PACS and DICOM Workflow
PACS issues often feel mysterious because images pass through several systems before a study is complete. A workflow view makes troubleshooting much easier.
The usual chain
- A modality creates images and study metadata.
- DICOM data is sent to one or more destinations.
- Intermediate systems may route, normalize, cache, or reconcile data.
- PACS stores the study and serves it to viewers.
- Other systems such as RIS, reporting, teaching files, or cloud archives may depend on the same identifiers.
Systems commonly involved
- Modalities and modality worklists
- RIS or scheduling systems
- DICOM routers or gateways
- PACS archive and viewer platforms
- Identity, networking, and storage infrastructure
Questions that quickly narrow the problem
- Did the modality create the study correctly?
- Did the transmission leave the modality?
- Did a router reject, hold, or rewrite the data?
- Did PACS ingest the study but fail to display it as expected?
- Is the problem tied to one modality, one AE Title, one site, or one date range?
Failure patterns worth recognizing
- Worklist issues usually show up before image acquisition
- Routing or networking issues often affect transmission timing or destination reachability
- Identifier mismatches can create split exams, duplicates, or missing priors
- Storage or viewer issues may appear only after the study has already arrived somewhere downstream
A practical engineering checklist
- Capture AE Titles, IP addresses, ports, and timestamps
- Preserve at least one example accession number or study UID
- Check whether the issue affects current exams only or historical retrieval too
- Confirm whether images are missing entirely or present in one system but not another
- Write down the first known good time and first known bad time