Although DVD storage and viewing systems for large imaging studies are still sitting on the back burner (pun intended), the industry hasn't yet reached the tipping point. Meanwhile, the use of CDs to transfer patient images can be troublesome since most studies require disk spanning - several CDs to cover one study, and a lack of industry standards means physicians often can't even retrieve CD images through their viewing system, rendering them useless.

Several CD burner vendors use proprietary viewers that store data in something other than DICOM standards, says Chuck Socia, vice president of operations for Arkansas Medical Imaging in Little Rock. In some cases, the images are stored as jpg files. Some PACS allow the importing of jpg files into a DICOM file - but that requires a fair amount of extra work, Socia says. Other vendors have a format that isn't viewable on anything other than their own viewer.

The solution to the problem is an industry-wide switch to Integrating the Healthcare Enterprise (IHE) protocols so everything is under a standard format, says Socia.

However, few customers are requiring IHE protocols so few vendors are offering them. "Until they do, there's nothing we can do to standardize," he says.

The solution to the CD problem is going to be using the PDI (Portable Data for Imaging) protocols from IHE so everything is under a standard format. Even if that happens, CDs only hold so much and scanners keep producing larger and larger data sets.

Complicating Socia's role is that physicians often don't understand the technology. They want to pull a study into their PACS so they can compare the old study to the new study and get like interpretation. If they don't understand why that might not work, nothing is going to change, Socia says. "Physicians are making demands, but PACS administrators and qualified IT people are not making decisions on what's being purchased," he says.

Automatic breaks and destination

"Large studies have always been a challenge," says Doug Ricci, PACS manager for the department of radiology at Johns Hopkins Hospital in Baltimore. Fortunately, he only needs to publish exams to CD in cases where a physician at another institution wants to view the study. Typically, it's a neuroradiology study that requires the special treatment.

Ricci uses software from ODS Medical to burn studies onto CDs and has worked with the company to help enhance the software's special features. For example, the software automatically determines whether the study should go to CD or DVD, based on its size. It also automatically breaks the study at appropriate points. "ODS is working with us for new ways to send these exams to our equipment." In fact, it was at Ricci's suggestion that ODS developed a feature that allows a radiologic technologist standing at the CT scanner to press a "destination" button on the scanner to send the exam to a CD or multiple CDs for "hands-free" publishing. A destination on the scanner called "autoburn" lets the technologist select a series and simply pick up the CD five minutes later.

Although many of the PACS in place have the ability to store large studies online, the data sets are only getting larger, which become cumbersome to manage and slow systems down. Many CT and MR studies now require 3D imaging which require two or three versions of a study going from CD into a 3D workstation. Most portable media devices were put on the market before studies became as large as they often are today. And once a study is over 650 megabytes, it needs to go to CD or DVD. DVD systems are more expensive and most physicians don't have DVD viewers. So, the other option is disk-spanning which can be cumbersome to look at. Different disks will contain different types of views and it's not easy to toggle between the two.

Image size, number will only grow

Phil Larkin, PACS manager for Ohio State University, works in a completely filmless department that typically puts 100 studies on CD a day and will probably conduct 400,000 studies this year, generating 35 million images.

Despite the problems with study size, Larkin says the industry won't switch to DVD until physicians decide it's important enough. "It's not worth putting studies on DVD so somebody can come back and complain" that it was only viewable on their home DVD player, he points out. Meanwhile, physicians are struggling with CDs. For example, "there's no point putting 3,500 images on a CD when they can only look at 200 anyway," Larkin says. "With functional MR and those kinds of things coming down the pike, those are 7,000 images."

Working with software from Sorna Corp., Larkin can construct multipatient CDs without "consciously going to another application to put more than one patient on one CD." The software generates labels. The first five exams, listing patient name, medical record number, and date of birth, go right on the label. Since one CD can contain up to 65 chest x-rays, the CD itself includes a list detailing all the studies on it.

Larkin's physicians complain that viewing images is too slow, although not as slow as a film library. Aside from the ability to avoid splitting studies, he looks forward to switching to DVD as a solution to catastrophic downtime. DVD "would allow us to archive to DVD as a temporary solution. We can archive more data on DVD which means less cleanup work on the back end." However, Larkin doesn't see this coming any time soon "unless the PC vendors start putting DVD hardware in everything they sell."