Re-send after correction
Send a corrected study back to a PACS as a new study beside the original, or replace it through IOCM where the archive supports it.
Desktop app only, 1.6.0 or newerdcmage pro
Problem
You opened a study from a PACS and corrected its metadata. DIMSE and DICOMweb can add images to an archive but have no way to overwrite them, so the correction can’t simply replace the original. It goes back as a new study, and someone has to remove the old one afterwards: you, in the PACS, or the archive itself when it acts on an IOCM rejection note.
Steps
-
Open the change preview with Send to archive…. It lists every change, as it does for an Orthanc commit. Over DIMSE and DICOMweb the button is Store corrected copy. On Orthanc you choose Store corrected copy instead of the default Commit correction when you want the original kept.
-
Check both identifiers. The preview shows the original study’s ID (its Study Instance UID) and the new ID the copy will get. It also says the original stays in the archive, so the PACS will hold two versions of the study until you delete one.
-
Tick the confirmation, then store. The button stays disabled until you tick the box. dcmage writes your changes into every image, gives the study, each series and each image a new ID, and sends the copy in batches. Cancel store stops the send between batches.
-
Read the result. Once the archive has accepted every image, dcmage searches the archive for the copy and downloads one image to check your values arrived.
Result
You get one of three results, each showing the copy’s new study ID:
- A corrected copy is in the archive. The archive accepted every image, a search finds them all, and the downloaded image carries every value you approved. The result reminds you the original is still there: delete it in the PACS once you have checked the copy.
- Accepted, but not confirmed. The archive took every image, but the search didn’t find the copy yet, or a value came back changed.
- Partly stored. The result says how many images the archive confirmed. Send the rest finishes the same copy under the same IDs and sends only what is missing. If you have changed your edits since, dcmage refuses, because that would build one copy out of two different corrections. Delete the incomplete copy in the PACS instead. Send the rest also goes away once you disconnect, because the archive only takes the rest of a copy from the connection that started it. Delete the incomplete copy in the PACS, then open the study again and repeat the correction.
A copy sends your tag changes only. If you changed the study’s pixels or its compression, or drew a redaction box, the preview refuses the copy and says so in place of the button. Undo the change, or delete the box in Anon mode, or export the file instead. A copy already started stops the same way: Send the rest returns once the change is undone.
dcmage won’t send a copy if the archive already holds a study with the copy’s new ID, so a copy never adds images to an existing study. An Orthanc connection can also replace the study, with no duplicate left.
Replace original
Some archives, dcm4chee among them, act on an IOCM rejection note: they stop listing the images the note names. If yours does, tick This archive removes images when it receives a rejection note for them when you connect over DIMSE or DICOMweb. dcmage remembers it with the archive. The preview then offers Replace original beside Store corrected copy.
dcmage stores and checks the copy exactly as above. Only when the copy is confirmed does it send a rejection note for the original (Incorrect Modality Worklist Entry), and then it searches the original study again:
- Replaced. The archive holds the note and lists none of the original images. There is nothing left to delete.
- Not replaced. The archive holds the note but still lists the original. It may not act on notes, or may act later. Check again searches first and sends the note again only if the archive has no record of it.
- Refused. The archive, or dcmage itself, refused the note. dcmage refuses when the original study changed since you opened it, or when the copy and the original are filed under different patients. The original stays; delete it in the PACS.
- Not confirmed. The archive’s answer to the note was lost, or the search after it failed. Check again finds out. After a disconnect it only searches: the archive only takes the note from the connection that stored the copy. If it has no record of the note, delete the original in the PACS once you have checked the copy.
A copy that was not confirmed is never followed by a note, so the original stays untouched.