qiicr/slicer-petdicomextension

3D Slicer extension containing tools for importing DICOM PET series

Slicer PET DICOM Extension

The PET DICOM Extension provides tools to import PET Standardized Uptake Value (SUV) images from DICOM into 3D Slicer. SUV computation is based on the vendor-neutral "happy path only" calculation described on the Quantitative Imaging Biomarkers Alliance (QIBA) wiki page Standardized Uptake Value (SUV).

Documentation: https://www.slicer.org/wiki/Documentation/Nightly/Extensions/PETDICOM

Acknowledgements

This work is funded in part by Quantitative Imaging to Assess Response in Cancer Therapy Trials NIH grant U01-CA140206 (PIs John Buatti, Tom Casavant, Michael Graham, Milan Sonka) and Quantitative Image Informatics for Cancer Research (QIICR) NIH grant U24 CA180918.

slicer-petdicomextension's Issues

SUV conversion will fail if there are extra files in the directory with the input series

The class that is used to generate file names specifically makes the assumption "assuming there is only one study/series".

https://itk.org/Doxygen/html/classitk_1_1GDCMSeriesFileNames.html

https://github.com/QIICR/Slicer-SUVFactorCalculator/blob/master/SUVFactorCalculator.cxx#L791
https://github.com/QIICR/Slicer-SUVFactorCalculator/blob/master/SUVFactorCalculator.cxx#L832

Extension is not functional on mac

Steps to reproduce:

1. download QIN-HEADNECK case 73 (screenshot 1)

2. try to parse from DICOM module (screenshot 2)

3. error is in screenshot 3

SUV conversion error - Can’t find tag 18 1072 - but tag 18 1078 available

Hello,

I can’t convert my DICOM files in Bq/mL into SUV.
I think the problem comes from the metadata of my data which do not contain the tag (0018,1072) RadiopharmaceuticalStartTime, as the error message of the “SUV Factor Calculator” module tells me:
itk::ExceptionObject (0000005A9AB9D0D0)
Location: “unknown”
File: D:\D\S\S-0-E-b\PETDICOMExtension\SUVFactorCalculatorCLI\itkDCMTKFileReader.cxx
Line: 89
Description: ITK ERROR: Can’t find tag 18 1072

However, in the documentation of this module, they mention that the tag (0018,1078) RadiopharmaceuticalStartDateTime​ can be interchangeable with the tag (0018,1072) RadiopharmaceuticalStartTime, and my metadata contains this tag.

How can I overcome this error message?

Thank you very much for your attention !

PixelData should not be precached

The dicom module uses the plugin's tags member to know which fields to put into the TagCache database:

https://github.com/Slicer/Slicer/blob/bc6482c9a4abaabe3f93b0bfee31d5a8a2c4ae97/Modules/Scripted/DICOM/DICOM.py#L176-L185

But the PETSUV plugin includes PixelData in the tags member:

https://github.com/QIICR/Slicer-PETDICOMExtension/blob/master/DICOMPETSUVPlugin/DICOMPETSUVPlugin.py#L65

This probably leads to a large tag cache and poor performance, although I haven't timed it.

Possibly related to #21

Very slow loading when many studies in the DICOM database

I downloaded all the studies from IDC with the PET modality (about 1700 studies) and loaded them into the Slicer dicom database. I installed the PET DICOM extensions in Slicer 5.0.3-2023-08-03.

I'm able to examine the series and see that the PET SUV Plugin correctly identifies that the PET series and is selected to load, but when I try to load it appears that Slicer is in an infinite loop, or at least something very very slow.

If I export the same study from the database to a directory by itself, and then import that directory into a newly created database the study loads quickly. I take this to mean that the plugin is looping over all studies in the database or something similarly inefficient.

Environment to Build PETDICOMExtension

I would like to build the Slicer-PETDICOMExtension on my own computer as a command-line tool to use. PETDICOMExtension seems to need to pre-install some libraries, such as Slicer, DCMTK, ITK, VTK, etc, to compile.

Any documentation to install all required libraries for compiling PETDICOMExtension? It would be great if there is a Dockerfile for compiling environment configuration.

Calculation of SUV values

I believe the calculation for SUV is as follows:

SUV = PET_activity * weight / decay_corrected_dose

Your implementation does NOT convert the PET_raw intensity values (0 to 32767) to PET_activity values using rescale slope and rescale intercept by the following equation:

PET_activity = PET_raw * rescale_slope + rescale_intercept

Can you please verify this and confirm the same? If the issue is valid, can you explain the fundamentals behind your implementation?

Visualization of PET volumes need to be improved: LUT and w/l

SUV volumes should probably have PET-specific lookup table (PET-heat?) and also some window/level pre-set that makes sense for clinicians. @ejulrich do you know if there is a certain way SUV volumes are typically windowed?

PETDICOMExtension is not available in Slicer ExtensionManager

see http://slicer.cdash.org/viewBuildError.php?buildid=1066505 for further information

Cause by most recent DCMTK update which set DCM_OtherPatientIDs to retired

Implemented SUV calculation - documentation

Guys, can you confirm - is it correct that this plugin implements the calculation as suggested by QIBA in the document on this page: http://qibawiki.rsna.org/index.php/Standardized_Uptake_Value_(SUV) ?

If yes, I suggest we add a link and explanation to the README. If not - some explanation why not and what are the differences.

What do you think?

builds failing due to ofstream

Nightly builds are failing.

https://slicer.cdash.org/viewBuildError.php?buildid=2193804

I'm going to see if I can make a fix.