our preliminarily efforts of ex vivo bladder tumor imaging by CLE using an antibody against a known human tumor-specific surface biomarker and a peptide that binds human epidermal growth factor receptor (EGFR) as contrast. Studies have shown that both the antibody and peptide bind surface markers abundant in cancer but absent on the surface of normal urothelium.
Methods: The tumor-specific antibody or the EGFR-binding 12-amino acid peptide was labeled with fluorescein isothiocyante (FITC) and instilled intravesically into a fresh cystectomy specimen (n=4 for antibody; n=1 for peptide). For a negative control, tumor-specific antibody was instilled into the renal pelvis of a radical nephrectomy specimen removed for RCC. After 15–30’ incubation at 37°C, the bladder was drained, opened, and imaged with a CLE system (Mauna Kea Technologies, Paris, France). Imaging was performed of normal-appearing, tumor, and suspicious areas, followed by excision of imaged tissues for histopathology with H&E. Tumor-specific staining with the antibody was further confirmed with immunofluorescence.
Results: All bladders had high-grade urothelial carcinoma on pathology. The tumor-specific antibody consistently showed greater fluorescent staining with CLE in areas of tumor, compared to normal urothelium in all 4 bladders (Figure 1). An erythematous area was not stained by the antibody and was confirmed to be inflammation by pathology. Immunofluorescence showed tumor-specific antibody on the most superficial layer of tumor but not on normal urothelium. Similarly, preferential staining of tumor compared to normal urothelium was observed when using the EGFR-binding peptide as a contrast agent. Imaging with tumor-specific antibody on normal urothelium from the renal pelvis of a nephrectomy specimen also did not show any fluorescent signal on CLE.
CLE of normal and tumor areas after administration of tumor-specific antibody. Corresponding H&E confirmed pathology.
Conclusion: Both the known tumor-specific antibody and EGFR-specific peptide show specificity for bladder cancer by CLE, ex vivo. Our study shows the feasibility of optical imaging of bladder cancer with CLE and tumor-specific contrast agents and raise the possibility of using intravesical molecular contrast agents for in vivo endoscopic targeted imaging of bladder cancer.
LiuJen-Jane1,2HsiaoShelly T.2PanYing1MachKatherine E.12McMillanAlex3JensenKristin2LiaoJoseph C.12
Department of Urology, Stanford University, USA
Veterans Affairs Palo Alto Health Care System-Palo Alto, USA
Department of Health Research and Policy, Stanford University, USA
Real Time Diagnosis of Bladder Cancer with Probe-Based Confocal Laser Endomicroscopy: A Prospective Diagnostic Accuracy Study
Introduction: Confocal laser endomicroscopy (CLE) is an emerging technology for in vivo optical biopsy of the urinary tract that enables micron scale resolution reminiscent of histology using probes that fit in standard cystoscopes/resectoscopes. White light cystoscopy (WLC) has well-recognized shortcomings in bladder cancer diagnosis, particularly in differentiating nonpapillary urothelial carcinoma from inflammation. Accuracy for cancer is reported to be 70–80% for WLC. We previously established suggested criteria to differentiate normal, benign, and neoplastic urothelium with CLE. We report the results of our ongoing prospective diagnostic accuracy study of CLE for bladder cancer diagnosis.
Methods: Patients scheduled to undergo transurethral resection of bladder tumor were recruited. Patients first went WLC, followed by CLE with intravesical fluorescein, tumor resection/biopsy, and histologic confirmation. Suspicious (targeted) areas were marked with electrocautery, imaged with CLE, and resected/biopsied. Normal-appearing areas also were imaged and biopsied as controls. Diagnostic accuracy is determined during cystoscopy by the surgeon and offline in a blinded fashion after image processing. Using histology as the standard, the diagnostic accuracy of WLC and WLC+CLE was calculated.
