DVD PRO Conference
V2B Conference
TechVideo Expo
Current Issue
Article Archive
News Archive
Buyer's Guide
nbsp; Home Magazine eNewsletters Events Contact Navigation

Current Issue Current Issue
Buyers Guide 2001

Buyers GuideCompany SearchProduct Search

News Indices

CD TrackerCD/DVD-ROM IndexFact, Figures & FindingsConference Calendar

Tech Video Expo

DVD TodayDigital EarfulSubscript
Ad Links

Archives and Interchange: Will Medical Imaging Prove a Healthy Market for DVD?

Philip De Lancie

November, 2000 | Ask most people how DVD is used, and they're likely to focus on the field of entertainment, and perhaps mention corporate communications as well. But the same attributes that make DVD attractive for video entertainment and corporate information–such as a widely available playback platform and the capacity to handle large, high-resolution files–make the format a contender in areas far removed from Hollywood, Wall Street, or Madison Avenue. One such area is medical imaging, where the way has already been paved by the acceptance of CD-R technology.

The field of medical imaging involves several distinct applications, including archiving, interchange, and informatics. In the medical field, "archiving is the preservation of patient images and information for many years in a secure manner," says Robert Petrocelli, president and CEO of Heartlab, Inc. in Westerly, Rhode Island, a company that sells DVD-based medical archiving systems for cardiology. "Regulations typically require that patient images be kept for seven years, and there are some institutions that never discard this data. In the ideal scenario, media is only removed from an archival system to be discarded."

In contrast, Petrocelli describes interchange as "the exchange of patient images and information, typically between hospitals. This can happen electronically or using media. In either case, the exchange should be secure and not subject to modification after it has left the sender. Interchange media is also sometimes discarded after its contents have been transferred to the receiving image-management system. Interchange media must be highly standardized and widely readable."

The patient information (as distinct from images) that Petrocelli refers to falls into the category of informatics, according to Chuck Alcon, vice president of sales and marketing at MicroBoards Technology of Chanhassen, Minnesota, a longtime supplier of CD and DVD recording media and production systems that has recently opened a medical imaging division. "Informatics," Alcon says, "refers to the processing by hospitals and the medical industry of generic data such as patient records, insurance information, and past history." While the volume of this data for a given patient does not approach the size of the images themselves, storing, updating, and retrieving the information along with the images adds another consideration to the overall picture.

The primary distinction in requirements for systems and media, however, is between archiving and interchange. "For the purposes of interchange, it's important that two different vendors of systems can read and write the same media," says David Clunie, development director of medical imaging products at Comview Corporation in Pleasanton, California, a vendor of capturing, viewing, distributing, and archiving systems for cardiology. "But for archiving, in theory, it doesn't matter what media you use inside. For example, some people use 12-inch optical disc platters, some use 5.25-inch MO, some use Exabytes or DLT or AIT tape-based technology. And recently, an increasing number of people have been trying to use DVD-R and DVD-RAM in that role as well."


Archiving, interchange, and informatics are concerns across all areas, or modalities of medical imaging, which includes disciplines such as MRI/CT, ultrasound, and cardiology. Classifying other areas under the general umbrella of radiology, Heartlab's Petrocelli describes the difference between cardiology and the rest. "Cardiology produces more data per patient than radiology," he says. "But the number of radiological studies is much greater. So radiology is a high-transaction rate, low-transaction size application, while cardiology tends to be moderate-transaction rate, high-transaction size." The higher the transaction size, the greater the need for a large-capacity medium like DVD.

Also of interest in relation to DVD is the fact that cardiology is out in front of other modalities in terms of implementing new technologies. "Although the installed base of cardiology systems is very small in comparison to the rest of the modalities," MicroBoards' Alcon says, "cardiology typically leads the technology curve, and then the rest of the modalities begin to adopt the new technology." This description seems to have held true with regard to the advent of CD-R initially adopted as a replacement for cine film in cardiology. "Cine film is simply 35mm movie film that records a full motion record of an x-ray angiogram in the cardiac cath-lab," Petrocelli explains. "This film is expensive to produce, can't be duplicated, and is frequently misplaced."

Despite its drawbacks, cine film was used in cardiology because alternatives based on video lacked the required resolution. "The x-ray beam, once it's been through the patient, can be used to expose film without actually having to go through a video chain," Comview's Clunie says. "So you get very high quality, much higher than one would get off a video image."

As computerization took hold in all types of workplaces in the late 1980s and early 1990s, vendors of imaging tools developed systems called PACS, for Picture Archiving and Communication System. "There are multiple workstations set up as viewing stations," Alcon says, "so that doctors and technicians can review the results of patient procedures over the network."

