ENH: Adds length measurement import #235
Conversation
This is a first pass at reading imaging measurement reports. Currently length measurements are mapped to annotation rulers. This has been tested using the measurements converted by @dclunie as part of the CrowdsCureCancer project.
|
@pieper I am thinking we should wait until the structure of the CrowdsCureCancer project SRs is finalized, and we have a public sample dataset that can be used for testing this feature. What do you think? |
|
Right, there's no particular rush. @erik Ziegler <[email protected]> and
I are working on the javascript version of this and we'll have more worked
out examples as a result of that effort.
…On Thu, Oct 11, 2018 at 1:06 PM Andrey Fedorov ***@***.***> wrote:
@pieper <https://github.com/pieper> I am thinking we should wait until
the structure of the CrowdsCureCancer project SRs is finalized, and we have
a public sample dataset that can be used for testing this feature. What do
you think?
—
You are receiving this because you were mentioned.
Reply to this email directly, view it on GitHub
<#235 (comment)>,
or mute the thread
<https://github.com/notifications/unsubscribe-auth/AAHsfdnZkaRjqkY5TecRcH8zJ0GUaUQVks5uj3qigaJpZM4WV1d3>
.
|
|
Sounds good - once David confirms the final structure, and the sample datasets are established, please update this PR so I know what to test it against! Thank you! |
|
@malaterre FYI this is the PR I mentioned in the email exchange earlier today so you can monitor it directly. |
|
@nroduit Just FYI CrowdsCureCancer folks are working on DICOM SR with measurement (non-image based). |
|
@fedorov shall we find some time to talk about merging this? It's not a blocking issue before merging but we should discuss how we want to handle variants of report and measurement types. |
|
Steve, I've been working to migrate/reorganize DICOM4QI content, and want to finish that first, but I will get to testing this right after, now that I have the confirmed sample dataset from David, which I will also add to DICOM4QI. |
|
@pieper as I was combing through email threads, it looks like I still do not have a sample SR with the matching image series. Can you share one with me? David only sent the sample SR. |
|
@pieper I searched this page for "Crowds", but didn't find the matching collection: https://www.cancerimagingarchive.net/. Can you point me to the location of data? |
|
For the record, it is quite awkward to load and trace back stuff in that collection, since the only option is to download the whole dataset of images, which I really don't need. So to locate the matching image, I opened the SR provided by David, looked up patient ID, from Study figured out it is a Kidney annotation, then searched TCIA main page for "Renal", which led me to TCIA-KIRC collection, which I then searched for subject TCGA-BP-4343. The study for that subject contained 6 series, and I located the matching one (series 3) by UID ... what a pain. Why would they not put those SRs into TCIA DICOM database proper? |
Good question for Justin? |
| rulerNode.SetLocked(True) | ||
| rulerNode.SetName(contents['personObserver']) | ||
|
|
||
| referenceFilePath = slicer.dicomDatabase.fileForInstance(measurement['referencedSOPInstanceUID']) |
There was a problem hiding this comment.
It will be good (and consistent with the behavior of the module for SEG-based measurements) to populate references (which can be done, since SeriesInstanceUID is present) and offer the user to load the corresponding image series on load of the SR with annotation/measurement. Can you add that functionality? (see https://github.com/QIICR/QuantitativeReporting/blob/master/DICOMPlugins/DICOMTID1500Plugin.py#L74)
This is a first pass at reading imaging measurement reports.
Currently length measurements are mapped to annotation rulers.
This has been tested using the measurements converted
by @dclunie as part of the CrowdsCureCancer project.