Results: To date, 35 patients and 118 areas were able to be imaged with CLE; 58 of these areas were deemed suspicious by WLC, and 62 classified as “normal.” Of the 58 suspicious lesions, 41 were confirmed to be urothelial carcinoma, and the remainder were benign or normal. Ninety-eight percent of normal-appearing lesions were confirmed to be normal on histopathology, and one area had CIS. For targeted lesions, accuracy for cancer diagnosis by WLC+CLE was 88%, compared to 84% for WLC alone. For normal-appearing lesions, accuracy for cancer diagnosis for both WLC alone and WLC+CLE was 98%. Representative CLE images of normal, benign, and cancer areas are shown in Figure 1.
CLE probe and representative CLE images of normal, benign/ inflammatory, low grade, and high grade lesions
Conclusion: CLE is a promising adjunct to WLC in the diagnosis of bladder cancer. Preliminary results suggest that the addition of CLE to WLC may improve accuracy of bladder cancer diagnosis compared to WLC alone, although additional data is required to show whether CLE offers a clinically significant benefit.
BurkeD.12ZhouX.12RottyV.2KonetyB.1SweetR.12
Urologic Surgery, Medical School, University of Minnesota, Minneapolis, MN
Center for Research in Education and Simulation Technologies (CREST), University of Minnesota, Minneapolis, MN
Web-Accessible 3D Anatomy Software of Urologic Pathophysiological Conditions and Procedures for Patient Education
Introduction: The poor transfer of knowledge to patients with standard informed consent is well documented. Providing easy-to-use, web-accessible tools for clinicians and patients should enhance informed consent. While passive 2-D animation models exist on the Internet, it has been shown that interactivity and the use of 3-D virtual models enhances understanding of anatomic structures and relationships. Our goal was to provide a detailed anatomy guide, tailored to displaying specific ailments and common procedures in urology with an easy-to-use interactive 3D web format.
Methods: Using current WebGL software, we access and display our library of 3D human anatomical structures. Models are derived from real human MRI/CT, and built under the guidance of consulting physicians. Users move and rotate models, remove layers and isolate areas of interest in finer detail, and even replace anatomy with available, diseased versions using a simple point-and-click interface. There is also the option to view animations of actual procedures for specific ailments. We utilize several drop-down commands and anatomy lists.
Results: A screen shot of the web-based 3-D model is shown in Figure 1. A high-quality model is displayed over the Internet in realtime. The user can strip layers off the model from skin to bone with an infinite number of viewpoints. Disease specific applications have been developed for BPH and prostate cancer. The user is able to continue interacting with the models during the procedures. This system has been deployed in the Urologic Clinic at the University of Minnesota for pilot testing.
3D gross anatomy of male pelvis, available for individual dissection.
Resection step of BPH treatment animation, with additional information and internal view.
Conclusion: The web-based anatomy viewer can be used online in real-time and warrants study as to the ability to enhance informed consent about urologic diseases and procedures to patients.
CilipChristopher M.1PierorazioPhillip M.2RossAshley E.2AllafMohamad E.2FriedNathaniel M.1
Department of Optical Science and Engineering, University of North Carolina at Charlotte, NC
Department of Urology, Johns Hopkins Medical Institutions, Baltimore, MD
High Frequency Ultrasound Imaging During Noninvasive Laser Coagulation of the Canine Vas Deferens, in vivo
Introduction: Approximately 500,000 vasectomies are performed annually in the United States, making it the most common urological procedure in the U.S. Although more effective and less likely to have complications than tubal ligation, the number of men undergoing surgical sterilization is approximately three-fold less than women. A safer, less invasive approach to vasectomy may eliminate male fears associated with surgery and reverse these trends. Our laboratory has recently demonstrated successful noninvasive laser coagulation and thermal occlusion of the canine vas deferens, ex vivo and in vivo, with the goal of developing a completely noninvasive vasectomy procedure. During conventional surgical vasectomy, occlusion of the vas is confirmed visually. However, during noninvasive laser vasectomy, a noninvasive diagnostic method may be helpful to confirm successful targeting and closure of the vas to insure consistent and reproducible results. High frequency ultrasound (HFUS) imaging has been shown previously to quantify the anatomical dimensions of the human vas accurately, in vivo. The objective of this study was to use HFUS to confirm successful targeting and thermal coagulation of the canine vas during noninvasive laser vasectomy.
