Is Telepathology Right for Your Lab


Is Telepathology Right for You r Lab?

telepath main 1 flat

Despite a few obstacles yet to be surmounted, telepathology is a value-added product that can enhance the service provided by a pathologist.

By Robert Rainer, MD

More than two decades ago we entered the information age with the introduction of the personal computer. By combining this multipurpose machine with digital networks, we are able to collaborate across great distance with ease in real time. This combination has the potential to radically alter the way we practice pathology.

Telepathology is a term that describes using the tools of the information age to perform the daily tasks of a pathologist in a virtual world. Yet despite all the hype associated with this technology, few practical, working models exist.

Altering Applications

Telepathology has remained a nebulous concept that offers great promise to increase the efficiency of a pathologist. The role of this technology is ever changing and seems to be increasing in scope. Although telepathology is not a new concept, the use of technology to render a remote diagnosis in real time has been the major driving force behind it. The reality is that remote diagnosis may represent only a small fraction of actual telepathology applications. From personal experiences, my intentions were to use telepathology as a remote consultation tool. However, I quickly discovered the power of this technology to remotely manage laboratories from afar. The uses and definition of telepathology are constantly changing.

Perhaps even more challenging than a changing definition is the rapid pace at which the actual physical hardware is evolving. Moore’s law states that computer hardware will roughly double in processing power every 18 months. Thus, capital outlays quickly become obsolete. When adopting this technology, therefore, it’s wise to be as modular as possible when purchasing a system.

Telepathology consists of at least four basic parts:

1. a computer, needed to coordinate other devices;

2. software, to collaborate and process the information;

3. a digital network, which will span any physical distance; and

4. an array of image acquisition devices to capture just about anything a pathologist might need for diagnosis.

Digital Images

The basic element of a telepathology system is a digital image, which is a binary representation of real world objects. All laboratorians are familiar with digital images since they are integral to the functionality and means in which we interact with a computer. However, it’s the inherent properties of digital images that make them so powerful. For example,

* The objects can easily be beamed across a digital network at the speed of light and reassembled to be exact replicas.

* The images have plasticity since they can be manipulated to enhance features within them.

* The images are stable; they can be copied an infinite number of times and be stored for long periods of time.

Digital Hurdles

Despite all of these attractive features, however, digital information is essentially virtual and can be lost. Thus, routinely scheduled back ups are an essential component for any telepathology system.

One of the largest obstacles faced by telepathology systems is the lack of an adopted standard for the technology. Many questions remain such as: What are optimal compression ratios? What bandwidth is required for a given application? What storage standard is best?

Images stored in one format, for instance, may be useless in a few years if a better imaging standard emerges. Radiology is further along in this arena than pathology with regards to remote imaging systems–the American College of Radiology (ACR) has adopted a set of clearly defined standards for teleradiology.

This is not to say, though, that pathologists have been idle. In reality, pathologists have been very active in defining the first telepathology standard. The College of American Pathologists developed a supplement to the ACR’s DIgital COmmunications in Medicine (DICOM) standard. DICOM supplement 14 describes what information should be attached to a given image and specifies a means that any DICOM device can display and store that image. This is a first step in adopting telepathology standards, but many more need to be taken.

Communication Standards

The DICOM standard does not specify what types of devices are to be used to capture images. Two basic types of devices can be attached to a microscope to acquire images: an analog and a digital system.

An analog system is a video camera. The technology used for these devices is mature and widespread and used in the consumer video market and by professional broadcasters. The standard used by this modality, NTSC, is the same standard used by every television set in the United States. It has a sync signal and an interlaced video signal at 15 frames per second. Their RCA type plug can easily identify most composite NTSC video systems, which put out a VHS signal.

Analog video systems also can put out a red, green, blue (RGB) signal or super video (SVHS) signal to a recording device. These two standards enhance the quality of an NTSC signal by separating the RGB color information and the chrominance (color) and luminance (brightness) information. RGB and SVHS systems will cost more than VHS systems. The SVHS system will integrate easily into most computer systems, whereas RGB systems usually require special hardware.

PAL Standards

Almost all video equipment sold in the United States conforms to the NTSC standard. This is an old standard with low resolution and accounts for many of the intrinsic problems pathologists encounter when looking at video images. It quickly becomes apparent that the image displayed on the video screen is not nearly as good as what we see in the microscope.

One way to immediately improve image quality, however, is to utilize the PAL standard used by the Europeans and Japanese. PAL is intrinsically better than NTSC and provides a better picture. Most computer capture cards will have a PAL mode, so integration is easy.

Finding a PAL-compatible monitor in the United States, however, can be difficult. PAL is only a stop-gap measure. We are on the cusp of a new and better standard–High Definition TeleVision (HDTV) is starting to be realized in the United States, offering twice the resolution of the NTSC standard and a different aspect ratio. It’s currently very expensive but should improve the quality of analog video devices.

