Return to CAP Home
Printable Version

  Digital cytopathology and telecytology: a method coming of age

 

CAP Today

 

 

 

January 2011

Walid E. Khalbuss, MD, PhD
David C. Wilbur, MD

Telecytology, a component of the broader field of telepathology, is the practice of cytology at a distance, using telecommunication networks (for example, the Internet) to transfer digital images from one site to another. The emergence of fast and high-resolution digital imaging tech-nology and computers with high processing capacity has made telecytology possible.1-3 Three methods are currently available for telecytology, each having its own strengths and drawbacks:

  • Static image capture and transfer is the most simple, requiring only the capture of microscopic images via a digital camera and the transfer of those images for remote evaluation (store-and-forward).
  • Dynamic (real-time) evaluation requires a constant linkage between sending and receiving station with the ability to move about the specimen, focus, and change magnifications, using either remote robotic or local control.
  • In whole-slide scanning and transmission systems, entire specimens are captured at high resolution, with transmission and manipulation of the entire specimen possible from one site to another.

The latter two systems are generally more expensive and demanding on networks, but they offer the telecytologist remote control of better quality digital images. Both gynecologic and nongynecologic cases are amenable to telecytology for immediate interpretation (rapid adequacy assessments) and for primary interpretation and secondary opinion consultation.

Transmission of static images is relatively simple; it requires only a camera and a network connection. Images can be relatively small (in terms of memory and transmission requirements). However, they suffer from representing only limited portions of the specimen and hence there is a potential bias of the image acquirer relative to the image observer, who sees only what the sender wants the observer to see. In “easy” cases this bias may not be a problem, but in more difficult cases (the type that would be more routinely shared in consultation), this biased partial representation might be very much an issue. In addition, lack of focusing ability and the issue of image manipulation (contrast, brightness, and color) may all be important impediments to a successful outcome.

Real-time image transmission involves an image stream, sent immediately upon acquisition and updated continually as the specimen is re-viewed. This type of imaging potentially allows for review of the entire specimen, with focusing and changes in magnification as required. In addition, there can be real-time interaction with the sender during the interchange. There are several systems for real-time image transmission, some of which are controlled at the local site and others that the distant observer can control robotically. Such systems allow for an unbiased review of the specimen because the observer can control the review or, at a minimum, instruct the local site to, for example, “move left, go to high magnification, focus.” Real-time systems can be cumbersome to use, however. They require large bandwidth network connections and can therefore be slow. Particularly in cytology situations in which screening or review at high magnification is required, slow image refresh rates and remote command transmission can lead to observer frustration and even overload networks. Static image and real-time systems have been successfully deployed for cytology case consultation and for rapid evaluations (adequacy assessments).

Whole-slide imaging, or WSI, technology offers significant advantages over static and real-time image transmission but, as usual, these advantages come with a cost. WSI equipment allows for image capture with high enough magnification to produce an image of the entire specimen with a resolution similar to that which is used routinely in a standard light microscope. Hence, instead of partial images of the specimen, the observer of a WSI can review the entire specimen in a fashion similar to reviewing the actual slide under the microscope. WSIs are memory-intensive. They require in the hundreds of megabytes of memory for a single image, compared with a standard static image, which might require three to five megabytes of memory. Further, two-dimensional images do not allow for focusing, and this is a particular problem with cytology specimens, which are routinely more three-dimensional than a standard histology tissue section. But WSI has methods to tackle this problem as well. Multiple scans of the same specimen, taken at different focal planes, can be “stacked” into a final composite image (referred to as a “z stack”). As would be expected, though, each of these planes requires the same amount of memory, and as such z stacks can be slow to load, transmit, and manipulate, and therefore lead to significant observer frustration. Storing large numbers of such scans will also require a very large server capacity.

As in all things computer- and network-related, these deficiencies can be looked at as mere “engineering” problems that are sure to be sorted out in the near future. In the meantime, two-dimensional images of cytology specimens can be made to exhibit a more three-dimensional appearance through multiplane “upfront” scanning followed by a software “trick” that incorporates the best focused image at each pixel into the final two-dimensional composite (or integrated) image. Improvements in the focusing of cytology specimens has been significant when using such technology and provides an excellent interim solution to the cytology-specific three-dimensionality problem while the field waits for the technology to catch up with a fully focusable image.

Other disadvantages of WSI are the high cost of the currently available scanners and the time scanning requires, which can range from three to five minutes for a liquid-based slide and up to 10 minutes for a conventionally prepared smeared slide. It can take hours of scanning time for the multiplane scanning that is necessary to generate composite two-dimensional or z stacked images.

