Over the past decade, picture archiving and communications systems (PACS) have matured to a point at which they are becoming essential to the operation of modern radiology departments. If, however, one looks at the total number of installed PACS around the country, one sees that most radiology departments still use film for one or more reasons. Next-generation PACS are looking to sweep away the remaining issues that prevent radiology departments from reaching filmless operation, a key plateau in establishing the full financial and operational value of these systems.

One of the most important issues in a move to filmless operation is that the PACS becomes a mission-critical information system. It must work, and the stored data must be reliably accessible. PACS technology has always been designed and engineered with mission-criticality in mind. However, next-generation PACS products will feature many enhancements to raise the standards of reliability, data availability, and maintainability.


Many radiology departments and imaging centers are likely to experience substantial growth in the years ahead through a combination of aggressive marketing outreach, acquisitions, and consolidations. In addition, the amount of data being generated by new imaging modalities is growing significantly. It is imperative, then, that PACS be more fluidly able to grow in tandem with the enterprise. PACS will need to be highly scalable, and the next-generation systems will be precisely that.

Software maintenance will require a more aggressive approach in the future in order to ensure not only that the system stays up and running, but that it runs in optimal fashion. Updates and upgrades become more challenging on a mission-critical system. The reason for this increasing difficulty is that one has to perform upgrades and updates without incurring any significant downtime. One needs ways of being able to do that in parallel with a system that is operational, as the new PACS models will now permit.

The next generation of PACS will deliver image files more quickly than ever before. That will prove invaluable, given that the number of users on a single system is likely to be many times greater than that which is typical today. One of the challenges is that PACS, just like streets, have rush hours in which traffic becomes congested and, with enough users, snarled. Unfortunately, in a mission-critical application, one cannot shunt users to back roads. One needs to be able to handle the bursts of high activity whenever they happen. More important, one needs the experience to model the PACS activity accurately so that there are no surprises. The good news is that the next generation of PACS, in combination with experienced vendors, will be able to address this challenge.

Since PACS changes the way people work within an enterprise, it will be more essential than ever to have well-trained users. The PACS installations of tomorrow will not only be more user friendly, they will also better facilitate the training cycle. The differences in work processes between film-based and filmless environments are significant, and they can disrupt a department’s operations if the users of the system are not up to speed. The training does not end after initial PACS installation, either. There will probably be minor updates several times per year and a major upgrade every year or two. Users need to be trained each time those changes are made. It is worth noting that a majority of the calls we receive at our support centers are from PACS owners and users who need application help, not those who call to report a problem. This suggests to us just how crucially important user training is.

System Support tools

Next-generation PACS will include a more comprehensive and sophisticated array of support tools to provide operators better information about what the system is doing and how it is being done. Mission-critical information systems need to provide system administrators with a detailed view into the system’s operation so that corrective actions can be taken before performance has been severely impacted. Such information might include a log of database transactions, data-cache activity, and connectivity problems with imaging modalities or with other information systems. Good diagnostic information also allows system administrators to identify and address difficult-to-pinpoint problems: for example, a noisy or intermittent network connection that is causing a large number of communication retries. Other problems that would be easier to uncover include the discovery that some users of the system are executing certain tasks improperly (which could be resolved with additional or remedial training) and that there is insufficient system capacity available. The key to being able to identify any of these problems is to have good information to start with, and that is one area where the next generation of PACS will really prove its worth.

The sharing of images and information among hospitals, clinics, departments, and departmental work groups is becoming increasingly common, and the next generation of PACS will be equipped to abet pooling and collaboration of a grander magnitude. Such sharing might occur between a diagnostic radiology operation and referring physicians from another department or enterprise. It might occur between two different radiology departments. Multiple clinical departments (such as cardiology, pharmacology, and radiology) might also want to integrate their information to form a single report to be used by a primary care physician. Handling the connectivity issues between multiple operations will be a key feature of next-generation PACS. The challenge faced by PACS owners is trying to integrate operations that do not work the same way. The issue can be as basic as the assignment of a patient identification (ID): the patient going into the clinic may be assigned a different ID when later going into the hospital, but it must be possible, somehow, to take the two records being built up on that patient and make the system understand that they are records for the same individual. What, in practice, has happened in many cases is that the hospital has built up its PACS solution and the clinic has separately built up one of its own, so that when one attempts to integrate, one actually has two different databases running and two different systems overall.

With integration will come, of course, challenges of making the connectivity protocols work. I believe that both the Digital Imaging and Communications in Medicine (DICOM) and Health Level 7 (HL-7) communications standards will be supported concurrently, and perhaps universally, by next-generation PACS. In fact, they will coexist. Already, we see DICOM extending to manage some of the types of data that are, at present, managed under HL-7. As a good example of how they coexist, I would point to the Integrating the Healthcare Enterprise initiative, in which the functional centers of activity (actors) associated with a PACS or radiology information system environment interact with other actors by using a combination of HL-7 and DICOM protocols. It is a very pragmatic view of how to make things work. That will probably continue for quite some time.