In many areas of medical imaging, the main need for viewing is within the same facility where the image is generated. "Outside cardiology," Clunie says, "most of the focus has been on exchanging images via the network. With CT scans, MRIs and ultrasounds, it turns out that it's unusual to want to transfer images between institutions; much more often they're used within the same organization. So a high-speed local area network is more appropriate for storing those images and retrieving them." The pattern in cardiology, however, is quite different. "There is a referral pattern," Clunie says, "where a cardiologist will do a study, and then if the patient needs surgery, they will pass the angiogram on to the cardiac surgeon who does the coronary artery bypass, for example. So they have to interchange these very large collections of images."

With cine film sticking out like a sore thumb in an otherwise digital world–"They used to literally ship a can of film to the surgeon," Clunie says–the consensus by the mid-'90s was that a digital medium was needed that would allow image interchange between different PACS. Vendors, radiologists, and cardiologists formed a standards committee dealing with digital imaging and communication for medicine, dubbed DICOM. Clunie currently serves as industry co-chairman of the group, as well as chairman of the working group on inter-change media.

"DICOM looked around for a medium," Clunie recalls, "and recordable CD–which was already being used informally–seemed to be the best choice. So DICOM defined an 'interchange format,' including a network protocol as well as a choice of media, objects, and file system to put onto CD-R. It's been a very popular application in cardiology. You can record a patient's entire angiogram onto one CD, and then take it to another station and plug it into a DICOM viewer and look at the images, without having to worry about film and so forth."

Alcon adds that the protocol is Windows NT-based and runs on off-the-shelf architecture. "But the actual DICOM software," he says, "has very specific requirements as far as how images are transported over a network, how they're viewed, and how they're stored." As a result of DICOM, Alcon says, CD-R usage has been widely accepted in the health-care environment.


"CD-R has been wildly successful in cardiology, especially in the cardiac cath-lab," Heartlab's Petrocelli says. "It was selected because it is standardized, widely available, secure once written, and stable. But its main limitation has been capacity. As additional information is added to the basic unit of interchange for a cardiac patient, denser media with similar attributes will be needed."

"Capacity is a big problem," Comview's Clunie agrees. "The images we capture at present are 512 x 512 x 8-bit. We want to move to using 1024 x 1024 x 10-bit. One angio session just will not fit on a CD-R, and handling multiple volumes of CD-R is a pain."

"The other thing," Clunie continues, "is that we're trying to integrate images from multiple acquisitions onto the same disc. For example, a patient might come in for a single session, and have various different kind of imaging studies: echocardiography, angiography, intravascular ultrasound. Reports and measurements may be generated, waveforms may be recorded, and ideally we'd like to put all that on a single disc that can be sent to the next physician down the chain."

One might think that in the age of the network, no physical media at all would be required for interchange. But the size of the image files is sufficiently large to discourage a move toward network-based interchange between health facilities, which would hog too much band-width. "We exchange stuff over the Internet all the time now," Clunie says, "but media has become well-established as a very low-cost way of shipping stuff around. You can talk about all this electronic technology as much as you want, but couriers are fast, reliable, and secure."

Magneto-optical (MO) media are used in some storage and archival applications, but are largely out of the running when it comes to interchange because of a lack of standardization on form factor and capacity. Tape-based media are also used for archiving, but Clunie says that "optical media are so much more robust in the face of being stuck in a mailer and sent across the country, compared to tape, that no one has really seriously considered using tape as an interchange media."

DVD-R, on the other hand, fits the description of being "denser media with similar attributes" to CD-R, and has the added attraction that the DVD-ROM drives used to read DVD-Rs will read existing CD-Rs. "DVD-R is in a similar position to where CD-R was just a few years ago," Petrocelli says. "The great majority of PCs are shipping with DVD-ROM readers. DVD-R writers are still relatively expensive for desktop use. But DVD-R has much greater capacity, and will see significant increases in capacity during its life-cycle. It is by far the most interchangeable media next to CD-R."


Clunie agrees that DVD-R has tremendous potential for medical imaging. But while Petrocelli's company has moved aggressively to incorporate the format into its archiving products, Clunie says neither DVD-R or DVD-RAM is "quite ready to be standardized" for interchange applications.

For DVD-R, one of Clunie's concerns is the high cost of the writing drives. Another is the fact that the drives are available only from Pioneer. "The rules under which DICOM is formed do not allow us to standardize on a single-vendor solution," Clunie says. "Even if we wanted to, it's not good to OEM from a single source. So the fact that other people haven't signed on to make DVD-R drives is a real killer for that particular format."

While acknowledging the single-vendor problem as "the most important barrier to DVD-R's acceptance as an interchange format," Petrocelli says that the problem will fade when lower cost, consumer-oriented, 650nm drives enter the market early next year. MicroBoards' Alcon agrees, saying that "as soon as another DVD-R drive vendor/ manufacturer enters the marketplace, then DICOM's restrictions will be lifted, and they'll be able to officially endorse DVD-R."

But Clunie sees a downside to the circumstances under which other manufacturers will join the market, meaning the advent of version 2, with its split of DVD-R into, at a minimum, 650nm General and 635nm Authoring varieties. "It means that your old drives are history," he says, "and you have to wait until December to buy new drives and new media. This churn, in terms of what there is to standardize on, causes us tremendous problems. For instance, had we picked DVD-R last year, then we wouldn't be able to leverage the consumer format, which has low-cost media and drives, in our products."

