Hello Again,
Our DICOM parser is failing to process some of the DICOM Files, on debug it looks like the following is causing the issue:
(0008,0005) CS [ISO_IR\100] # 10, 2 SpecificCharacterSet
The value was probably intended to be:
(0008,0005) CS [ISO_IR 100]
An example is SOP Instance UID 1.4.361.1.0.3155254.5.222.3028793833016864425
It is not just our DICOM Tool Kit that cannot process the record, Micro DICOM struggles as well, it will display the image but then cannot display the text based DICOM Tags.
For the challenge, should we just highlight such studies ? Or should we repair (0008,0005) and carry on with the de-id ?
Regards,
Chris
Created by Chris Ablett ChrisAb @ChrisAb, Could you please send your Synapse Team ID to my email (linmin.pei@nih.gov) by Sept. 29, otherwise, I will your Synapse ID (ChrisAb) on the whitelist allowing the test phase. Thanks! Hi @ChrisAb,
Thanks for bringing this to our attention. I have confirmed that this invalid character set is present in 306 instances from 2 series and only in the validation data, not the upcoming test data.
As it stands, this element is evaluated by a "text_retained" action and will be counted as a failure if you do repair it, however, the effect from two series on the final scoring should be negligible. It's your call.
Regarding your mention that other software has issues displaying it, I tried other viewers, including Micro DICOM and I'm not seeing the same behavior. All displayed it simply as "ISO_IR\100". The tag does have a VM of 1-n, so it is a valid structure, just invalid values.
Let us know if we can further assist.
Michael @ChrisAb ,
We are looking into the issue and will respond to you soon.
Another question: could you please provide your Synapse Team ID, or one Spynase ID for the test phase? One team is only allowed to submit once in the test phase. Thanks!