It should be noted that in a truly filmless environment, the electronic distribution of images and information to the users of this information (specialty physicians and primary care physicians) must be simple, fast, and secure. The solution must also allow for scalable implementations that can grow to support many users. The wider the area for the distribution of images and information, the more intensified the need for stringent system security. Forthcoming PACS technology will enable owners to enjoy greater control over who is permitted to access the system and to prevent unauthorized access. Indeed, next-generation PACS will utilize Internet protocols as never before for low-cost network connections and, increasingly, to partake of services such as off-site storage management and application hosting. As PACS makes use of telecommunications channels that can be hacked, however, security issues will become more pronounced. What we can expect are PACS with stronger and more sophisticated firewalls, password-security controls (which might well involve use of biometrics as a means to provide log-in control), and much more powerful data encryption-decryption technology.


The importance of Internet technology to next-generation PACS cannot be overstated; of that, I am quite certain. I expect that PACS will operate more quickly, more cheaply, and more securely, thanks to emerging advances in the Internet itself. Even now, we are witnessing progress in making available to the public dramatically increased bandwidth at reasonable prices. This should provide greater opportunities for PACS users to transfer the large data blocks associated with PACS more economically at greater speeds.

Internet technology, coupled with next-generation PACS technology, will allow system owners to manage data better by allowing data to be processed and stored centrally. For most users, this is a key step toward a practical implementation of a mission-critical information system. Most of the hard issues associated with managing a mission-critical information system revolve around protecting the data and making them reliably available. By outsourcing this task to a central data center with the specialized skills to deal with these issues, a mission-critical data management solution is within the reach of every PACS user. The actual ability to put the PACS data at a central location is a reality and, within 5 years, could be quite commonplace.

PACS in the future will also rely more extensively on so-called thin-client technology, in which the PACS application is stored and managed at a central location. This has as its primary advantage the ability to maintain the system better and more easily. There is simpler technology at the remote or individual client sites, so there is less system management required at that level. This does not necessarily make the system less expensive?although it could. What it really does is make the system more mission critical, and makes it more available to the individual users.

My view is that the economics of the next generation of PACS will be influenced significantly by matters related to telecommunications, such as the cost of bandwidth. Obviously, the amount of data being transferred to and fro in a PACS is potentially quite high. Compression of data helps reduce the size of the data packages, but significant bandwidth will still be required to achieve acceptable performance. Fortunately, the telecommunications costs are coming down, which will help make tomorrow’s application service provider (ASP)-based PACS implementations economically feasible for far more enterprises.

Still, I am of the opinion that the economic advantages of PACS will not materialize until the department, clinic, center, or hospital operates in a truly filmless fashion, which means that issues associated with mission-criticality are completely addressed. The economic benefits that PACS promises derive mainly from reduction or elimination of costs stemming from image storage and distribution, but also from the accelerated delivery of images and reports to the referring physician.


Agfa’s IMPAX? Solutions has been hard at work developing the innovations that will support current and prospective PACS owners in an environment calling for mission-critical systems. From us, customers will see more and better solutions to those mission-critical issues. Specifically, that means higher-performance systems and larger implementations of reliable Web distribution solutions. Our service and maintenance tools will be all the more comprehensive and sophisticated. Our products already allow much of the application or storage capability to be deposited in a central location, and we will continue to improve the functionality of our solutions when used in this manner.

The ASP model of PACS will become a cardinal element in Agfa’s offerings. ASP is, in reality, a means of providing existing IMPAX functionality using a centralized storage and application management solution. By allowing PACS owners to shift most of the mission-critical data-management chores to a central location and by using highly trained information-management people to keep the system up and available, the ASP architecture provides a way to store data cost-effectively and very efficiently (while, at the same time, maintaining their security, reliability, and availability). It is, therefore, a vital tool for achieving the goal of making a PACS solution mission critical. The ASP architecture is also a very attractive option for smaller sites that cannot afford that initial investment in PACS. An ASP operational model allows users to pay on a per-use basis. Enterprises that have difficulty justifying the initial investment to implement a PACS solutions will, therefore, find an ASP operational model quite an attractive proposition.

Another Agfa/IMPAX Solutions innovation will be enhanced workstations that provide the targeted functionality required within specific medical disciplines. It is not enough to generate and store diagnostic image information; Agfa believes that it is essential that this information be provided to clinical physicians, along with the tools that help the physicians work with it effectively.


I believe strongly that the most successful PACS installations of the future will be those in which the system owners use vendors that understand both earlier and next-generation PACS at the deepest levels. For PACS to become mission critical, system owners must work with vendors that have the ability to avoid the configuration issues that can make a system run suboptimally and that can prove difficult to remedy. Training programs must be first rate, based upon experience. No matter how PACS is approached from here on, however, mission criticality will be the defining issue for the next generation of systems.

Carter H. Yates is global product manager, IMPAX? Solutions, Agfa Corp, Greenville, SC. The Agfa ASP is a works-in-progress not currently available for sale in the United States.

Related Links: