Hi David,

I'm struggling with what to do to make some files DICOM compliant and not produce any outright errors with dciodvfy.

Laterality is 2C, based on whether Body Part Examined has a certain type. But Body Part Examined is Type 3, so in this case, whether Laterality should be present is formally undefined.

However, in this case, dciodvfy reports an error. Is that correct behaviour?
Error - Missing attribute Type 2C Conditional Element=<Laterality> Module=<GeneralSeries>

The reason this matters currently is that images I have deidentified are being rejected by a SECTRA PACS, but it gives no diagnostics and the best I can do is to try and ensure the images throw no errors with your tool.

Best wishes,

Simon

Is the correct thing to do in this case to include the element but set it to an empty string, signifying "unknown"?

I would be surprised if Laterality issues had anything to do with why Sectra PACS is not accepting your images, so I would continue looking for other reasons.

The "correct" thing to do is to include a valid Body Part Examined (or similar) value and then send Laterality if it is appropriate with a valid value..

My dciodvfy tool is very aggressive about Laterality and Body Part Examined in order to encourage them to be populated correctly and consistently rather than be left empty or absent, using the Type 2(C) definition of required to have a value unless unknown as a reason to force the issue (begging the question of unknown to whom?).

So if you just include Laterality with zero length, and BodyPartExamined is not present or has a value that is a recognized unpaired body part (e.g., ABDOMEN rather than ARM), it will still complain until you get it right, reporting either:

Warning - is only permitted to be empty when actually unknown; should be absent (not empty) if an unpaired body part, and have a value if a paired body part - attribute <Laterality>

or

Error - Attribute present when condition unsatisfied (which may not be present otherwise) Type 2C Conditional Element=<Laterality> Module=<GeneralSeries>

David

On Saturday, May 27, 2023 at 5:52:20 PM UTC-4, Simon Doran wrote:
> Is the correct thing to do in this case to include the element but set it to an empty string, signifying "unknown"?

On Friday, May 26, 2023 at 3:38:41 PM UTC-4, Simon Doran wrote:

> I'm struggling with what to do to make some files DICOM compliant and not produce any outright errors with dciodvfy.
>
> Laterality is 2C, based on whether Body Part Examined has a certain type. But Body Part Examined is Type 3, so in this case, whether Laterality should be present is formally undefined.
>
> However, in this case, dciodvfy reports an error. Is that correct behaviour?
> Error - Missing attribute Type 2C Conditional Element=<Laterality> Module=<GeneralSeries>
>
> The reason this matters currently is that images I have deidentified are being rejected by a SECTRA PACS, but it gives no diagnostics and the best I can do is to try and ensure the images throw no errors with your tool.

Hi David,

Thanks very much for the speedy reply.

I agree with the desirability of what you are trying to do, but surely it's slightly unfair to be "aggressive" when a decision was obviously made somewhere along the line to make Body Part Examined a Type 3 element?

In the files that I received as input, neither Body Part Examined nor Laterality were present, making them non-compliant and leading to your error code above. Under those conditions, then there's nothing else I can do for a valid DICOM file other than set Laterality to an empty string, right? That way, at least the error converts to the warning.

Best wishes,

Simon

fredag 26 maj 2023 kl. 21:38:41 UTC+2 skrev Simon Doran:
> The reason this matters currently is that images I have deidentified are being rejected by a SECTRA PACS, but it gives no diagnostics and the best I can do is to try and ensure the images throw no errors with your tool.

Hi Simon,

I can confirm that the Laterality issue is not the reason why Sectra PACS rejects the image. Sectra PACS accepts images lacking Laterality information when DICOM requires it.

System administrators of the receiving system should be able to have a look in the Sectra PACS import log for the relevant DICOM Storage SCP, where the reason for the rejection usually is easy to identify.

Since you have deidentified the images one reason could be a mismatch between administrative information and the UIDs or the Patient ID. Sectra PACS usually requires that what you have in (0008,0050) Accession Number and (0020,0010) Study ID is consistent with (0020,000D) Study Instance UID and (0020,000E) Series Instance UID. For instance, if you change Accession Number and/or Study ID but keep the original Study and Series Instance UIDs and Sectra PACS already have registered the original image, the new image would be rejected, using the default configuration of Sectra PACS. Another possibility is a similar mismatch between (0008,0050) Accession Number/(0020,0010) Study ID and (0010,0020) Patient ID.

-- Krister

Thanks, Krister.

That's very helpful. I was just starting to come round to the same conclusion myself.

Best wishes,
Simon