Telemedicine in Gastrointestinal Endoscopy

In the paper on tele-endoscopy by Rabenstein et al. published in the September 2002 issue of Endoscopy [1], the authors present their view that tele-endoscopy (i. e., conveying video endoscopy images over a network for the purposes of diagnosis from a remote location) is feasible only with high-end picture coding/compression schemes - such as Moving Pictures Experts Group 2 (MPEG2) or Motion Joint Photographic Experts Group (MJPEG) - which require high-capacity transmission routes, such as asynchronous transfer mode (ATM). They claim that it is not feasible with simpler picture-coding schemes that are tailored to low-capacity routes, such as those provided by the public telephone network, such as the Integrated Services Digital Network (ISDN).

We commend the thorough approach taken by the authors and their desire to apply scientific objectivity to a highly subjective measurement, such as perceived picture quality. We do, however, have some difficulty in supporting their view against low-end solutions, particularly since their high-end solution is so costly and can only be deployed in specialist locations where there is access to high-capacity network connections. Our own research in this field [2] [3] [4] [5], which dates back to the early 1990s, approaches tele-endoscopy from the opposite position - that of minimal cost (i. e., up-front, as well as running costs) and availability across the medical profession regardless of geographic location, by using the public telephone network. As a consequence, our findings are rather in contrast to those of Rabenstein et al. - namely that tele-endoscopy is indeed feasible over low-capacity transmission routes, provided that certain sensible precautions and operating methodologies are adopted. In addition, if tele-endoscopy of this kind was to be made widely available, medical training in this field would be certain to benefit significantly [2] [3] [4] [5].

In 1994, we successfully demonstrated (for the first time in the United Kingdom) the transmission of video endoscopy images over the public telephone network. Our objective was to develop the underlying principle of tele-endoscopy as a means of remote diagnosis and a training methodology, and to show that this could be achieved over the public telephone network with minimal terminal costs [2] [3]. By exploiting the ubiquity of the public telephone network, tele-endoscopy links can be set up between almost any location in the world, bringing key medical skills to locations in which patients hitherto would not have benefited. By contrast, high-capacity links, either in the form of dedicated point-to-point installations or specialist switched network services (e. g., ATM), are generally only available between large infrastructure facilities, such as hospital sites, universities, etc., and they incur significant on-line costs in comparison with a simple telephone call.

In our research [2] [3] [4] [5], we conducted a number of trials over a single ISDN telephone line at a data rate of 128 kbit/s, and also three aggregated ISDN lines, providing 384 kbit/s. The terminal apparatus consisted of commercially available video conferencing stations - one at the patient site, connected to the video endoscope system, and the other at the remote site. Figure [1] shows the apparatus at the hospital site (Ipswich National Health Service Trust), and Figure [2] shows one of our colleagues (Dr. Brian Saunders) at the remote site in a nearby town. His image appears on the video monitor at the hospital site. During trials, we carried out a number of endoscopic examinations with patients’ consent. On some occasions, the attending endoscopist was under the direction of a specialist at the remote site, while on others the attending endoscopist directed the proceedings, which were monitored by trainees at the remote site. In either configuration, we found that the picture quality was entirely adequate for the procedure, and a suitable diagnosis was reached for each patient [2] [3]. As can be seen in Figures [1] and [2], the technology was available back in 1994 to combine the endoscopic view, the view of the patient and/or the view of the magnetic endoscope imaging system, which we first described in 1993 [6] and have improved upon since [7].

Figure 1 The system used by the authors in 1994 to send the endoscopic view and magnetic endoscope imaging view simultaneously to a distant expert using a single ISDN2 telephone line.

Figure 2 The combined endoscopic and magnetic endoscope imaging view as seen by the distant expert.

It would be wrong of us to imply that the picture quality at the remote site was as good as the original. The picture quality did deteriorate while the endoscope was in motion, but the type of coding schemes we used (H261, etc.) intrinsically improve picture quality when the image is stationary, ultimately reaching full resolution within a second or so. Alternatively, or in combination, we could use the ”freeze” facility at the terminal station, which captures a video frame and displays it at full resolution. In either case, the resulting picture quality was very close to the original, and on many occasions was indistinguishable from it. Examples of this are shown in Figure [3].

Figure 3 Examples of picture quality using a single ISDN2 telephone line (a) and three such lines (ISDN6) (b), in comparison with the original video of the colonoscopy (c).

Of course, coding schemes specifically designed for the narrow-band channels afforded by the telephone network have improved greatly since we conducted our research (some 8 years ago), so it can reasonably be expected that picture quality has also improved. In addition, the new breed of digital subscriber line (DSL) technologies give certain telephone lines bandwidths of megabits per second, which can deliver excellent picture quality while still retaining the geographic coverage of the public telephone network. All of this strengthens our belief still further that tele-endoscopy over the telephone network is entirely feasible.

In addition to the discussion that has taken place so far regarding picture quality, we strongly advise that feasibility studies on tele-endoscopy should also address issues of voice quality (in instances in which a commentary channel is used). In our experience, establishing satisfactory voice communications over a tele-endoscopy link proved as challenging as picture quality, although for entirely different reasons. It is surprisingly difficult to prevent echoes and reverberation loops arising between the two end locations, whilst at the same time creating for the participants an environment in which normal conversations can be conducted.

To sum up, we fully agree with the need to specify an appropriate picture coding scheme and data rate for video-based medical applications, such as tele-endoscopy that will guarantee an acceptable picture quality after transmission over a network. Indeed, we should all strive to create a set of standards that define these parameters for all pertinent areas of telemedicine. However, we feel that it is premature and inappropriate at present to recommend or reject coding schemes purely on the basis of picture quality, vital though that undoubtedly is. It is important to take in wider considerations, such as cost, flexibility, availability, medical practices, audio requirements, etc. We believe that our work on tele-endoscopy [2] [3] [4] [5] has demonstrated that simple coding schemes of the kind developed for the public telephone network can deliver an entirely acceptable picture quality in most circumstances and at minimum cost. Our approach also affords significant flexibility in the placement of the terminal apparatus, limited only by the geographic coverage of the telephone network.

References

• 1 Rabenstein T, Maiss J, Naegele-Jackson S. et al . Tele-endoscopy: influence of data compression, bandwidth and simulated impairments on the usability of real-time digital video endoscopy transmissions for medical diagnoses.  Endoscopy. 2002;  34 703-710
• 2 Heatley D JT, Bell G D, Saunders B. Tele-endoscopy: the application of telecommunications to medical endoscopy. London; Royal Society Lecture on Telemedicine July 1995
• 3 Heatley D JT, Bell G D, Saunders B. Telecommunications in medical endoscopy.  J Telemed Telecare. 1995;  1 237-238
• 4 Bell G D, Rowland R S, Heatley D JT. Tele-endoscopy need no longer be prohibitively expensive.  Curr Med Lit Gastroenterol. 1998;  17 63-67
• 5 Bell G D, Rowland R S, Heatley D JT. Inexpensive and readily available tele-endoscopy using a modified Pentium PC and a single ISDN2 telephone line [abstract].  Gut. 1998;  42 (Suppl 1) A18
• 6 Bladen J S, Anderson A P, Bell G D. et al . Non-radiological imaging of endoscopes.  Lancet. 1993;  341 719-722
• 7 Rowland R, Bell G D. Non-radiological technique for three-dimensional imaging of intestinal endoscopes: a new improved method of computerized graphical 3-D representation of the endoscope and patient’s skeleton.  Med Biol Eng Comput. 1998;  36 285-290