Horizontally Gantry Tilt

Hi everyone
I’m a medical image processing specialist, and now I’m working on Dicom.
I have a question about Tilt in Dicom, which is defined using the (0018, 1120) Dicoms tag.
Based on my research, we have 3 types of tilts, Normal, Vertical, and Horizontal.
1- Is there anybody to tell me what are the differences between these types?
2- Why do the applications support the Vertical and Normal but they don’t support the horizontal?
I will be so thankful if you answer my questions.

These 3 types of tilts don’t sound familiar. Maybe it refers to “no tilt”, or rotation around horizontal axis and rotation around vertical axis. Rotation around horizontal axis it means towards the patient’s head or feet, this is what (0018,1120) DICOM standard can describes, and I have ever encountered in CT images. Rotation around vertical axis might be feasible, too, but would be harder to implement (there is not much space available to do that) and I don’t think it is commonly done (if at all).

Maybe gantry rotations that you have come across describe C-arm rotations (RAO/LAO, CRA/CAU), which are very different from Gantry tilt of a CT described by (0018,1120).

Thanks for answering dear Andras
I have seen these types in some applications, you might be right and (0018, 1120) is not the same thing that I’m working on.
My main problem is that I can’t get the MPR for some datasets, cause of this error: Does not support [horizontally gantry tilted images] as a source dataset for MPR.
I can’t figure out why there is no problem with vertical and normal images, but for horizontal?
In my point of view, the acquired images should be the same in the 3 directions, so what is the problem?
Is my question clear?

That doesn’t sound like a Slicer error. Where does it come from?

Slicer uses a more general method to account for gantry tilt (and other geometric issues like irregular slice spacing) so it doesn’t rely on the tag you referenced.

Thanks for answering dear Steve
The link you’ve sent is really helpful.

It seems that the message comes from ClearCanvas. The small illustration that they provide is very nice and clear:


However, I don’t think this “horizontal tilt” technique is used in clinical imaging. It may be hard to implement because the gantry would collide with the table, it would need an extra rotation joint for the gantry, etc.

What ClearCanvas documentation calls “vertical tilt” is used for clinical imaging (I’ve seen mostly in brain imaging) and this is what (0018,1120) refers to as well:


If you indeed use ClearCanvas then I would recommend to replace it with some more up-to-date software. If you host the DICOM server yourself then you can use DCM4CHE or Orthanc; if you are on the cloud then Google, Microsoft, etc. have their own hosted DICOM server implementations that you can rely on (and may save you some maintenance efforts). If you want to just have something simple then you may use DCM.js DICOMweb server or Slicer’s built-in DICOMweb server.