ENH: Add support for multiple metrics and regularization by Leirbag-gabrieL · Pull Request #1437 · SuperElastix/elastix

Leirbag-gabrieL · 2026-04-28T07:15:05Z

Hi, apparently it was not possible to do a multi-image registration with a transform metric like this:

(Metric "AdvancedMattesMutualInformation" "AdvancedMattesMutualInformation" "AdvancedMattesMutualInformation" "AdvancedMattesMutualInformation" "TransformBendingEnergyPenalty")
(Metric0Weight 1.0)
(Metric1Weight 1.0)
(Metric2Weight 1.0)
(Metric3Weight 1.0)
(Metric4Weight 10.0)

(NumberOfHistogramBins0 32)
(NumberOfHistogramBins1 32)
(NumberOfHistogramBins2 32)
(NumberOfHistogramBins3 32)

Now it is possible :)

N-Dekker · 2026-05-15T14:56:14Z

Thank you very much @Leirbag-gabrieL

Can you possibly make a minimal example of a registration that is made possible by your proposed fix? Ideally in the form of a library unit test, like this one:

elastix/Core/Main/GTesting/itkElastixRegistrationMethodGTest.cxx

Lines 3006 to 3048 in 0c0b19f

    
           GTEST_TEST(itkElastixRegistrationMethod, EuclideanDistancePointMetric) 
        
           { 
        
             static constexpr auto ImageDimension = 2U; 
        
             using PixelType = float; 
        
             using ImageType = itk::Image<PixelType, ImageDimension>; 
        
             using SizeType = itk::Size<ImageDimension>; 
        
             using OffsetType = itk::Offset<ImageDimension>; 
        
             const OffsetType translationOffset{ { 1, -2 } }; 
        
             const SizeType   imageSize{ { 5, 6 } }; 
        
             const auto fixedImage = CreateImage<PixelType>(imageSize); 
        
             const auto movingImage = CreateImage<PixelType>(imageSize); 
        
             elx::DefaultConstruct<ElastixRegistrationMethodType<ImageType>> registration{}; 
        
             using PointType = itk::Point<float, ImageDimension>; 
        
             const PointType fixedPoint{}; 
        
             PointType       movingPoint = fixedPoint; 
        
             for (unsigned int i{}; i < ImageDimension; ++i) 
        
             { 
        
               movingPoint[i] += translationOffset[i]; 
        
             } 
        
             registration.SetFixedImage(fixedImage); 
        
             registration.SetMovingImage(movingImage); 
        
             registration.SetFixedPoints(::MakeVectorContainer(std::vector<PointType>{ fixedPoint })); 
        
             registration.SetMovingPoints(::MakeVectorContainer(std::vector<PointType>{ movingPoint })); 
        
             registration.SetParameterObject(CreateParameterObject( 
        
               ParameterMapType{ // Parameters in alphabetic order: 
        
                                 { "ImageSampler", { "Full" } }, 
        
                                 { "MaximumNumberOfIterations", { "2" } }, 
        
                                 { "Metric", { "AdvancedNormalizedCorrelation", "CorrespondingPointsEuclideanDistanceMetric" } }, 
        
                                 { "Optimizer", { "AdaptiveStochasticGradientDescent" } }, 
        
                                 { "Registration", { "MultiMetricMultiResolutionRegistration" } }, 
        
                                 { "Transform", { "TranslationTransform" } } })); 
        
             registration.Update(); 
        
             const auto transformParameters = GetTransformParametersFromFilter(registration); 
        
             EXPECT_EQ(ConvertToOffset<ImageDimension>(transformParameters), translationOffset); 
        
           }

But then, having something like { "Metric", { "AdvancedMattesMutualInformation", "TransformBendingEnergyPenalty" } } instead.

Otherwise, a complete ("dummy") test set, including elastix parameter file, would also be of help to us. Again, preferably "minimal", so that it won't take much time to process.

N-Dekker · 2026-05-15T15:00:07Z

  using MovingImageDerivativeScalesType = FixedArray<double, Self::MovingImageDimension>;

+  /** Typedef for transform penalty metrics. */
+  typedef TransformPenaltyTerm<TFixedImage>             TransformMetricType;


Nowadays, we are using using, instead of typedef. But a coding style issue like that is not a show-stopper. I can adjust it afterwards. Also, I think I would move the definition of TransformMetricType into the function body of CombinationImageToImageMetric::SetMetric, as that seems to be the only place where it is needed.

