Research

My PhD research primarily involved the development of a confocal microendoscope for optical biopsies in surgery. My specific contributions to this project were:

  1. developing the surgical imaging catheters
  2. developing the contrast agent delivery system
  3. developing the focus / depth scan mechanism
  4. making the overall instrument smaller for operating room use
  5. writing the instrument software interface
  6. developing a higher resolution and larger NA objective lens targeted for use in esophageal cancer

We are currently in human trials using the device to image ovarian cancer in high risk women.

I am still adding content to this page. Below are figures illustrating the device. More information will be added soon.

The Confocal Microendoscope

System in use during surgery

System in use during surgery

The surgical microendoscope system imaging the epithelial surface of an ovary during surgery. In this image the surgeon has located the left ovary using a wide field endoscope (left most display) and is inspecting the epithelial cells using the microendoscope via the second display from the left. The mobile instrument is on the far right.

In vivo confocal microlaparoscope video of the human ovary using topically applied fluorescein sodium as the contrast agent. The pathology report diagnosis was serous cystadenoma. (Circular field of view is 450μm.)

Need for real-time optical biopsies in surgery

Successful treatment of cancer is highly dependent on the stage at which diagnosis occurs. Early diagnosis, when the disease is still localized at its origin, results in very high cure rates—even for cancers that typical have poor prognosis. Unfortunately, many cancers are not found until later stages due to inadequate diagnostic techniques. Approximately 90% of cancers arise from the epithelial cells that cover organs [1]. Development of surgical devices that can better interrogate epithelial surfaces for abnormalities would enable earlier detection of cancer and significant gains in overall patient survival.

Diagnosis of cancer is often done by pathologists using thin sections of stained and processed biopsy tissue. Bright field microscopy has played a central role in the diagnosis of disease from carefully prepared biopsy slides. Confocal microscopy, a more recent innovation, is also being used with greater frequency because it can directly image bulk sections of tissue with high clarity. Bright field images of bulk tissue appear very blurry due to simultaneous collection of out of focus planes. However, confocal images of bulk tissue are sharp because the microscope only collects light from in focus planes; light from out of focus planes is rejected.

Since the confocal microscope alleviates the need for cutting tissue into thin sections, it has significant potential as an in-vivo imaging device that could supplant biopsies. However, a standard confocal microscope is a large device that is not especially suited for accessing the epithelial surface of most organs where biopsies are acquired. Realizing the potential of confocal imaging to ultimately supplant biopsies via in-vivo imaging, we worked on the initial technologies to enable in-vivo confocal imaging via coherent fiber optic bundles [2]. Since the initial work, our research group has continued developing technologies to allow live in-vivo human cellular imaging via confocal microendoscopy during surgery.

Case study: patients at high risk of ovarian cancer

The American Cancer Society estimates that over 230,000 new cases of ovarian cancer were diagnosed in 2007 while more than 141,000 women died from the disease worldwide. [3] For those diagnosed with the disease, only 30 to 45 percent will survive five years. [4] [5] The survival rate for ovarian cancer is poor because the disease frequently has no obvious signs until it has reached an advanced stage. Only 19 percent of all cases are diagnosed during the early localized stages of the disease; most of the population are diagnosed during the later stages when the treatment is expensive and generally unsuccessful.

For the female population at large, most are at a low risk of acquiring ovarian cancer. It is estimated that approximately 1 in 58 newborn females (1.7 percent lifetime risk) will develop ovarian cancer in their lifetime. [6] However, the lifetime risk for the disease increases by two fold if the individual has one first or second-degree relative with the disease. With two first-degree relatives the risk increases by 25x (a 40 percent lifetime risk). [7] [8] [9] [10]

For the subgroups of women at increased risk, few options exist to allow early detection of the disease. The delicate epithelial surface of the ovary is not amenable for biopsy. The NIH 1994 consensus stated that there is no single acceptable screening test for ovarian cancer and no evidence that combining the available screening tests—CA125, transvaginal ultrasound, and pelvic exam—has an acceptable sensitivity and specificity. [11] Ovarian cancer is thought to metastasize early in the course of the disease and many experts believe that a different biology occurs during Stage I cancer as compared to the later stages. In Stage I, it is thought that the cancer often metastasizes before a lesion in the ovary becomes grossly visible. Without a viable method for early detection of the disease, most women at high risk are provided the option of prophylactic oophorectomy. Preemptive oophorectomies have the negative side effects of sterility and loss of natural hormone production. Hormone replacement is typically not an option for this subgroup since they are also at an increased risk for breast cancer and hormone replacement can further increase this risk.

