Next-generation PACS tackles mammography and cardiology

March 2, 2005

The challenges that PACS addresses are not limited to radiology. Other image-producing areas that have been slower to adopt PACS technology, such as mammography and cardiology, can benefit from the same improvements seen in the radiology department through the implementation of PACS for image storage and transmission.

October 2004 - The challenges that PACS addresses are not limited to radiology. Other image-producing areas that have been slower to adopt PACS technology, such as mammography and cardiology, can benefit from the same improvements seen in the radiology department through the implementation of PACS for image storage and transmission. But the method by which a PACS is introduced in these areas should differ from a radiology-based implementation to allow for differences in workflow, application requirements, and regulatory approval of the equipment used. With careful planning, radiology, mammography, and cardiology can all make the transition to a digital image archive and achieve cost savings and productivity improvements. The differences between cardiology and radiology that prevent the effective use of a radiology PACS for cardiology include the technology, applications required, and information systems used in each department. A radiology PACS may integrate through DICOM or HL7 interfaces only. A cardiology PACS must integrate with a number of different systems that may all speak a different language: hemodynamic monitoring systems, ECG machines, and nursing/technologist documentation systems.

The cardiology PACS must also integrate with the system that will feed it patient data: a cardiology information system (CIS). Many departments do not have a CIS in place prior to investigating PACS, and many more use their system only for scheduling or billing rather than creating reports or sending information to the modalities. So it is imperative that the cardiology PACS include these capabilities along with image storage to keep all patient information in one place.

A radiology PACS just may not be capable of meeting all the unique needs of cardiology such as capturing sound, measurements, and M-mode acquisitions from echocardiograph devices or supporting DICOM structured reporting for cath lab procedures. Because of these needs and the inability of most radiology PACS to support them, a cardiology PACS capable of all these requirements may have to be implemented.

MAMMOGRAPHY CHALLENGES Mammography presents its own technical, operational, and regulatory challenges. Unlike cardiology, it may be feasible to extend a radiology-based PACS to include mammography, but certain measures and workflow enhancements should be considered. Mammography storage requirements are the first challenge. Due to the extraordinarily high resolution of mammography images, digital mammogram files can exceed 150 MB per study. Although this is not necessarily larger than a cardiac cath study, which can approach 1 GB of data, it is likely that a single department will perform many more mammograms than catheterizations in a single year and will review prior mammograms much more often than prior catheterizations. This impact on a radiology PACS archive and network must be taken into account before purchase, or the archive could be monopolized by mammography studies and the network brought to a crawl. Mammography also involves a regulatory concern, since few PACS are FDA-approved for interpretation of mammograms. In most cases, an independent review workstation is supplied with a digital mammography system to serve these needs, but this solution presents technical issues of its own. This workstation must be integrated with the other systems that PACS is tied to, including the RIS and voice recognition. In many instances, a separate voice recognition workstation needs to be implemented in mammography, with the workflow adjusted for these changes.

Computer-aided detection throws another wrench in the workflow. CAD is often integrated with the mammography review workstation but not necessarily with the PACS. The ability to send images and the CAD interpretation to the PACS may require an additional interface if the mammography and CAD markings are to be saved.

By overcoming these speedbumps, a department may successfully integrate mammography into a functional PACS environment.

FUTURE POTENTIAL The future may see a single system with a shared database across multiple -ologies but with unique front-end applications that give radiologists, cardiologists, mammography specialists, pathologists, and endoscopists the individual tools they need to optimally interpret their studies. The advantages of such a system would be cost savings and data consistency. Cost savings would come from sharing archive space as well as servers and possibly interfaces, depending on what other clinical systems are in use in mammography, cardiology, and beyond. Maintenance and support of the equipment may potentially reduce costs more, since less hardware purchased means less hardware to maintain. Aside from cost reduction, a system with a single database would be advantageous from a data consistency perspective and with regard to speed.

A cardiologist wanting to view an MR angiography study stored on a separate radiology PACS from a cardiology PACS workstation will likely have to initiate a DICOM query to the radiology PACS. This process is going to be much slower than pulling from cardiology's own database, even if the two systems share the archive on a storage area network. Additionally, the patient could be registered in the cardiology PACS under a different spelling or nickname from that in the radiology PACS record, introducing potential problems in querying that patient's records.

While the perfect system may be a fantasy, its potential benefits need not be. Radiology, cardiology, and mammography PACS, although purchased separately, can share archive space on a SAN and a long-term archive, thereby saving hardware and maintenance expenses. Open-standard-based systems can integrate with one another to provide cross-functional checking for consistency of data, and Web-based systems can improve the speed by which information from each PACS can be accessed and viewed. By taking into account the various challenges that each -ology is presented with and acquiring a system that meets not only the departmental workflow but the greater needs of the institution at large, decision makers can attain many of the advantages of a future system today.

Mr. Meitzner is an associate with the Health Systems Group at ECRI, specializing in PACS planning, procurement, and implementation.