Digital Cameras

Video signals–composite, SVHS or RGB–must have their voltages converted to digital values to be used by a computer. Digital cameras will read the amount of light and pass this information to the computer’s graphics system. These cameras are relatively new and rapidly evolving. They have two means of interfacing with the computer: Cameras are either Universal Serial Bus (USB) compliant or IEEE 1394 (Firewire) compliant.

Firewire offers the ability to move greater amounts of information than USB, but more USB options exist. In addition, Firewire can move full screen video at more than 30 frames per second through a computer. These cameras offer great ease of use and interface directly into a computer’s hardware and operating system. Although digital cameras appear to be the trend toward the future, the technology behind them is in its infancy.

Sensing Devices

Whether a system is analog or digital, the sensing device will be a Charged Coupled Device (CCD) array, which converts the physical light energy into electrical signals. These devices have physical limitations; the most important factor is related to the image they produce relative to their size. A camera’s CCD chip, for example, will most often be either one-third inch, one-half inch or two-thirds inch in size. Bigger is better since more physical information can be captured.

Some higher-end cameras will have bigger CCD chips along with a substantially higher price tag. Many of the lower cost cameras will utilize a one-third inch CCD chip and, therefore, only capture a small fraction of what’s displayed in the microscope.

Half-inch CCD chips suffer from the same problem but offer a compromise between price and performance. The ideal size is a two-thirds inch CCD. These cameras are still affordable and can capture most of the field of view.

Also affecting the price of a video camera is the number of CCD chips present in a camera. Many higher-end cameras contain three chips that will sense RGB-filtered light. This produces a superior picture to the single chip camera.

Coupling Devices

The final factor that determines an image’s quality is the microscope-to-camera coupler. This device will extend the focal plane of the microscope to the CCD array. The image can be magnified or reduced. This coupling, along with the CCD size, will determine how much of the field of view will be presented to the camera.

The cost of a coupling device is equivalent to a single chip camera. There are two types, depending on the type of camera:

1. A “C” mount is a simple screw-on device.

2. A bayonet mount has a locking device. Generally, a bayonet mount is associated with professional video equipment and by having such a distinction, it’s associated with a higher price.

Each device will allow an image to be captured from a microscope. This is a difficult task fraught with numerous technical challenges. The primary reason: A microscope was not designed for machine vision. Telepathology systems of the future may rely more on a slide scanning device than a microscope. The problem faced by this method is the huge files that must be generated to display a virtual slide on a computer system with all of the possible focal planes. Thus, today’s telepathology systems rely on sending a video representation of a slide. Until a few years ago, computers did not possess the raw processing power to videoconference using software COders and DECoders (CODECs) alone. Specialized boards were required to handle the digital video. However, computers with a Pentium II or better possess the raw processing power to use software CODECs for video conferencing.

Software CODECs greatly simplify hooking up to computers for a telepathology session. Instead of identical hardware, the two computers only require having identical CODECs installed. Even more interesting is that both machines do not have to run the same software to collaborate. This opens the door for users to utilize software packages that suit their preferences.

There are several software packages that will accommodate video conferencing. Examples include Netmeeting, CuSeeMe and VDOphone. All have similar functionality–they will allow receiving of remote video and sharing of applications and data via a white- board. The remote video will allow the entire case to be viewed by a remote pathologist; the shared whiteboard will allow regions of interest to be discussed and diagnosed.

Case in Point

An example of a dynamic telepathology system–where pathologists are interacting with one another in real time–is as follows: A pathologist in one location calls a pathologist in a different location using a videoconferencing package. The sessions start with the remote pathologist talking in a face-to-face conversational tone. Details of the case are discussed and specific questions are asked. The video switches to a recording of processing the gross specimen. The video is paused or replayed so pathologists can discuss salient points. An area of interest is noted, so a high-resolution still is placed on the white- board for further discussion. The video then switches to a live picture of a glass slide on a microscope. Again, areas of interest are captured and placed on a shared whiteboard. Specific areas are pointed out, and higher or lower power images are obtained. The case is discussed, worked-up and diagnosed across great distance.

The other type of general telepathology system is a static, or store-and-forward system. The remote pathologist simply captures a series of images, then e-mails them to a consultant. Because the consultant is dependent on the remote pathologist to select the appropriate images, this method is limited by selection biases. The main advantage to this method is the relatively low costs associated with maintaining a static telepathology system. All you need is an e-mail account and Internet access. The initial capital outlays are essentially the same between a dynamic and a static system.

Dynamic systems have advantages over static systems, but they also have a higher overhead associated with them. The greatest expenses encountered in telepathology are realized in the monthly bill generated by the idling and use of the digital network. When setting up a telepathology system, it’s vital that this fact be well understood.

