iusca / sqan Goto Github PK

This is a brainstorming task before we execute any plans. We would like to allow any/all of the following

• Template override and/or reQC with alternative template at subject level
• Template override and/or reQC with alternative template at research study IIBIS level
• One or both of the above for selected set of series only?
• Do the options above go in the existing ReQC button?

In 2010-00166 PPE080 the errors were a little confusing as to what they mean. Perhaps a summary page that explains what an error indicates that we could refer to would be useful.

When doing an exam-level QC override, the missing series field is not cleared out, leaving erroneous missing series warnings in place.

• use the same coloring scheme as elsewhere (i.e fail = red, etc.)
• use arrows or a visual indicator of the change instead of the current verbiage; for example "Updated QC1 state <red fail/error> => <green/OK> (2019-04-04 10 PM)"

This is an FYI for @jdwest0711 and @mparamas -- I had demo'ed our use of Zenhub (kanban boards for project management). I am afraid you won't have continued access to it. I had hoped our infinite 'free' github org status would apply to the Zenhub marketplace app but it's not, therefore we are having to pay for it.

You can still create/respond to/close issues directly on github.com/IUSCA/rady-qc ...

Issues in epic focus on the work the RADY and the SCA teams will have to do after the operational shutdown to upgrade the UI/API, the database, etc. prior to a mass data resend. Note that among the many changes, the SCA side will be able to store the entire DICOM metadata header following the completion of this set of tasks.

Will meet 1-3 PM Wed, Feb 12th 2019 via zoom https://iu.zoom.us/j/8128560187

It appears we are not picking up radiotracer for CT/PT images.

From John @jdwest0711

For PET/CT, I think we are seeing what could be a problem going forward with how we classify these. The PET is associated with a tracer while the CT is not associated with anything other than being a CT. For 0000-00004 IIBISID this causes the problem that for the PET data the templates fit fine but for the CT they don’t know which to use. I’m wondering if there is anything in the CT headers that will allow us to link them with the appropriate PET. CT and PET are different modalities, but they come from the same scanner and there is always CT data associated with a PET data set. I know in the beginning we said to list it as a separate modality so I’m not sure how this can be rectified but it is worth a discussion.

Discussed in video meeting today (Wed, Feb 20 2-3p)

Request to add a Data Received by SCA data to template information similar to how it exists on data sets.

Following up from https://github.com/IUSCA/rady-qc/issues/51

• Develop application/process or add to process above to move these to SDA (or stage tarballs off SDA) - may be a separate ticket in the future

It might be useful to add expand/collapse across the board as well (e.g. IIBIS or modality+tracer level on the LHS).

This fixes https://github.com/IUSCA/rady-qc/issues/39

Some data appears as having two images with repeated InstanceNumbers. This will be solved by using the StudyInstanceUID as an exam's identifier, instead of using the StudyTimestamp, as we have been doing so far.

If we want this change to be applied to the entire dataset, we will have to re-ingest all data. I am not sure that this is necessary, but, it is not trivial to single out all the cases where there are multiple StudyInstanceUID with identical StudyTimestamp.

Regardless, we need to first standardize the header files (i.e. tarballs vs .json files) before we can start re-ingesting any data at all

In 2010-00166 PPE080 the events on the right side are not in chronological order. I think they should be in order from newest to oldest so that the latest change is at the top. Is this doable?

The summary color bar on the left side is not reflecting warnings in the data. A good example of this is 10786 in 0000-00001 which has warnings on everything but shows all green in the summary plus light blue for missing scans.

For e.g.

[REDACTED] See for screenshot.

Let's consider doing the following

• Grey out the SeriesDescription for ignored series (the rows on which the QC status background color is lightened )
• Make the Show/Hide toggle a 3-way toggle with a tooltip informing the user about this so we can hide the QC status or we can hide the ignored Series (rows) altogether

FYI @aiavenak , I sent this to @youngmd

Edited by @agopu on 2019-08-01 - moved screenshots with and/or information stakeholder prefers to keep private to an issue on that project's private repo.

I would like us to add a "Are you sure?!" confirmation for user actions especially deletes (e.g. templates in the template summary view).

John to schedule F2F demo to both:

• Technologists
• Researchers

... preferably the same day, shooting for the week of Apr 29-May 3...

• The template view is a bit confusing. When one is expanded, we should probably grey out or add space before we show more IIBIS, etc.

FYI @youngmd

Allow users to click on a template series from the template summary, and go to the template view to inspect template headers

Let's get in the habit of using as much standard (boilerplate/template) text as possible, and annotate that here.

Sundar and I will be meeting with PET/CT researchers to identify if the data captured is meeting needs.

We have set 2:30 on 2/19 as the date for this. We will push back if we cannot have the data to look at ready by that time.

