Skip to content

Prior studies

When a new study arrives on a route with Prior studies on, the router asks the archives for the patient’s earlier studies, and sends the relevant ones to the route’s destinations (or chosen ones), so the radiologist has them when the new study is read. Configuration › Routes › Prior studies.

The archives are searched by Patient ID, and the most recent matching studies are fetched:

Setting Default
Most recent studies 3
Not older than (years) 5 (0 = any age)
Only studies with the same modality off
Only studies of the same body part (and not of the other side) off
Also studies whose body part cannot be told on
Archives all of them
Send them to the route’s destinations

The body part of a study is read from keywords in its descriptions: body part examined, study, series, protocol and procedure descriptions, whole words, any case. A chest CT’s priors are its chest studies; a knee MRI’s are knee studies.

  • When a prior’s study description does not say, its series are asked about (one series-level search of its archive).
  • Priors of the other side (LEFT/RIGHT, LT/RT) are left out: a left knee does not bring the right knee.
  • Whole-body studies are relevant to any body part, and any body part to them.
  • A new study whose body part cannot be told brings priors of any body part.

The keywords are under Configuration › Settings › Relevant prior studies: a built-in list of eleven body parts (head, neck, chest, heart, breast, abdomen, pelvis, spine, upper limb, lower limb, whole body) that you can copy and edit, one line per body part: Name: KEYWORD, KEYWORD. A name ending in * makes its studies relevant to any body part.

  • Retrieved through the router, priors are queued like any instance and show in the history as received from the archive, with source Prior studies. Archives that retrieve by C-MOVE send them straight to each destination.
  • Each new study is looked at once per node (again after 12 hours), however many instances it has.
  • Requests are kept on the node’s disk until done, so a restart resumes them. When the archives cannot be reached, a request is retried after 1, 2, 4… minutes (at most an hour apart) and given up after a day.

Lists the requests: waiting, retrying (with the next attempt and last error), done (priors fetched of those found), given up or cancelled. Administrators can retry requests (including those given up; priors already fetched are skipped) or cancel those still waiting. The list names patients, so it is for history users, and viewing it is audited.