N-Dekker · 2026-05-15T15:01:19Z


+  /** Typedef for transform penalty metrics. */
+  typedef TransformPenaltyTerm<TFixedImage>             TransformMetricType;
+  typedef typename TransformMetricType::Pointer         TransformMetricPointer;


TransformMetricPointer is never used, so I think I would remove it. (No show-stopper for now.)

N-Dekker · 2026-05-15T15:05:11Z

+  TransformMetricType * testPtr2 = dynamic_cast<TransformMetricType *>(metric);
+
+  // Increase newly defined numberOfMetric counters
+  if (testPtr1)


I would make it:

if( dynamic_cast<PointSetMetricType *>(metric) )

And remove the local testPtr1 variable.

Similar for testPtr2, oldTestPtr1, oldTestPtr2. No show-stopper to me, though. Just a style thing.

N-Dekker · 2026-05-21T11:51:41Z

@Leirbag-gabrieL How does your pull request relate to pull request ENH: Add support for multiple metrics and regularization #482 (@alvarez-pa, 2021, addressing issue Multiple metrics and regularization #481)? Do I understand correctly that you used it as a starting point, and you made some further improvements?

N-Dekker · 2026-06-18T13:35:27Z

-  if ((this->GetNumberOfMovingImages() != 1) && (this->GetNumberOfMovingImages() != nrOfMetrics))
+  if ((this->GetNumberOfMovingImages() != 1) && (this->GetNumberOfMovingImages() != nrOfImageMetrics))
  {
-    itkExceptionMacro("The NumberOfMovingImages should equal 1 or equal the NumberOfMetrics");
+    itkExceptionMacro("The NumberOfMovingImages should equal 1 or equal the nrOfImageMetrics");


@stefanklein @mstaring Would this change (comparing the number of input images with the number of image metrics, instead of the total number of metrics) break legacy use cases that would use "dummy" images when they use multiple metrics?

If so, shall we still allow such legacy use cases, by relaxing this check?

stefanklein · 2026-06-19T09:37:14Z

Hi Niels, I think it would indeed be a nice backward compatibility approach, if we still could allow the case when NumberOfMovingImages equals the NumberOfMetrics. Best, Stefan Van: Niels Dekker ***@***.***> Verzonden: donderdag 18 juni 2026 15:36 Aan: SuperElastix/elastix ***@***.***> CC: Stefan Klein ***@***.***>; Mention ***@***.***> Onderwerp: Re: [SuperElastix/elastix] ENH: Add support for multiple metrics and regularization (PR #1437) @N-Dekker commented on this pull request.

________________________________ In Components/Registrations/MultiMetricMultiResolutionRegistration/itkMultiMetricMultiResolutionImageRegistrationMethod.hxx<#1437 (comment)>:

- if ((this->GetNumberOfMovingImages() != 1) && (this->GetNumberOfMovingImages() != nrOfMetrics))

