I have an odd problem I’m trying to work out.
One of my users reported that some images were not visible in the Orthanc browser via the web. I confirmed that the images will not come up in the viewers. I also confirmed that the meta-data from the DICOM was not accessible to the Orthanc web browser.
I went digging in the back end Orthanc Storage on disk. I have the Orthanc set up to store DICOM to disk rather than into the SQL database.
I tracked down the location of one of the files.
It looks like what is stored where the DICOM would be is the JSON dump you might get from /instances/OID/tags. By that, I mean it’s a JSON dump of the DICOM meta data in a text file - rather than the binary DICOM I would expect to find.
In that same folder, there are actual DICOM binary files from other studies. Given how Orthanc generates its own UID and stores files on the back end, it makes sense that any given folder might have DICOM from a number of different studies.
These particular problematic files went through our system back in 2018. I generally update Orthanc each year or so, but I need to dig into my notes to see what version we were running at the time.
Even if I had alternative copies of these DICOM, I’m not sure how to clean this up. I would need a way to both delete the incorrect JSON dumps in OrthancStorage AND somehow tell Orthanc to re-index its meta data and “forget” about those problem files.
Has anyone run into this issue before?
Thanks,
John.