In the following sections we discuss a confocal microendoscope system that is generally applicable to in-vivo imaging of epithelial cells. The technology is applied to the subgroup of women at high risk for ovarian cancer who have no suitable methods to detect cancer in its early stages. Our mobile device provides the surgeon with real-time in-vivo confocal images. The long term objective is to allow the physician to determine if the ovary is healthy and circumvent unnecessary prophylactic oophorectomy.

Surgical System

Surgical cart

Mobile surgical system.

Our surgical confocal microendoscope system consists of an imaging catheter connected to a mobile cart. The mobile system can be easily moved into an operating room when live cellular imaging is needed. All the system components are housed on a mobile endoscopy cart. The above figure depicts the system’s components. The top shelf houses the sterile imaging catheter and other necessary surgical supplies. The second shelf contains the optical scanning system—discussed in the following paragraphs—which directly connects to the imaging catheter. The third and fourth shelves house the laser, function generator, and computer. The bottom shelf houses the medical grade isolation transformer and laser safety goggles. The operator display and the controls for collecting data during surgery are located on a platform at standing height.

Optical components

Optical components

Diagram of the optical scanning system. The left side illustrates the standard mode of operation. A 488 nm laser is anamorphically shaped into a line and scanned onto the coherent fiber bundle in the imaging catheter. The excited signal re-enters the system, is descanned and filtered through a confocal slit. Then the light is rescanned onto a two dimensional detector. In the spectral imaging mode, the image scan mirror is turned to its extreme position redirecting the light through a dispersing prism.

At the core of the confocal microendoscope system is the optical scanning system. The components and layout are shown above. The left side of the figure shows how the system operates in the standard live imaging mode. In this mode a 488 nm solid state laser beam is expanded and anamorphically shaped into a line via a cylindrical lens. This laser line is then reflected into the image path by a dichroic filter and scanned across the coherent fiber bundle face at the proximal end of the imaging catheter’s connector.

Tissue fluorescence is collected by the imaging catheter, collimated back into the optical scanning system, and de-scanned using the object scan mirror. The dichroic filter passes the fluorescence signal, which is focused down to a stationary line. A slit serves as the confocal aperture. The light exiting the slit is then re-collimated and rescanned using the image scan mirror. A final filter removes residual excitation. Finally the beam is refocused back into a line that sweeps across the camera to collect a two-dimensional image every 1∕30th of a second. In addition to live two-dimensional imaging, the system can also collect multi-spectral data. [12] [13] This multi-spectral mode is instantly achieved via a software button that deflects the object scan mirror to its extreme position. In this configuration, the light passes through a prism and the dispersed signal is collected by the camera. Since one spatial dimension on the camera is used for spectral collection, the second image spatial dimension is collected over time. The complete spectral data collection procedure executes in a few seconds. Once spectral collection is complete the system reverts back to its grayscale operating mode.

Surgical software

Surgical software

The mobile system has been designed to streamline all operations during surgery. Once the system is plugged in and the safety interlocks engaged, the system boots and all hardware is initialized. Hardware initialization includes the solid state laser, camera, dye delivery system, and function generator for scan mirror control. After the automatic initialization, the operator is presented with the software control system auto-initialized for live imaging.

The above figure shows the software control system. It provides a simple interface for viewing and collecting live images during the surgical procedure. The software has controls to: (1) start live acquisition, (2) save the current frame, (3) record video, (4) delivery dye, (5) load dye, and (6) adjust histogram optimization. In addition to the basic controls, the system also records procedure and patient information, which is archived with the images. Basic diagnostic information such as image dynamic range and frame rate are also visible. During operation the surgeon can easily see real-time imagery in the main window along with data acquisition and dye delivery status.

Imaging Catheters

Endoscopes

Endoscopes

The confocal imaging catheters. A laparoscopic version is shown on top with an integrated dye delivery system that uses a piezo valve and pressurized syringe. The lower left of the figure shows the flexible endoscope version of the device. Both devices have the same handle that uses a depth/focus knob to translate the coherent fiber bundle routed through the center of the device (handle detail shown to the right).

The imaging catheter comes in two varieties, a rigid laparoscope 5 mm in diameter and a flexible endoscope 4.2 mm in diameter; both provide the same functionality. This compact instrument contains the micro-objective, coherent fiber optic bundle, dye delivery system, and depth/focus mechanism. In use, the imaging catheter feels like a wide field endoscope except that imaging is done with the probe in contact with the tissue.

