0

I know the basic flow or process of the Image Registration/Alignment but what happens at the pixel level when 2 images are registered/aligned i.e. similar pixels of moving image which is transformed to the fixed image are kept intact but what happens to the pixels which are not matched, are they averaged or something else?

And how the correct transformation technique is estimated i.e. how will I know that whether to apply translation, scaling, rotation, etc and how much(i.e. what value of degrees for rotation, values for translation, etc.) to apply?

Also, in the initial step how the similar pixel values are identified and matched?

I've implemented the python code given in https://simpleitk.readthedocs.io/en/master/Examples/ImageRegistrationMethod1/Documentation.html

Input images are of prostate MRI scans:

Fixed Image Moving Image Output Image Console output

The difference can be seen in the output image on the top right and top left. But I can't interpret the console output and how the things actually work internally.

It'll be very helpful if I get a deep explanation of this thing. Thank you.

bl pooja
  • 1
  • 1
  • Did you read the [registration overview](https://simpleitk.readthedocs.io/en/master/Documentation/docs/source/registrationOverview.html#lbl-registration-overview) of simple itk before? – NoDataDumpNoContribution Apr 29 '19 at 08:34
  • Yes, but I didn't understand what happens to the image pixels which are mapped and which aren't when registration happens @Trilarion – bl pooja Apr 29 '19 at 09:11
  • The registration overview specifies the available transformations. These transformations are applied to all(!) pixels of the registered/aligned image. Usually these transformations are kind of affine transformations. Your console-output for example says that you have a simple translation transformation. – NoDataDumpNoContribution Apr 29 '19 at 10:56
  • Thank you for replying. Yes, I've used simple translation transformation but I'm still in the confusion state regarding the pixel/intensity level registration. I want to know what happens to the pixels which are mapped from both the images, i.e. whether they are kept intact or there values remain the same in the output image and what happens to the pixels which don't map, whether there average is taken or they are discarded in the output image. @Trilarion – bl pooja Apr 29 '19 at 12:56

1 Answers1

1

A transformation is applied to all pixels. You might be confusing rigid transformations, which will only translate, rotate and scale your moving image to match the fixed image, with elastic transformations, which will also allow some morphing of the moving image. Any pixel that a transformation cannot place in the fixed image is interpolated from the pixels that it is able to place, though a registration is not really intelligent.

What it attempts to do is simply reduce a cost function, where a high cost is associated with a large difference and a low cost is associated with a small difference. Cost functions can be intensity based (pixel values) or feature based (shapes). It will (semi-)randomly shift the image around untill a preset criteria is met, generally a maximum amount of iterations.

What that might look like can be seen in the following gif: http://insightsoftwareconsortium.github.io/SimpleITK-Notebooks/registration_visualization.gif

Czorio
  • 105
  • 2
  • 7