Nothing will have a greater impact on the usability of a system than the network. When a network is purchased from the phone company, four factors determine what you will get, including:

1. the size of the digital pipe,

2. the distance between the two points of presence,

3. the time you use the network and

4. the infrastructure of the phone company.

The amount of information that can be moved across a network per second is bandwidth. The size of a video image and the frame rate of that image are directly related to the bandwidth of a network. The higher the bandwidth, the better quality your video will be and the more it will cost. The monthly cost will also include a fixed line and variable charge based on usage. The network can cost anywhere between $200 and $2,000 a month to maintain, depending on the configuration.

Why not just use the Internet for telepathology? After all, it is a packet-based network, versatile and permeates all nooks and crannies of the United States. Unfortunately, it cannot be reliably used for videoconferencing since you can’t guarantee the bandwidth. Thus, you might have a fast connection to the Internet, but once you are on the backbone, your packets do not get any more priority than a teenager playing a game. So, digital networks needed for telepathology have to run on leased lines.

ISDN lines, for instance, provide bandwidth at 128 kilobits/second, which allows for adequate transmission of video information. ADSL lines are distant dependent, but they can reach speeds of 1.5 megabits/second. T1 lines are not distant dependent and will provide similar bandwidth. Multiples of T1 lines come in the form of OC3, OC12 and OC48, which are typically fiber optic lines used by the phone companies to serve large customers and central offices. It’s doubtful that telepathology will ever require more than a T1 line; many applications will run on a couple of combined ISDN lines.

New Services

A relatively new service is Asynchronous Transfer Mode (ATM). It provides up to 155 megabits/second and offers the ability to set up virtual networks. Unlike other digital services, voice, video and data can travel on separate networks using ATM. Thus, one modality will not gain better performance at the expense of another. This feature allows a pathologist to set up a network with a given amount of bandwidth for a given application to a given physical location. ATM is not widely available and is expensive. However, if you are considering setting up a telepathology application in the same town, this option should be explored because of the benefits it provides.

Two major factors have led to lack of sustainable telepathology systems. The first and most powerful force is the relatively inexpensive means of moving diagnostic material from one place to another. The second factor, which hindered the sustainability of telepathology, is that until recently no one was willing to pay for this type of service. It’s pure overhead to a practice, and this fact, coupled with falling reimbursement for diagnostic services, has led to little interest in utilizing this technology. Although reimbursement for telepathology services is available, it’s limited and applicable in rural, under-served areas only.

Do telepathology applications ever make sense to implement? If assessed as a value-added service, the technology can be a useful adjunct to a pathology group. The biggest impact of the technology is the ability to extend the geographical reach of a pathology group. Problematic cases can be handled in a time-efficient manner, and more times than not, the issue with the case can be resolved.

From experience, three types of cases lend themselves to telepathology consultations. The first case is the handhold. A pathologist is fairly sure of the diagnosis, but wishes to have a colleague confirm his/her impression. The second type is one in which an inexperienced pathologist might not recognize a described entity. When the case is shown to a more experienced pathologist, the diagnosis is usually rendered. The final case: the just plain hard case. Special stains may be required, re-cuts may be needed to visualize the changes or questions about micro invasion arise. By utilizing a telepathology system upfront in these cases, a great deal of the groundwork needed to achieve a diagnosis can be accomplished before critical deadlines to initiate special studies. The average time for a telepathology case is roughly 15 minutes; handholds can last as little as two minutes, and difficult cases can last for 30-40 minutes.

Additional Models

Several other telepatholgy models are available. One is for a group that covers only one hospital. Pathologists can set up a fairly inexpensive link to the operating rooms that will allow the pathologist to show the surgeon images and discuss the relevance. For instance, a gross specimen can be discussed with the surgeon and orientated. Exact areas of concerning margins can be conveyed both grossly and microscopically. Thus, better communication between the two parties can be achieved.

Use of this technology also can enable one pathologist to provide specific services to several hospitals in an efficient manner. By setting up a virtual onsite office, laboratory problems can be addressed as they arise. Consultation can be provided in a real-time manner and quality assurance issues can be addressed in a scheduled or as-needed basis. Virtual management is a very powerful use of this technology, allowing pathologists to cover laboratories in many of the smaller hospitals that cannot support a full-time pathologist. This modality can also be extended to physician office laboratories.

Despite the lack of direct reimbursement, and the overhead associated with maintaining the system, telepathology is a value-added product that can enhance the service provided by a pathologist. The technology has the potential to break geographic barriers that have confined pathologists to their local regions.

Dr. Rainer is a pathologist at Spartanburg (SC) Regional Healthcare System.

Telepathology consists of at least four basic parts:

1. a computer, needed to coordinate other devices;

2. software, to collaborate and process the information;

3. a digital network, which will span any physical distance; and

4. an array of image acquisition devices to capture just about anything a pathologist might need for diagnosis.