Figure 4 depicts the two varieties of the imaging catheter. A twenty foot flexible housing connects to the optical scanning system and protects the coherent fiber optic bundle and electrical connections. The distal tip of the instrument contains a micro-objective [14] lens that images tissue onto the coherent fiber bundle. The micro-objective connects to a housing that contains the fiber bundle. Running parallel to the housing is a 21.5 gauge channel that delivers controlled volumes of fluorescent dyes onto the tissue in the imaging catheter’s field of view. The outer surface of the lens and the fiber housing (excluding the face of the lens) are sealed inside a medical grade teflon sheath. The imaging catheter is designed for reuse; it can be quickly disconnected from the optical scanning system and sterilized using Ethylene Oxide.

We have investigated a variety of depth/focus mechanisms [15] [16] [17] and now present a new method that has proven to be more reliable and easier to use during surgery. Rotation of the depth/focus knob on the handle causes the fiber optic bundle to translate along the optical axis. Translation of the fiber via the depth/focus knob allows the surgeon to select the desired imaging plane in the tissue. Since the system is designed for contact imaging on the epithelial surface of organs, once optimal focus has been obtained refocus is not necessary. The surgeon can simply move across the tissue and change sites while maintaining focus since contacting the tissue will bring the epithelial cells into focus. The surgeon can use the depth/focus knob to image planes below the surface for further interrogation of abnormalities.

Dye delivery

Dye delivery

Demonstration of the imaging catheter's ability to locally deliver small volumes of dye to the field of view. In this sequence of images (at 1/30th second intervals) the operator presses the dye delivery button and the dye is delivered to the distal tip via actuation of the piezo valve near the imaging catheters's handle. The delivered volume in this example is approximately 1 uL.

We have developed a localized dye delivery system that minimizes patient exposure to fluorescent contrast agents and provides a method to mark imaged tissue sites. At the tip of the imaging catheter, a 21.5 gauge channel conforms around the face of the micro-objective to deliver dye directly to the field of view. The system is capable of delivering very small dye quantities on demand, down to 0.1 μL. Once the operator pushes the dye delivery button, a piezo valve located behind the imaging catheter handle opens for a few milliseconds. A pressurized syringe behind the valve containing a fluorescent contrast agent supplies dye through the dye channel to the field of view. The above figure shows the dye delivery process in 1∕30th second intervals; full delivery occurs in approximately 1∕6th of a second.

Human results

The confocal microendoscope system is currently being evaluated in clinical trials to image the epithelial surface of the ovary at the University Medical Center in Tucson, Arizona. The device was granted “non significant risk” status by the University of Arizona’s Institutional Review Board and has been approved for use in humans using a protocol that includes topical application of sodium fluorescein as the contrast agent. [18]

Ovarian tissue with Acridine Orange

Images of human ovary epithelium obtained ex-vivo using acridine orange. Sub-captions contain pathology diagnosis.

Ex vivo video of normal appearing ovary epithelium tissue stained with acridine orange. (Circular field of view is 450 μm.)

Ex vivo video of abnormal appearing ovary epithelium tissue stained with acridine orange. The pathology report diagnosis was serous cystadenoma. (Circular field of view is 450 μm.)

The images with acridine orange, shown above, demonstrate the excellent diagnostic ability of the instrument. The epithelial surface of a healthy ovary is characterized by a homogeneous distribution of bright nuclei seen in (a). The epithelial surface of the ovary is delicate and partial denuding can occur, exposing the underlying stroma (b). Below the epithelial surface, healthy stroma also exhibits a characteristically homogenous structure albeit with a different nuclear size distribution and shape (c).

In the case of carcinoma, the tissue structure is visibly differentiable (d) from healthy epithelial cells (a). The epithelial surface is irregular and the high degree of heterogeneity indicative of ovarian cancer.

We have perviously shown [19] that the confocal microendoscope system can easily differentiate normal epithelium from ovarian cancer—providing a diagnostic advantage when the neoplasia is small and not visible at the gross anatomic level. It also appears that the confocal microendoscope system may also be able to visualize cellular changes that happen prior to the onset of cancer. Less distinct tissue changes such as tissue sclerosis and endometriosis may also be detectable

Ovarian tissue with Fluorescein

Images of ovary epithelium obtained in-vivo during clinical trials using fluorescein. The suboptimal cellular detail is a result of the minimal preferential binding exhibited by fluorescein.