Methods: Bilateral noninvasive laser coagulation of the vas deferens was performed in a total of 6 dogs using an Ytterbium fiber laser with a wavelength of 1075 nm, output power of 9.0 W, 500 ms pulse duration, pulse rate of 0.5 Hz, and 3 mm diameter spot. A cryogen spray cooling device was used to cool the treatment area during the procedure and prevent the formation of scrotal skin burns. A standard clinical US system with 13.2 MHz transducer was used to image the canine vas before and after the procedure. Simultaneous application of Doppler US at a frequency of 6.15 MHz helped to distinguish between the vas and the spermatic cord. Burst pressure measurements were recorded for the excised thermally coagulated vas samples to quantify the degree of closure.
Results: Acute burst pressures averaged 291±31 mmHg, significantly greater than previously reported normal ejaculation pressures of 136±29 mmHg. Representative HFUS images of the canine vas deferens both before and after the procedure are shown in Figure 1, demonstrating the ability to distinguish native from thermally coagulated vas. Doppler US indicated normal blood flow through the testicular artery and no detectable collateral damage to proximal structures.
Representative ultrasound images showing (A) native vas before the procedure and (B) thermal lesion in vas after the procedure.
Conclusion: High frequency ultrasound can be used as a diagnostic method to confirm successful targeting and thermal coagulation of the canine vas during noninvasive laser vasectomy.
HungAndrew J.UkimuraOsamuHoHenryChengChristopherGillInderbir S.DesaiMihir M.USC Institute of Urology, Keck School of Medicine, University of Southern California and Singapore General Hospital, Singapore
Novel Robot for Trans-Perineal Prostate Needle Intervention: Phantom Study
Purpose: To evaluate in phantom prostates the accuracy of a novel robotic ultrasound-guided intervention device, which may be utilized for active surveillance and focal therapy of prostate cancer in the near future.
Materials and Methods: The BioXbot (Biobot Surgical, Singapore) is an automated robotic intervention delivery device that utilizes 2D ultrasound imaging to create 3D planning models for needle delivery. Once the targets are designated in the 3D model, the robotic positioning system serially positions the delivery arm for each target. Needle is delivered transperineally with conformation to a dual-cone template in which a maximum of two skin puncture sites are required to cover the entire prostate. We first targeted the center of pre-made hypoechoic lesions randomly distributed within three commercially available prostate phantoms (Model 053MM, CIRS, Norfolk, Virginia). As we retracted the needle from each target, we injected colored dye to mark the needle tracts. After targeting and delivering the needles to the center of the hypoechoic lesions once, we re-positioned the BioXBot device and re-captured the images of the phantom prostate to create a new planning model. Re-targeting of the hypoechoic lesion centers was then performed a second time, and the needle tracts were inked as described. Additionally, we also targeted and delivered needles to additional areas of the phantom prostate away from the pre-made hypoechoic lesions following a sextet template. The first and re-targeted sextet needle tracks were all inked. Finally, we acquired 1-mm step MR imaging of the phantom prostates and utilized the images to measure the spatial distance between tips of the needle tracts and their intended targets.
Results: Each hypoechoic lesion was targeted twice (six needle tracts per phantom x three phantoms=18 needle tracts) with the average distance from needle tip to lesion center being 3.81±1.25 mm. Of note, the needle tip was successfully placed within all nine pre-made lesions in both rounds. The tips of the sextet-based needle tracts (18 additional needle tracts) were each re-targeted by ultrasound imaging with the average distance between the first and second needle tips 4.68±1.18 mm (targeting hypoechoic lesions versus sextet biopsy tracts, p=0.06. Some difficulty was encountered in visualizing the tips of the first sextet needle tracts by ultrasound and likely contributed to an increased error in accuracy.
Conclusions: The BioXBot allows automated delivery of intervention needles, robotically targeting and re-targeting lesions within the prostate currently with reasonable accuracy. Although presently limited by its lack of real-time imaging during targeting to confirm actual needle location (versus planned target), this device has a promising future for the management of low-risk prostate cancer by active surveillance or focal therapy.