The agenda includes:

1. Are the parameters you really need getting captured?
2. How do we best go about setting up good templates for QC’ing against?
3. Can we identify someone to be more involved in Sundar and I’s meetings with the RADY SCA team working on this project to be able to give more input on the PET/CT side going forward?

Will post feedback from that meeting here.

Issues in this epic focus on prep work necessary before we can shutdown production RADY-SCA and begin significant refactoring upgrades.

This epic includes tasks associated with the mass resend itself including resend of templates (including phantoms) and image datasets + following up with technologists/researchers with status updates and feature update info.

No private tags are currently in the list of integer/float conversions in api/qc/instance.js done as part of the ingestion process. This leads to false QC errors for these fields where exact string matching fails on acceptable numerical differences. Additionally, if there are private tags which are exam-unique (e.g. timestamps, patient descriptions) they should be added to the list of ignored fields in api/qc/template.js

• Radio tracer in the Research summary view should be Left Justified
• Attempt to swap subjects and series so we go from those two entities being Columns/Rows => Rows/Columns
  • Actually Andrea is making is a 'Transpose' switch/button instead that the user can click on at any time.
• Andrea plans to refactor the top area where we show IIBIS details to waste less white space and/or allow user to hide IIBIS details with a +/- buttons.
• SEE #57 (comment) for additional requirements

Crud:

• I think we need to paginate the research summary view
  • UPDATE 2019-05-28: Andrea / Arvind unclear what the request means, so striking out this requirement.

This information is not being picked up when the series is not qc-ed.
Also, the information about when series is received is now duplicated because we have it as an event, and we also have it in the 'createdAt' field. We should stick to one and remove the other to avoid redundancy

Looking at 2010-00166 subject PPE0978 there is a missing headers error for both series. This seems to be unclear for two reasons. One we have no information on what headers are missing. Two, this is confusing for researchers as when they consider the word "header" they consider the entire dicom header to be an entity and each piece of information in that header to be a field (i.e. SeriesDescription, PatientName, ScanDate, RepetitionTime, etc.). We may need to clarify this and add extra information.

Mike suggest on a slack conversation we should add radiotracer to the exam search

This is tied to #79

It can be hard to tell if an error at first glance is due to missing images. Could we have something that for that particular error pops up? For example next to ERRORS: 1 maybe it could say Missing individual images in series. This would allow us to quickly see that a subject is missing data and realize the send may have gone bad.

Second, could we add it so when you click on the series that has missing images, those images are highlighted. Right now there is just less boxes, but it would be nice to have something similar to when those images are missing in a the template there are blue boxes. This would allow us to see exactly which images are missing and investigate them further to see if they have issues on our side or if we resend data if the same images go missing again.

Sundar and I were discussing this and it would be great to be able to Query the DICOM receiver for Orthanc to verify it received all the images. Do we think this is possible?

We currently only allow the deletion of templates that are not used for QC. We would like to allow users to delete any template; if a template is used for QC and is deleted, then the series must be un-qced.

This is a two fold problem. For CT there is a CT template for each PET tracer, but unlike pet where each PET tracer gets a unique template, the two CT templates are considered the same template but different versions. We would like a way to link a CT template to a particular tracer.

For MR, some IIBIS IDs have two templates for different data sets. We would like a way to differentiate there as well. This may need an addition on Radiology side from scanner.

Template view needs to be updated to work with the new DB model

In video meeting today

5. John points out IIBIIS [REDATED] has two exams a year apart, this used to be listed side by side but that's no longer the case any more.

Mike: May be able to sort it hierarchically.

FYI @jdwest0711 @aiavenak @mparamas

I can still see errors for GRE field maps as before. I believe these are due to a persistent issue we've encountered already. IIBIS id's that show this (not all inclusive) 2018-00072 and 2015-00094

To test out the functionality we did the QC1 accept on one of 2010-00166 PPE080 series. Karmen entered a reason in the field that popped up but then we couldn't find that information anywhere on the page. Are we missing it or is there a bug?

Following up on https://docs.google.com/document/d/1U7Z6LOikpjccxNgrqVfXB4hM9GPQ1px3v3zhRYMG3ys/edit and discussion Mike & i had about this ... we may color code templates including one used for QCing in the portal. Let's brainstorm this and figure out feasibility.

FYI: @jdwest0711 @mparamas

When a researcher is looking at a particular series they would like the subject name to be more prominent so it is easy to see. They consider that piece of information more critical than study time stamp or series description as they are usually QC'ing on a subject basis, then series within a subect. So making the subject name more prominent would make it easier to keep track.

Recently found working through Karmen's list that there are three fields that have different HEX values but same name. This is probably true for MR as well, but have not verified.

[0029,0010],[0071,0010],[0071,0011] all are called PrivateCreator. How can we handle a situation such as this?