For initial in-vivo clinical studies fluorescein was selected because of its pre-existing approval for human use. The diagnostic utility of fluorescein when used as a topical contrast agent on the ovary is not very good as it provides little contrast and lacks preferential binding to cellular level structures of interest. However, it has served its purpose as a safe, pre-approved contrast agent allowing us to test the safety and feasibility of the confocal microendoscope system.

The above figure shows three examples of in-vivo images obtained with the confocal microendoscope system. The images demonstrate that the device functions as designed. The imaging catheter can deliver controlled volumes of dye to the image site and then display real-time image data to the surgeon. The focus mechanism works well; after an initial adjustment of the focus, the instrument can be moved to various sites on the ovary while maintaining good focus on the epithelial surface.

Esophagus tissue with Acridine Orange

Images of human esophagus epithelium obtained ex-vivo using acridine orange.

As previously discussed, any lumen or organ that is endoscopically or laparoscopically accessible is a suitable candidate for microendoscope imaging of the epithelium. Applications of the device to detect the transformation of normal esophagus to adenocarcinoma are also being investigated. Barrett’s Esophagus is the premalignant lesion for adenocarcinoma of the esophagus. Due to a severe chronic gastroesophageal reflux disease, these patient experience a transformation of the normal squamous epithelium (a) into tissue that closely resembles the intestine with columnar appearing mucosa and intestinal metaplasia (b). Once this transformation has taken place, the individual is at higher risk for adenocarcinoma of the esophagus. [20] Moreover, the 5-year survival rate for this cancer is only between 10-15%. [21] The American Cancer Society estimates that approximately 529,000 new cases of adenocarcinoma of the esophagus were diagnosed in 2007 while more than 442,000 people died worldwide from the disease. [22] Yet again, one of the predominant factors causing low survival rates is the late detection of this disease. Patients with Barrett’s Esophagus identified with an increased risk for adenocarcinoma typically undergo endoscopic surveillance and biopsy every one or two years. Again, the confocal microendoscope system’s ability to resolve cellular detail indicates that it would be a useful tool to improve the detection of dysplasia and adenocarcinoma in-vivo. In the above figure (c) illustrates the distinct difference when esophagus tissue has made the transformation to tumorous tissue.

Modeling System Performance

To better understand the axial and lateral performance of the surgical confocal system, a Monte Carlo model was developed to study the effects of tissue scattering. Confocal microscopes use a small aperture to reject out of focus light, allowing imaging of thin sections within thick samples. Standard confocal microscopes employ a single pinhole aperture that must be spatially scanned to collect a two or three dimensional image. To reduce im- age acquisition times, parallelized confocal systems use an array of pinholes or a slit aperture to simultaneously collect multiple image points, reducing acquisition times in proportion to the number of simultaneous detection points used. The drawback of parallelized systems is cross-talk between the individual apertures. In highly scattering media, such as tissue, the cross- talk can be large resulting in a significant reduction in system performance.

A Monte Carlo model was implemented that simulates the confocal system. The system consists of a laser source and optical elements that uniformly illuminate the confocal aperture. The aperture can be either a pinhole, slit, Nipkow disk, or linear array. The aperture and illumination beam are imaged into the tissue via the objective lens. Fluorescence signal is collected by the objective lens and imaged back onto the confocal aperture. Light passing through the aperture is brought back to focus. A dichroic beam splitter directs the emitted fluorescence signal to a detector. In the case of a pinhole aperture, a single detector is used; in the case of a parallelized aperture, an array of detectors are used.

Monte-Carlo simulation of confocal photon signal positions in tissue. Simulation results of the three-dimensional distribution of collected photon signal positions r_s for each aperture configuration (pinhole, slit, Nipkow disk, and linear array) in simulated tissue. Tissue surface is at z = −62.5 μm (grid plane), focus is at z = 0.

Monte-Carlo simulation of confocal photon error signal positions. Simulation results shown in terms of error in collected position ε for each aperture configuration (pinhole, slit, Nipkow disk, and linear array) in esophagus tissue.

Further information

To find out more about the research I am involved with visit the Biomedical Imaging Laboratory.