Petrocelli, however, sees no reason for version 2.0 to delay DVD-R implementation, because "all media written by either version 1.9 or 2.0 drives at either the 650 or 635nm wavelengths will be readable in any DVD-ROM drive. Different media will be needed for 650 and 635nm recorders, but that's not relevant once the media is written." In other words, media created on existing systems should continue to be readable even as newer systems are introduced using the newer, less expensive drives.

As for DVD-RAM–facing some changes of its own with the advent of 4.7GB-per-side version 2–Petrocelli says he thinks the format has "a smaller role to play, since it is erasable and not as universally readable as DVD-R. We have also found that its sequential read and write rates are rather disappointing for cardiac applications. There is no such thing as 'fast enough' in cardiology."

"DVD-RAM has the advantages of multiple vendors, very low-cost media, and very low-cost drives," Clunie says. "But the disadvantage is that you can't just take out a type-2 cartridge DVD-RAM, stick it in your PC, and play it back on most of the installed base of DVD-ROM players. That may change a bit in the future, as more DVD-ROM drives have the firmware and the hardware to be able to read DVD-RAM, but that's not the current state of the installed base. And that's a big point against DVD-RAM for interchange purposes."


Whatever the relative merits of DVD-R and DVD-RAM for interchange, their suitability for archiving is, as Clunie points out, "a completely separate question. The issues are much more about performance, access, speed, longevity of the media, stability, and physical handling. There are proponents of both DVD-R and DVD-RAM. Both are excellent technologies for the purpose, and I've seen medical imaging archives that use one or the other or both. I kind of think of them as MO replacements that are cheaper and have a longer life in terms of support from the vendors."

Clunie's expectation of longer life reflects the broad support enjoyed by the DVD standard as a family of formats. "Since archive media is not interchanged," Petrocelli says, "it need not be based on a standard. But an open standard is nonetheless superior to a proprietary system because the life of the data often exceeds the life of the image management system. We have been faced with many instances where a customer cannot access their own data without the proprietary system that created it, and that leads to complex and expensive data conversion projects."

In the archiving arena, however, tape jukeboxes become realistically competitive to DVD. Chuck Thomas, general manager of MicroBoards' Medical Data company, points out that while DVD's random access means that information is faster to get to, tape systems support much higher data transfer rates.

Because of this speed, Clunie says, "you can't use anything but tape in certain applications, where you really need the transfer rate. And tape is cheaper by the byte. But most people, when they get serious about a big archive, have a huge RAID that contains all of the immediately current information, and the other media–be it MO, CD-R, DVD-R, DVD-RAM, or some form of tape–is really only used as long-term storage. So both access time and transfer rates are slightly less of an issue. My personal opinion is that I don't really trust tape–I've had too many dead, jammed tapes and lost too many files."

Petrocelli agrees that DVD has the edge as far as reliability. "We used tape when it was the only practical choice for large storage systems," he says, "but anyone who has spent any time at all trying to make a truly robust tape archive system work in cardiology has probably learned that it is extremely problematic. Our service organization spends about 10 times the effort maintaining the DLT-based systems we have in the field compared with the DVD systems. That's why we have been in this market with DVD-R."

While DVD may be a newcomer to medical imaging, Petrocelli is excited about the reaction so far. "Since January 2000," he says, "Heartlab has shipped over 75 terabytes of storage capacity on DVD-R for long-term archiving. The industry can be slow to change, but the customers are not. They have a very high level of interest in DVD technology for image archiving, and I expect that we will see DVD added to the interchange standard when additional vendors begin manufacturing recorders early next year. We believe that this technology will provide the convergence of archival and interchange media."

Companies Mentioned in this Article

Comview Corporation
5870 Stoneridge Drive, Suite One, Pleasanton, CA 94588; 800/444-3022; 925/416-8950; Fax 925/416-8961; http://www.comview.com

Heartlab, Inc.
101 Airport Road, Westerly, RI 02891; 800/959-3205, 401/596-0592; Fax 401/596-8562; info@heartlab.com; http://www.heartlab.com

MicroBoards Technology, Inc.
1721 Lake Drive West, Chanhassen, MN 55317; 612/556-1614; Fax 612/556-1620; http://www.microboards.com

Pioneer New Media Technologies, Inc.
2265 East 220th Street, Long Beach, CA 90810; 310/952-2111; Fax 310/952-2990; http://www.pioneerusa.com

Philip De Lancie (pdel@compuserve.com) is a freelance writer covering media production and distribution technology based in Berkeley, California.

Comments? Email us at letters@onlineinc.com.

Copyright 2000-2001 Online, Inc.
213 Danbury Road, Wilton, Connecticut 06897-4007
203/761-1466, 800/248-8466
Fax 203/761-1444