Using DICOM C-FIND and C-MOVE for Controlled PACS Migration

The DICOM Query/Retrieve migration workflow

A standards-based PACS migration commonly uses DICOM Query/Retrieve to discover studies and instruct a source archive to send them to a configured destination. C-FIND discovers matching records. C-MOVE requests delivery to a named destination AE Title. The source then opens one or more C-STORE associations to transmit objects to that destination. These are separate operations and must be monitored separately.

Use C-FIND to build the migration inventory

At Study Root level, C-FIND can query by patient, date, accession, modality and other supported keys and return identifiers such as Study Instance UID. The exact keys and matching behaviour depend on the source implementation. Store the query criteria and response data so the inventory can be reproduced and reconciled later.

• Test broad and narrow query behaviour before production discovery.
• Use stable date or facility partitions for large archives.
• Preserve Study Instance UID as the primary study key.
• Detect repeated responses, incomplete attributes and query timeouts.

Understand what C-MOVE confirms

A C-MOVE request tells the source PACS which destination AE Title should receive the matching objects. The source reports sub-operation status such as completed, failed, warning and remaining. A success response is valuable operational evidence, but it should not be treated as final proof that the destination retained a complete, clinically usable study.

Configure AE routing before migration

The source must know the destination AE Title, host and port, and the destination must accept the source calling identity and object types. Firewalls must allow the actual C-STORE route, which may differ from the connection used by the migration controller.

• Migration controller to source: C-ECHO, C-FIND and C-MOVE.
• Source archive to destination: C-STORE associations.
• Correct called and calling AE Titles at each hop.
• Supported presentation contexts and transfer syntaxes.
• Network address translation, firewall and idle-timeout behaviour.

Control workload and clinical impact

Large C-MOVE workloads can compete with clinical retrieval, storage and network traffic. Use bounded batches, defined schedules and concurrency limits. Measure practical throughput with representative studies, then adjust without assuming that maximum transfer speed is the safest project speed.

• Partition work by a stable, non-overlapping rule.
• Limit concurrent requests and associations.
• Pause or slow migration during operational stress.
• Record request, response and destination-verification timestamps.

Track study-level status and evidence

Create an operational record for every inventory item. Keep the request identifier, current state, attempt count, DICOM status, failure detail and destination verification result. This lets teams stop, resume or retry selected work without losing accountability.

Retry with diagnosis, not repetition

Before retrying, decide whether the failure was transient or deterministic. A timeout may justify another attempt after infrastructure recovery. An unknown destination AE, unsupported object or rejected presentation context requires correction first. Uncontrolled retry loops can produce duplicate load and hide the real fault.

• Retain every attempt rather than overwriting history.
• Apply a maximum retry policy with visible exhaustion state.
• Retry selected studies after remediation.
• Keep manual stop, skip and resume actions auditable.

Validate independently at the destination

Use destination queries, database reports or supported APIs to confirm receipt, then reconcile identifiers with the source inventory. For higher-risk migrations, compare series or instance counts and validate representative studies in the clinical viewer. This closes the gap between protocol success and accepted migration outcome.