- changed milestone to 0.8.2
-
assigned issue to
Philips Azurion dosereports rejected by openrem
When I send a dosereport from an azurion system it does produce an entry in the database but is not showing up for any modality and django seems to throw an error:
django.core.exceptions.ValidationError: [u"'' value must be a decimal number."]
Analysis by Ed suggests that what is happening is as follows: The calibration factor information is present, but the values are empty The numeric values (factor, uncertainty) can be blank, but they can't be a string We don't check what the value is - if the CodeMeaning is there, we assign the NumericValue to the table field When we then save the table, it tries to save the DecimalField with the value '', which isn't allowed. We'll need to put some if statements in there to ensure we have a number before assigning the value
Comments (20)
-
-
Temporary local fix for empty calibration factor for ref
#673. However, structure of RDSR is different so accumulated data isn't being extracted. Need to walk through the RDSR and the code to establish what is going on!→ <<cset 6b3bc813e9ac>>
-
- First issue is that there is no
StudyInstanceUID
! So the duplicates check code is skipped. - No
AccessionNumber
- has the example you sent me been excessively 'anonymised'?
Darn, mouse stopped working in my VM. Have to save this comment and restart it :(
- First issue is that there is no
-
The
AcquisitionDeviceType
is presented asIntegrated Projection Radiography System
. My reading of the standard is that this should beFluoroscopy-Guided Projection Radiography System
.As a result, none of the accumulated fluoroscopy or acquisition data is extracted by OpenREM as the relevant function isn't triggered. ThereforeOpenREM presents the study as radiography, and if you use the function to specify the modality all of the accumulated fluoroscopy information is missing.
@dickrsch - is this a full release of the software? This seems wrong to me. Do you think I have interpreted the standard wrongly?
-
reporter That sounds reasonable. This is a full release of the software though and seems to deviate significantly from the previous (Allura) platform implementation. I should dive into the standard to see if I can understand why they went for Integrated Projection system....
-
Hacky fix for non-conformant Philips Azurion to get accumulated fluoro data. Refs
#673→ <<cset c6dc99a97167>>
-
This fluoro system has Molybdenum filters - I've not seen that before outside mammo...
-
Made the numeric value test more generic. Tests still pass. Refs
#673→ <<cset 87f798ffb567>>
-
PyCharm code reformat due to extra long lines from last commit. Refs
#673→ <<cset a4d871f4215f>>
-
PyCharm imports optimised. Refs
#673→ <<cset 1a48d8cf20ec>>
-
Forgot to remove test_numeric_import from _calibration. Refs
#673→ <<cset a3c58e6780f5>>
-
Used the same hacky method to define as RF as was used to trigger accumulated fluoro fields. Refs
#673→ <<cset 78349ed7c57a>>
-
@dickrsch - would you be able to test the changes in this branch -
you'll need to replace your
rdsr.py
with this one: https://bitbucket.org/openrem/openrem/raw/78349ed7c57a904c9251f9a7efae351b79a72350/openrem/remapp/extractors/rdsr.py,and replace your
tools/get_values.py
with this one: https://bitbucket.org/openrem/openrem/raw/78349ed7c57a904c9251f9a7efae351b79a72350/openrem/remapp/tools/get_values.py -
You might like to use the
Config
->Display names and modalities
->Review
option to remove any remnants of the previous import attempts for this system. You might find it in DX or Other, or it might not be there. -
Added ref
#673to changes and updated release notes.→ <<cset 7f99dd119303>>
-
reporter Tested here and it seems to work.
-
Added test for ref
#673, but requested new sample file as this one is large and takes a while to import - will eat up all the test compute time!→ <<cset 707a89328687>>
-
- changed status to resolved
Merged in issue673emptycalibrations (pull request #229)
Fixes
#673Empty NumericValues and workaround for Azurion AcquisitionDeviceTypeApproved-by: David Platten
→ <<cset 23628a0c4883>>
-
Disabling Azurion testing until we have a smaller file to import. Refs
#673→ <<cset a00f2bf47ad0>>
-
- changed milestone to 0.9.0
- Log in to comment