Is deid.dicom aquivalent to CTP's pixel.anonymizer.script?

[fimafurman]

I noticed that deid.dicom is missing entries in CTP's pixel.anonymizer.script - is deid's deid.dicom config file modeled after CTP's script?

[vsoch]

Yes! Would you care to PR to update? Would be greatly appreciated!

[fimafurman]

Sure! It is likely that there several entries missing as the model I noticed was quite old from 2013, so I will try to come up with a comprehensive list of gaps.

…

On Wed, Sep 12, 2018 at 10:15 AM Vanessa Sochat ***@***.***> wrote: Yes! Would you care to PR to update? Would be greatly appreciated! — You are receiving this because you authored the thread. Reply to this email directly, view it on GitHub <#63 (comment)>, or mute the thread <https://github.com/notifications/unsubscribe-auth/Ak2KXEHMehLWV_zhc-TV1Q2jXhDz3Vcfks5uaRcTgaJpZM4Wlchq> .

[fimafurman]

Question:

I have an example image which gets both blacklisted and greylisted with coordinates. It seems to hit this blacklist test:

LABEL Burned In Annotation # (CTP)
contains ImageType SAVE

• contains Modality CT|MR
  || contains SeriesDescription SAVE
  || contains BurnedInAnnotation YES
  || empty ImageType
  || empty DateOfSecondaryCapture
  || empty SecondaryCaptureDeviceManufacturer
  || empty SecondaryCaptureDeviceManufacturerModelName
  || empty SecondaryCaptureDeviceSoftwareVersions

With:

"group": "blacklist",
"reason": " ImageType contains SAVE and Modality contains CT|MR or SeriesDescription contains SAVE or BurnedInAnnotation contains YES or ImageType empty or DateOfSecondaryCapture empty or SecondaryCaptureDeviceManufacturer empty or SecondaryCaptureDeviceManufacturerModelName empty or SecondaryCaptureDeviceSoftwareVersions empty "

1. Should the image be cleaned or deleted?
2. Is there a way to group ad and or conditions in the script?

[vsoch]

hey @fimafurman ! The idea of the recipe is that you are intended to write your own criteria to supplement or replace the default. There is no perfect default that everyone would be happy with. If you need to change the above, you can freely do so and distribute your own file, or if you think it should be default, do a PR to change the default! Now to answer your questions:

• the blacklist is hit first, just by way of being higher in the file, since it's more "urgent" it takes priority (but it meets criteria for both of the groups as you can see) I don't see issue with this second point because it's reasonable that a user could want to have fuzzy categorization with respect to groups.
• the image being cleaned or deleted is up to you - it's your tool / decision, etc.
• the conditions are grouped based on the filter groups! So you can make your own group with a set of criteria! The names of "blacklist" and "greylist" etc are just arbitrary - you can call whatever you like.

[fimafurman]

I would assume that it tries to mirror CTP functionality which does attempt to make a decision for you. So my question is more what is the default behavior with such condition if there are no override rules?

[vsoch]

This default I had help developing with users that were actively invested in the filter criteria - honestly this is not something that I want to take responsibility for, and indeed I don't have the expertise. For this situation I would recommend that you figure out the filters that are missing / need to be changed, suggest fixes, and PR with them. If there are no "override rules" for something and it doesn't get flagged, it's not going to be flagged. I can't offer much more explanation than that - I'm not in the business of managing PHI or rules about them other than developing this initial default to really get users started, not enforce anything. Deid is not published or approved by any privacy board, so it's use at your own decision.

[fimafurman]

Thanks - that's fair - I just wanted to understand a bit better how it behaves by default. I will play with it some more and based on testing, update the default recipe. Thanks!

[vsoch]

Sounds good.

[vsoch]

@fimafurman do you have any additions (logic for the parser) to contribute, or shall we close the issue?

[fimafurman]

Yes. How would you like me to share it with you?

[vsoch]

Any way that works for you! The easiest is likely dropping the file you are using here, and then I can open a PR to add to deid.

[fimafurman]

Attached below.

I was surprised how incomplete the CTP published file is, and, as the result the deid, so we had to populate file fairly manually. If you know of any sources of complete scripts for CTP (for example) I would appreciate it.
ffur_deid copy.dicom.zip

[vsoch]

Excellent, thank you! I will get these integrated asap.

I share the same sentiment - I was really surprised that "industry standard" was lacking so badly. I think it must be the case that centers add their own custom logic and don't share, because the culture (I've noticed) is fairly closed off when it comes to sharing code, etc.