Once images (in whatever form) are acquired, what can they be used for? On the clinical side, they are used most commonly for teleconsultation, either on final specimens4 (expert consultations and quality assurance reviews being most routine) or real-time rapid interpretations, such as specimen adequacy assessments on fine-needle aspirations performed at a remote site lacking cytology expertise.5 Such transmissions could be from office to office within the same campus, but theoretically, in the digital environment, these consultations could be from anywhere in the world to anywhere else having a high-speed Internet connection. Although technology today makes interinstitutional clinical use possible, there will be significant practical challenges to widespread implementation. Information technology restrictions of institutional firewalls, other security concerns, and routing of HIPAA-protected patient information along with images—all challenging issues at this time—will require solutions.2

Slide archiving is another clinical use. Institutions receive numerous requests for patient consultations, for expert opinion and for patient care continuity among institutions, for which slides need to be returned. Receiving institutions can now keep a permanent, near perfect record of the slides via WSI scanning. In addition, adding slides to digital databases (PACS) will make it possible for them to be merged with other clinical information, providing a permanent and complete electronic record of all information—including pathologic samples.

Telecytology has also been used effectively for distance-based continuing education with teleconferences using static images accompanied by a lecture, real-time microscopy sessions, or tumor boards being broadcast regionally within health care systems or out to other users at large.6

In the future, WSI may be adaptable to other uses such as cytology proficiency testing, where having a well-defined, well-validated WSI challenge would be easier to maintain and distribute than are the extensive glass slide collections that require maintenance now by PT providers. Use of WSI for archiving and presenting rare cases, unusual presentations, and classic examples of entities would have significant value.

Novel telecytology applications already being explored involve combinations of automated gynecologic slide screening with machine acquisition of relevant fields of view, which can be distributed automatically to remote reading stations for review. Initial trials of such a system have shown that this method can be effective in the initial triage of specimens, and, in many cases, adequate for final diagnosis.7

Telecytology procedures can be performed today, but the capability is in a relatively rudimentary stage of development and implementation. The problems to be addressed are those of multiplane (three-dimensional) specimens, scanning speed, bandwidth required for rapid transmission, and IT related to firewall security and the transmission of protected patient information. Investigators also need to work on the ergonomic issues of image review. Simple mouse-driven computer screens may not be efficient and can create significant user fatigue. Other methods of review, such as images in high-resolution displayed on large walls or bigger monitors (where magnification is controlled by the observer moving away or closer), may be necessary to increase efficiency and accuracy, and ultimately may lead to better acceptance of the method. As progress continues to be made, these issues will almost certainly be overcome. In addition, validation is an important issue—that is, can we interpret cytology specimens by these methods as well as we can using standard microscopes? The FDA is already considering standards for the practice of telepathology, including methods and equipment. Protocols will almost certainly be proposed for interpretation validation as well, to ensure quality interpretation results. Interest in telecytology applications is rising, and enthusiasm for its use is building. How cytologists will adapt remains to be seen.

References

1. Pantanowitz L, Hornish M, Goulart RA. The impact of digital imaging in the field of cytopathology. Cyto-Journal. 2009;6:6.

2. Williams S, Hendricks WH, Becich MJ, et al. Telepathology for patient care: What am I getting myself into? Adv Anat Pathol. 2010;17:130–149.

3. Weinstein RS, Graham AR, Richter LC, et al. Overview of telepathology, virtual microscopy, and whole slide imaging: prospects for the future. Hum Pathol. 2009;40:1057–1069.

4. Yamashiro K, Kawamura N, Matsubayashi S, et al. Telecytology in Hokkaido Island, Japan: results of primary telecytodiagnosis of routine cases. Cytopathology. 2004;15:221–227.

5. Kim B, Chhieng D, Jhala N, et al. Dynamic telecytopathology has equivalent efficacy with on site rapid cytology diagnoses for pancreatic carcinoma (abstract). Cancer Cytopathol. 2006;108:357.

6. Mulford DK. Telepathology education: reaching out to cytopathology programs throughout the country. The ASC Bulletin. 2006;43:25–30.

7. Eichhorn JH, Buckner L, Buckner SB, et al. Internet-based gynecologic telecytology with remote automated image selection: results of a first-phase developmental trial. Am J Clin Pathol. 2008;129:686–696.


Dr. Khalbuss, a member of the CAP Cytopathology Committee, is associate professor of pathology and director of cytology, University of Pittsburgh Medical Center-Shadyside, Pittsburgh, Pa. Dr. Wilbur, past chair of and now advisor to the CAP Cytopathology Committee, is professor of pathology and director of cytology, Massachusetts General Hospital, Boston.