Dept of Urology, University of Texas Southwestern Medical Ctr., Dallas, TX
Digital Light Innovations, Austin, TX
A Novel Laparoscopic Camera for Characterization of Renal Ischemia Using DLP® Hyperspectral Imaging: Initial Experience in a Porcine Model
Introduction: Digital light processing (DLP®) hyperspectral imaging is a non-invasive means of visualizing the chemical composition of in vivo tissues using reflectance spectroscopy. It is able to provide a real-time map of surface tissue oxygenation. Using a CoolSNAP HQ2™ CCD camera (Photometrics, Tucson, AZ), we have previously used this system to characterize renal ischemia during open partial nephrectomy in pigs and humans. By incorporating a light guide, 00 laparoscope, and a DVC 1200M CCD camera (DVC, Austin, TX), DLP® hyperspectral imaging was adapted for use during laparoscopic surgery.
Methods: Two adult female pigs underwent DLP hyperspectral imaging using the laparoscopic system, before, during, and after clamping the renal hilum for 90 minutes. Imaging during ischemia was obtained at several intervals. The relative percentage of oxygenated hemoglobin (relative %HbO2) was determined at each imaging interval using previously described methodology. Data were compared to those for 3 pigs imaged over a similar interval using the open HQ2™ camera.
Results: For Lap vs. Open, relative %HbO2, was 68.7 vs. 72.9% pre-clamp respectively, dropping by an average of 27% vs. 26% within 10 minutes after clamping, and rapidly rising to baseline after clamp removal, with similar trends for the two imaging systems.
Laparoscopic HSI camera.
%HbO2versus time.
Conclusion: The laparoscopic DLP hyperspectral imaging system performs similarly to the open system and has excellent spectral imaging capabilities.
KoWoo Jin1KangHyun Wook2RajabhandharaksDanop2RutmanMatthew3CharlesOsterberg E.4ChoiBenjamin B.4
Department of Urology, National Health Insurance Corporation Ilsan Hospital, Goyang, South Korea
American Medical Systems, San Jose, CA, USA
Department of Urology, Columbia University Medical Center, New York, NY, USA
Department of Urology, New York Presbyterian Hospital-Weill Cornell Medical Center, New York, NY, USA
Defining Optimal Laser Fiber Sweeping Angle for Effective Tissue Vaporization Using a 180 W, 532 nm Lithium Triborate Laser
Background: A side-to-side, rotational laser fiber sweeping motion during photoselective vaporization of the prostate (PVP) is considered standard technique contrary to axial (in and out) motions seen in traditional TURP. Scientific validation of the optimal laser fiber sweeping techniques currently is lacking.
Objective: To identify the most efficient tissue vaporization sweeping angle (SA) during PVP.
Design, Setting and Participants: Experiments were conducted with the GreenLight XPSTM laser system with MoxyTM fiber technology (American Medical Systems, Inc.) at 120W and 180W. Ten 2.5×2.5×1.5 cm blocks of porcine kidney were used for each SA (n=140). Specimens were placed in a chamber equipped with computer-assisted sweeping motion at variable SAs (0, 15, 30, 45, 60, 90, and 120°).
Measurements: Vaporization efficiency was assessed by the amount of tissue removed per time. The coagulation zone (CZ) thickness also was measured. Statistical analysis was performed with the two-sample Student's t-test (p<0.05).
Results and Limitations: Maximal vaporization rate (VR) was achieved at SA 15 and 30°. Irrespective of power, VR increased and CZ decreased linearly with decreasing SA from 120° to 30°. Across all SA, 180W achieved at least a 60% greater VR versus 120W. Specifically, the percentage difference of VR was statistically significant for 15° and 30° when compared to 60° for both 120W and 180W. The CZ was the thinnest at SA 30°. This study used an in vitro model that may restrict its clinical translation.
Conclusions: Optimal vaporization, regardless of the laser power, occurred at a SA of 15° and 30° with the lowest CZ at 30°. Contrary to anecdotal recommendations for a wider SA, a narrower SA (30°) achieved the maximal tissue vaporization efficiency. Collectively, these data highlight the capability to test PVP technique in a scientific manner, identifying the optimal surgical parameters to aid practicing urologists to achieve desired clinical outcomes.