+ if ((this->GetNumberOfMovingImages() != 1) && (this->GetNumberOfMovingImages() != nrOfImageMetrics)) { - itkExceptionMacro("The NumberOfMovingImages should equal 1 or equal the NumberOfMetrics"); + itkExceptionMacro("The NumberOfMovingImages should equal 1 or equal the nrOfImageMetrics"); @stefanklein<https://github.com/stefanklein> @mstaring<https://github.com/mstaring> Would this change (comparing the number of input images with the number of image metrics, instead of the total number of metrics) break legacy use cases that would use "dummy" images when they use multiple metrics? If so, shall we still allow such legacy use cases, by relaxing this check? - Reply to this email directly, view it on GitHub<#1437?email_source=notifications&email_token=AAF2LNKV7IUHUQI57BH7UET5APV3LA5CNFSNUABKM5UWIORPF5TWS5BNNB2WEL2QOVWGYUTFOF2WK43UKJSXM2LFO4XTINJSGUZTKNJYGYZKM4TFMFZW63VHNVSW45DJN5XKKZLWMVXHJLDGN5XXIZLSL5RWY2LDNM#pullrequestreview-4525355862>, or unsubscribe<https://github.com/notifications/unsubscribe-auth/AAF2LNN2T5GSIBEIEY54IM35APV3LAVCNFSNUABEKJSXA33TNF2G64TZHM4TCNJYGY4TINB3JFZXG5LFHM2DGNBRGI4DGNZTGWQXMAQ>. Triage notifications, keep track of coding agent tasks and review pull requests on the go with GitHub Mobile for iOS<https://github.com/notifications/mobile/ios/AAF2LNNILPEGO2JK2DTSZJ35APV3LA5CNFSNUABKM5UWIORPF5TWS5BNNB2WEL2QOVWGYUTFOF2WK43UKJSXM2LFO4XTINJSGUZTKNJYGYZKM4TFMFZW63VHNVSW45DJN5XKKZLWMVXHJKTGN5XXIZLSL5UW64Y> and Android<https://github.com/notifications/mobile/android/AAF2LNJ4DPPQXUV7Q7P5GF35APV3LA5CNFSNUABKM5UWIORPF5TWS5BNNB2WEL2QOVWGYUTFOF2WK43UKJSXM2LFO4XTINJSGUZTKNJYGYZKM4TFMFZW63VHNVSW45DJN5XKKZLWMVXHJLTGN5XXIZLSL5QW4ZDSN5UWI>. Download it today! You are receiving this because you were mentioned.Message ID: ***@***.******@***.***>>

Leirbag-gabrieL · 2026-06-19T13:50:17Z

Hi,

sorry for responding so late, to answer your question @N-Dekker yes I totally took @alvarez-pa 's pull request as a strating point to do mine.
I see that you made the small little changes you asked me yourself. I was about the do a pull request of the changes you asked me here https://github.com/Leirbag-gabrieL/elastix/tree/pr_adjustments .

I think I should not do it since you just did this one #1453 right ?

N-Dekker · 2026-06-19T14:12:44Z

Thanks for your work, @Leirbag-gabrieL My intention is indeed to merge PR #1453, and then close both PRs of you and @alvarez-pa. But your commit is still there in PR #1453, so your name will remain in our git log history 😃

ENH: Add support for multiple metrics and regularization

ca7b49d

N-Dekker reviewed May 15, 2026

View reviewed changes

N-Dekker reviewed Jun 18, 2026

View reviewed changes

N-Dekker mentioned this pull request Jun 19, 2026

Support multiple metrics and regularizations, add tests and backward compatibility ("legacy hack") #1453

Open

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

ENH: Add support for multiple metrics and regularization#1437

ENH: Add support for multiple metrics and regularization#1437
Leirbag-gabrieL wants to merge 1 commit into
SuperElastix:mainfrom
Leirbag-gabrieL:main

Leirbag-gabrieL commented Apr 28, 2026

Uh oh!

N-Dekker commented May 15, 2026

Uh oh!

N-Dekker May 15, 2026

Uh oh!

N-Dekker May 15, 2026

Uh oh!

N-Dekker May 15, 2026

Uh oh!

N-Dekker commented May 21, 2026

Uh oh!

N-Dekker Jun 18, 2026

Uh oh!

stefanklein commented Jun 19, 2026 via email

Uh oh!

Leirbag-gabrieL commented Jun 19, 2026

Uh oh!

N-Dekker commented Jun 19, 2026

Uh oh!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

3 participants

Conversation

Leirbag-gabrieL commented Apr 28, 2026

Uh oh!

N-Dekker commented May 15, 2026

Uh oh!

N-Dekker May 15, 2026

Choose a reason for hiding this comment

Uh oh!

N-Dekker May 15, 2026

Choose a reason for hiding this comment

Uh oh!

N-Dekker May 15, 2026

Choose a reason for hiding this comment

Uh oh!

N-Dekker commented May 21, 2026

Uh oh!

N-Dekker Jun 18, 2026

Choose a reason for hiding this comment

Uh oh!

stefanklein commented Jun 19, 2026 via email

Uh oh!

Leirbag-gabrieL commented Jun 19, 2026

Uh oh!

N-Dekker commented Jun 19, 2026

Uh oh!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

3 participants