Mammo RDSR crashes on import at Anatomical structure

Issue #368 resolved
Ed McDonagh created an issue

Fails with Hologic RDSR. Found issue and fix, but need to check that it won't stop fluoro RDSRs working if I make the change.

Issue reported by Erik-Jan Rijkhorst on the google group forum.

Comments (17)

  1. Ed McDonagh reporter

    The Anatomical structure sequence doesn't appear to exist in the Siemens or Toshiba fluoro RDSRs

  2. Ed McDonagh reporter

    Fixed 3 TID 10003 elements that have not previously been used as the extraction code didn't match what would be found in the wild or match the models! Mammo RDSR now imports but thinks it is a fluoro study... Refs #368

    → <<cset 58e6818a593c>>

  3. Ed McDonagh reporter

    Study now goes in as MG and mostly works, but I need to review the RDSR code to ensure it is fit for mammography and makes the most of the data available. Refs #368

    → <<cset d079e862025b>>

  4. Ed McDonagh reporter

    Mammo RDSR should only complete the accumulated mammo xray dose table, not the others. Refs #368. acuisition_device_type should have been a ContextID foreignkey not a charfield. The value is currently never used, but should be to determine which Accumulated template should be expected. Created new cid version.

    → <<cset 39f6a9d56777>>

  5. Ed McDonagh reporter

    Now recording TID 10001 acquistion_device_type, and using it along with procedure_reported to determine which accumulated templates should be populated. Needs testing against all modalities! Refs #368

    → <<cset 3c2feb7e35d0>>

  6. Ed McDonagh reporter

    Filtering order, study delete rf, used the wrong DAP and Dose RP totals; corrected. Removed DAP unit conversion function from AccumProjXRayDose model - it shouldn't be used. Changed a couple of references to it. Refs #368

    → <<cset 074747024780>>

  7. Ed McDonagh reporter

    Laterality at event level was overwritten by anatomical structure value due to error in code. Fixed. CT and Fluoro don't generally use this as far as I know, so shouldn't be an issue. Updated detail view to include breast laterality in view column. Refs #368

    → <<cset cb54b0aefadd>>

  8. Ed McDonagh reporter

    Tested this branch against current develop: imported various fluoroscopy RDSRs into a fresh database using both branches, then exported them using csv and XSLX. Results were identical, though the version on this branch was faster on import and export. XLSX export in particular could do with a lot of optimisation though!

  9. Log in to comment