Transforming of points between images having different voxel sizes

[cto...@gmail.com]

Hi all,

I've browsed through the Elastix manual looking for details on how Transformix deals with images having very different voxel dimensions but haven't managed to find it.

I've been playing around with DICOM image stacks and have managed to perform image registrations and used Transformix to transform points across the two domains.

I understand that  "outputpoints.txt" provides the indices ( "OutputIndexMoving" ) of the transformed points.

I'm curious to know what Elastix does for the case where, for example, the Moving image stack has half the slice thickness as the Fixed image stack.  Since any one voxel in the Fixed image maps to more than one voxel in the Moving image, presumably a decision is made as to what voxel in Moving image to transform the point to.

Likewise, if Moving image has double the slice thickness as Fixed image, presumably more than one voxel in Fixed image will map to the same voxel in Moving image?

It would be very helpful to know whether it's possible to know which voxels in Moving image coincide with any given voxel in Fixed image, i.e. a point at (x_F, y_F, z_F) in the Fixed image domain that corresponds to voxel (i_F, j_F, k_F) maps to voxels (i_M1, j_M1, k_M1), (i_M2, j_M2, k_M2), ..., and hence points (x_M1, y_M1, z_M1), (x_M2, y_M2, z_M2),... in the Moving image domain.

Also if anyone has any DICOM test datasets they could share, or knows of where I could obtain open-source DICOM datasets* that contain DICOM series of the same patient imaged with different pixel shape, or slice thickness, or ImageOrientationPatient (one variable changed at a time) that would be extremely useful, as I could see what results I get in outputpoints.txt for different cases.

Thanks.

Note:

* I am aware of sites where anonomised DICOMs can be found (e.g. www.cancerimagingarchive.net, demo.softneta.com) but I could spend an eternity looking for collections that specifically have the above mentioned single variable changes.  Most sets have a combination of different pixel shape, slice thickness, image position, orientation, etc.

[Stralen-2, M. van]

Hi,

 

Good question. The influence of difference in image resolution can be overlooked and can be relevant if it differs largely between fixed and moving images.

 

This behavior of elastix is influenced by the ImageSampler and Interpolator used.

Depending on the choices there, samples are taken on grid (exactly on voxel positions), or at random continuous coordinate positions, where then the Interpolator is used to sample the intensity at the location in between voxel positions. The interpolator determines how much of the neigghborhood of that location is taken into account in determining the intensity value. The combination of the sampling and the interpolation thus determines the behavior. RandomCoordinate sampling already solves most of the potential issues I think.

 

But indeed, using grid sampling in a scenario where the fixed image has much lower resolution than the moving image, it might result in the moving image information being only partially used, because the sampling only sees the corresponding voxels of the ‘sparse’ sampling grid, defined by the fixed image. (in my understanding)

 

Cheers,

 

Marijn van Stralen

Assistant Professor Medical Imaging

Image Sciences Institute | Center for Image Sciences | University Medical Center Utrecht | The Netherlands

+31 (0) 6 10505649 | https://www.isi.uu.nl | https://www.umcutrecht.nl/cis

m.vanst...@umcutrecht.nl | https://umc-utrecht.webex.com/meet/van.stralen

 

Visiting address || Heidelberglaan 100, Room Q02.4.46 | 3584 CX Utrecht

Postal address || Huispostnummer Q02.4.45 | PO Box 85500 | 3508 GA Utrecht               

--
You received this message because you are subscribed to the Google Groups "elastix-imageregistration" group.
To unsubscribe from this group and stop receiving emails from it, send an email to elastix-imageregis...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/elastix-imageregistration/b2331df0-b68a-482f-ab6d-154e1966711cn%40googlegroups.com.

---

De informatie opgenomen in dit bericht kan vertrouwelijk zijn en is uitsluitend bestemd voor de geadresseerde. Indien u dit bericht onterecht ontvangt, wordt u verzocht de inhoud niet te gebruiken en de afzender direct te informeren door het bericht te retourneren. Het Universitair Medisch Centrum Utrecht is een publiekrechtelijke rechtspersoon in de zin van de W.H.W. (Wet Hoger Onderwijs en Wetenschappelijk Onderzoek) en staat geregistreerd bij de Kamer van Koophandel voor Midden-Nederland onder nr. 30244197.

Denk s.v.p aan het milieu voor u deze e-mail afdrukt.

---

This message may contain confidential information and is intended exclusively for the addressee. If you receive this message unintentionally, please do not use the contents but notify the sender immediately by return e-mail. University Medical Center Utrecht is a legal person by public law and is registered at the Chamber of Commerce for Midden-Nederland under no. 30244197.

Please consider the environment before printing this e-mail.

[cto...@gmail.com]

Thank you for your helpful explanation.  It's much appreciated.