Attendees: Jay Gaeta, Carl Jaffe, John Perry, Justin Kirby, John Freymann, Eric Kascic
- Discussed progress of CTP development efforts to this point
- JP to provide new CTP version to Justin and Jay for testing in the next week
- This new version will include:
- changes discussed to handle invalid dates that CTP tries to increment (will provide blank date instead of 7/4/1776)
- new feature to apply Supp142 profiles to be applied against existing profiles
- Discussed outstanding issues pertaining to crafting the anonymization script for the Supp142 Basic Profile
- updated the list of issues with new resolutions and action items on outstanding issues
************
These are Draft notes below - needs to be cleaned up!! - JBF
Profile: Specified by DICOM 142
Script: Implementation of a Profile, or modification thereof
If you are editting a script
For example - if you created a private group, then applied the baseline DICOM142 profile to it - it would highlight the
CTP on startup it loads a configuration. Configuration specifies a script file. In absence of script file, it loads a 'default' script file. From then on it uses that particular script file.
Suggestion: Default script is based on the DICOM142 aggressive profile.
Levels of Script Instructions:
Replacement field of an element
Global Instructions:
- Global Keep
- Global Remove
Individual instructions (keeps, removes, and modification instructions) take precedence over Global Instructions (and keep takes precedence over remove)
Script file has 3 categories,
JP: New configurator will have a menu bar at the top with anonymizer instructions (so you would never have to edit the script file with a text editor)
If you have two anonymizer stages pointed at the same script file, modifications would be applied to both stages (not necessarily desirable)
"Key" management:
JP - a script attribute would point to a file that contains the key. Could be contained on a USB fob.
The way it works now every element can have its own key. Not desirable.
Suggest the key should be a
************CTP WG 18 Meeting
John Perry.
Profiles are things defined by DICOM.
Justin list of issues:
Group 2 it's in 'file metadata' , not in the main body
• Date error - the CTP was trying to modify what it thought was a date but it was not valid
In the DICOM files
There is a DICOM tag designed to hold whether it is not deidentified.
Tag should add in the whole anonymization history...
Whow wrote supp 142: David Clunie editor, Jay was a reviewer. Others involved. JP will ask
Group 12 is a DICOM standard group...
Jay- Within the clinical trial, when we manage it we keep the provinance in the group 12.
One at the acqui
Jay will send script for group 12
Jay will try to standardize and use the same scripting language... John P has wanted to change the scripting language for some time...
