Synchronizing study data changes between spoke sites and IDC: Concepts

Cases where data must be synchronized

The spokes usually send only reported and approved studies to a data center. (Note that even if unreported studies are sent to a data center, and then subsequently sent again as reported, any new objects in the study are added to the original study.) This limits the situations in which studies must be re-archived. The following types of study updates are normally performed before studies are reported and approved. However, in cases where they are performed on approved studies, data synchronization is required.

1. Some images from one study are split into another study for another patient

Splitting a study means moving some of the images from one study to another study. Split is typically used when images from one patient are mistakenly associated with a study for another patient. Selected images are moved and all others remain with the original study.

When a study is split, the moved objects are assigned new study instance UIDs and possibly new series instance UIDs. The original study keeps the original UIDs.

2. Two or more studies are merged into a single study

In a merge, two or more studies are combined into a single study. When studies are merged, the objects (images, series) that belong to the source studies are moved into the destination study. The moved objects are assigned new SOP instance UIDs. The spoke site stores these newly created objects to the data center.

3. One study is segmented into multiple studies for one patient

Segmenting a study means breaking or copying a single study into multiple studies for one patient. This involves selecting the images to be associated with each order, and may require duplicating some images. This could be necessary, for example, if three studies were ordered, but only one study was performed at the modality.

4. A DICOM attribute of an object changes

Study information attributes such as patient name or study ID may be corrected at the spoke.

5. An object is deleted

A study can have objects (series, images, presentation states, or markups) deleted from it at the spoke.

6. A study is fixed such that its UID is changed

An unverified study is an incoming study that does not match the information from the HIS. These studies are fixed by associating them with an order, and updating the study attributes accordingly. If the study UID changes as a result, this change needs to be synchronized.

Data synchronization methods

Newly created objects, such as those created by a split, are archived to the data center. Most other types of changes, such as object deletion and DICOM attribute changes, are handled through change context (cc) objects. IMPAX Data Center reads the changes in the change control object and applies them to the database. No changes are applied to DICOM objects stored in the online cache or nearline archive volumes. However, any changes made are applied when studies are transmitted at time of retrieval.

Study fixups that cause the study UID to change also require manual steps at the spoke site and at the data center to keep the spoke and data center in sync. For this type of change, this is true of both IMPAX sites (any version) and third-party PACS sites.

New objects or presentation states added to a study

A study may be changed at the spoke so that it now contains new objects in the same study, such as annotations or markups that are added as new presentation states, or secondary capture objects from 3D reconstructions. (For spokes using older IMPAX systems, objects such as markups or annotations are stored in the local IMPAX database and no new objects are created.)

The new objects follow the normal archive procedure as for any other new object, and are archived to the data center.

When IMPAX Data Center receives the new objects, the application associates them with an existing study, and stores them to a nearline archive volume in a new TAR file. Note that the images are grouped in a series-level TAR file. If the series has already been archived before the new objects were added, a new tar file for the series is created.

When a spoke requests a study that has new objects, IMPAX Data Center retrieves all instances of the study. IMPAX 6.5.1 systems support the publishing of presentation states.

Ownership of synchronized data

The originating spoke is the owner of its own data and therefore should be the only one to change the original study data (although this ownership is not enforced). And, when new objects are generated—such as when new presentation states or 3D reconstruction views are created—the spoke that created the new objects owns them.

As well, IMPAX Data Center does not broadcast any updates that it receives. For example, site B has recalled (and is storing locally) a study from the data center that belongs to site A. When site A makes an update to the study, the update is synchronized between site A and the data center, but the data center does not broadcast the update to site B.