↓ Skip to main content

Agfa HeartLab: Performance Issues or Normal Behavior?

·364 words·2 mins
John Sole
Author
John Sole
I help health systems implement imaging systems and fit them into the workflows already in place. DICOM, HL7, IHE, PACS and VNA, with SQL and Python as needed.

Originally published on JS PACS, May 22, 2009.

I typically don’t like to discuss vendor-specific problems in front-page articles, but I’m going to break that rule today. I’m hoping this article reaches many people, but specifically cardiology PACS administrators who manage the Agfa HeartLab cardiology PACS. I’m currently managing this C-PACS and find some recurring behaviors that are not only problematic but result in significant manual work to keep the system in a usable state.

What has been occurring in my implementation is routine loss of the DICOM services. If you’re familiar with Agfa as a cardiology solution, you’ll know that they run objects called DICOM Labs. Each DICOM Lab is responsible for image transfer from one particular imaging device. We have customized HeartLab in that we don’t store directly to the cardiology PACS; we store to the radiology PACS (which contains our enterprise archive) and forward a copy to the cardiology PACS for viewing. Since we only have one device sending to the cardiology PACS, we only have one DICOM Lab, and one would think that without the typical 10 or 15 DICOM Labs running simultaneously we would see high levels of performance. We do not.

The most frustrating problem I have with this system is that every so often (20 to 24 hours) the DICOM services hang on the server. There is really no indication that the services have hung, other than that an occasional spot check of our archive’s auto-router shows anywhere from 1,500 to 2,000 images pending transfer to HeartLab. Usually, logging on to the server and performing a stop and start of the DICOM services kicks all the processes back into motion. Sometimes, however, a complete reboot of the C-PACS server is necessary. While the resolution is manual, it is not overly lengthy; but the fact that it has to occur every day is unacceptable. Between the time the PACS administrator spends fixing the problem and the time lost by cardiologists waiting for image transfers, this cardiology solution significantly slows down workflow throughout the day.

I’m curious to know whether other cardiology PACS administrators experience similar issues, specifically with the Agfa solution, and what types of resolutions you’ve developed.