Early PACS workstations were composed of Unix-based hardware that had to be purchased from the PACS vendor and was expensive to maintain. Replacement parts were not available off the shelf and also had to be ordered through the PACS vendor. In contrast,
Early PACS workstations were composed of Unix-based hardware that had to be purchased from the PACS vendor and was expensive to maintain. Replacement parts were not available off the shelf and also had to be ordered through the PACS vendor. In contrast, most PACS workstation software today runs on standard Windows-based PCs, which hospitals can purchase directly at far lower prices, and replacement parts are relatively easy to obtain.
Although PACS vendors will usually sell the workstation hardware to the customer if the customer prefers, they will also provide guidelines for customers who wish to reduce costs by purchasing it themselves. Except for the display system, which needs a separate article, the basic feature set needed for PACS workstation hardware can be fairly inexpensive and summarized in a short review.
The value proposition for purchasing workstation hardware is strongly affected by the type of licensing agreement. The original cost model for PACS workstations bundled together the cost of the hardware and the cost of the license for the software, creating a range of $40,000 to $100,000 per workstation. As hospitals realized that they could save money by buying the hardware themselves, two changes occurred. First, the cost of the hardware was separated from that of the software, as software-only licenses could be purchased from the PACS vendors. Second, and more important, the hospitals realized that it is more logical to pay for a license for a user of the system than for the workstation. In other words, the software cost could be based on the number of concurrent users of the system rather than the number of workstations. This meant that the cost of adding another workstation to the system would equal the price of the hardware only, which can be as little as $2300.
In fact, some PACS vendors are beginning to do away with charges for the workstation software altogether and just absorb the cost into the yearly service contract based on volume. They realize that the number of workstations that will require support, as well as the overall need for support, is related to the yearly volume of studies. The greater the volume of studies, the greater the number of workstations required to read them and modalities to create them, and the greater the amount of storage and support required. If the yearly service contract price is based on the number of services purchased from the vendor and the yearly study volume, then there's no need to even charge for concurrent user licenses. This model is the best of all, as it avoids the headache of trying to figure out how many licenses are needed and how to monitor compliance with the license limit.
The PC platform is familiar to end users throughout the hospital, not just to the IT staff that needs to support the computers. Another advantage is that excess or moderately aged PC hardware can be reallocated as regular desktop computers, whereas Unix-based hardware can rarely find another use.
Vendors started providing PC-based software only about three years ago. Today, almost all PACS vendors support PC-based workstations, but some continue to offer Unix-based workstations. We made the transition to PCs at Massachusetts General Hospital about a year and a half ago.
COST VARIES
A high-end PC without monitors costs about $1000. More expensive workstations are available, including some with dual processors, but it is unclear how much additional benefit they confer. Typical computer applications will not take advantage of a second processor unless specifically written to do so, and most PACS software is not written that way. It is possible, however, that a dual-processor workstation will lead to a real performance gain when the user is trying to run on the same PC two resource-intensive applications, such as the PACS workstation software and speech recognition software.
In 2003, the processor's speed should be between 2.5 GHz and 3 GHz, and it may come from manufacturers such as Intel, AMD, or Cyrix. A gigabyte of RAM is usually sufficient. Although many machines can handle up to 4 GB, more than 1 GB is usually not necessary unless the machine is to be used for 3D processing as well.
If you anticipate that the memory may be upgraded, make sure additional memory slots remain available. If the machine has only two memory slots and both are filled initially, the only way to upgrade is by removing the existing memory chips, which is a waste if they can't be reused elsewhere. Bear in mind that most machines will undergo only a single memory upgrade, if any, in their lifetime. By the time a second upgrade is required, it will be time to replace the machine.
Most hospitals have either 10- or 100-Mb connections to the network available. A 100-Mb connection is preferable. Gigabit Ethernet connections are generally reserved for the network backbone. It is very important that the workstation's network card be configured properly, as not all cards automatically adjust correctly to the speed of the connection. This can have a major impact on the performance of a system but can be fixed with relative ease.
Two popular chassis models are the small form factor and the minitower. True to its name, the small form factor takes up less space, but it has some limitations. Some will accommodate only half-height, or "low-profile," videocards. They tend to have fewer PCI slots, which limits the ability to add videocards to support additional monitors. The greater flexibility of minitowers, therefore, usually makes them better suited for PACS workstations. Most workstations are physically situated in a computer desk that has ample room for the computer and monitors.
The most costly part of the workstation hardware is the display system, which includes the videocard(s), cables, and the monitors themselves. Since this is such a broad topic, it will be the subject of a future article.
The computer part of the PACS workstation, now a commodity to be had for about $1000, can be bought by hospitals directly from computer manufacturers, thereby avoiding vendor markup. These computers are so cheap that it is reasonable to keep entire spare systems in a closet so that any in need of repair can just be swapped out until they are fixed. There's no need to worry about purchasing a few too many computers-those not needed for PACS workstations can be easily reallocated to desktops for use by staff. Spare parts are readily available, and most hospital IT departments are well equipped to maintain PCs.
All of these factors make the process of purchasing PACS workstation hardware less expensive, less risky, and less daunting. Provided the licensing is negotiated properly, hospitals that could only afford 10 workstations in the early days of PACS can now afford to scatter more than 100 of them throughout the healthcare enterprise. This makes diagnostic-quality images that much more accessible to healthcare providers and will likely translate ultimately into improved patient care.
Dr. Hirschorn is a clinical fellow in radiology at Massachusetts General Hospital and Harvard Medical School. Mr. Schultz is senior programming consultant in radiology at MGH and for Partners Healthcare. Dr. Dreyer is vice chair of radiology information services at MGH and an assistant professor of radiology at Harvard Medical School.
Study Reaffirms Low Risk for csPCa with Biopsy Omission After Negative Prostate MRI
December 19th 2024In a new study involving nearly 600 biopsy-naïve men, researchers found that only 4 percent of those with negative prostate MRI had clinically significant prostate cancer after three years of active monitoring.
Study Examines Impact of Deep Learning on Fast MRI Protocols for Knee Pain
December 17th 2024Ten-minute and five-minute knee MRI exams with compressed sequences facilitated by deep learning offered nearly equivalent sensitivity and specificity as an 18-minute conventional MRI knee exam, according to research presented recently at the RSNA conference.
Can Radiomics Bolster Low-Dose CT Prognostic Assessment for High-Risk Lung Adenocarcinoma?
December 16th 2024A CT-based radiomic model offered over 10 percent higher specificity and positive predictive value for high-risk lung adenocarcinoma in comparison to a radiographic model, according to external validation testing in a recent study.