[1] L. J. Kleinsmith, Principles of cancer biology, Pearson Benjamin Cummings, San Francisco, 2006.
[2] A. F. Gmitro and D. Aziz, “Confocal microscopy through a fiber-optic imaging bundle,” Optics Letters 18, November 1993.
[3] M. Garcia, A. Jemal, E. M. Ward, M. M. Center, Y. Hao, R. L. Siegel, and M. J. Thun, Global Cancer Facts & Figures, American Cancer Society, Atlanta, GA, 2007.
[4] National Cancer Institute, “NCI issues clinical announcement for preferred method of treatment for advanced ovarian cancer.” http://www.cancer.gov/newscenter/pressreleases/IPchemotherapyrelease, January 2006.
[5] P. A. Wingo, “Cancer statistics,” Cancer journal for clinicians 45(1), pp. 8–30, 1995.
[6] K. Kerlikowske, J. S. Brown, and D. G. Grady, “Should women with familial ovarian cancer undergo prophylactic oophorectomy,” Obstetrics and Gynecology 80(4), pp. 700–707, 1992.
[7] K. Kerlikowske, J. S. Brown, and D. G. Grady, “Should women with familial ovarian cancer undergo prophylactic oophorectomy,” Obstetrics and Gynecology 80(4), pp. 700–707, 1992.
[8] J. F. Stratton, P. Pharoah, S. K. Smith, D. Easton, and B. Ponder, “A systematic review and meta-analysis of family history and risk of ovarian cancer,” British Journal of Obstetrics and Gynecology 105(5), pp. 493–499, 1998.
[9] I. Jacobs, “Genetic, biochemical, and multimodal approaches to screening for ovarian cancer,” Gynecologic Oncology 55(3), pp. 22–27, 1994.
[10] Y. Miki, “A strong candidate for the breast and ovarian cancer susceptibility gene braca1,” Science 266(5182), pp. 66–71, 1994.
[11] B. S. Kramer, J. Gohagan, and P. C. Prorok, “NIH consensus 1994: screening,” Gynecologic Oncology 55(3), pp. 20–21, 1994.
[12] A. R. Rouse and A. F. Gmitro, “Multispectral imaging with a confocal microendoscope,” Optics Letters 25(23), 2000.
[13] H. Makhlouf, A. A. Tanbakuchi, A. R. Rouse, and A. F. Gmitro, “Design of a multi-spectral channel for in-vivo confocal microscopy,” Endoscopic Microscopy II 6432(1), p. 643206, SPIE, 2007.
[14] A. R. Rouse, A. Kano, J. A. Udovich, S. M. Kroto, and A. F. Gmitro, “Design and demonstration of a miniature catheter for a confocal microendoscope,” Applied Optics 43, pp. 5763–5771, November 2004.
[15] A. R. Rouse and A. F. Gmitro, “Multispectral imaging with a confocal microendoscope,” Optics Letters 25(23), 2000.
[16] A. R. Rouse, A. Kano, J. A. Udovich, S. M. Kroto, and A. F. Gmitro, “Design and demonstration of a miniature catheter for a confocal microendoscope,” Applied Optics 43, pp. 5763–5771, November 2004.
[17] A. A. Tanbakuchi, A. R. Rouse, J. A. Udovich, and A. F. Gmitro, “Surgical imaging catheter for confocal microendoscopy with advanced contrast delivery and focus systems,” Endoscopic Microscopy 6082(1), p. 608202, SPIE, 2006.
[18] J. A. Udovich, A. R. Rouse, A. Tanbakuchi, M. A. Brewer, R. Sampliner, and A. F. Gmitro, “Confocal micro-endoscope for use in a clinical setting,” Endoscopic Microscopy II 6432(1), p. 64320H, SPIE, 2007.
[19] S. Srivastava, J. J. Rodriguez, A. R. Rouse, M. A. Brewer, and A. F. Gmitro, “Computer-aided identification of ovarian cancer in confocal microendoscope images.” in press, Journal of Biomedical Optics.
[20] P. Sharma and R. E. Sampliner, Barrett’s esophagus and esophageal adenocarcinoma, Blackwell Science, Malden, Mass., 2001.
[21] M. Garcia, A. Jemal, E. M. Ward, M. M. Center, Y. Hao, R. L. Siegel, and M. J. Thun, Global Cancer Facts & Figures, American Cancer Society, Atlanta, GA, 2007.
[22] M. Garcia, A. Jemal, E. M. Ward, M. M. Center, Y. Hao, R. L. Siegel, and M. J. Thun, Global Cancer Facts & Figures, American Cancer Society, Atlanta, GA, 2007.