Which is broken: the DICOM or the converter?

Question

I'm trying to convert a 4D DICOM image (x,y,z,time) to a different file format. Something goes wrong, because the output image has lost the time dimension.

I'm trying to decide whether:

• the DICOM series is broken -- it's possible that a 3rd party, who anonymized data, removed critical information from the header; or
• the conversion code is incomplete -- it simply can't handle this flavour of DICOM

The answer to this will determine whether I have to fix the DICOM, or fix the converter.

I've tried diving into the DICOM standard, to understand what specific header values mean, but I don't find this document helpful; it gives a mere word or two for each header field. I see fields in my data that look suspicious, but I don't know if it's actually wrong, or if I don't understand what it's supposed to be telling me.

I can think of several ways to answer my problem:

1. Are there any tools out there that can confidently classify a DICOM series as either valid or invalid?
2. Is there a document which describes precisely what each DICOM header value is supposed to contain?
3. Is there a better approach to figuring out which is broken -- the image, or the converter?

Answer 1

You are not looking at the right document, you should be looking at PS 3.3 - current. For example:

• A.4.3 MR Image IOD Module Table

Or as someone mentioned in the comments use dciodvfy from dicom3tools package.