Loading Image Data

The easiest way to load image data into rview is to simply select the image file and then drag it onto one of the display areas:
drag

Dropping the file loads the entire volume (all slices):

drop

Notes/Warnings:

For DICOM and other multi-file formats drag ONLY ONE slice file: rview will then check all the slice files to see which ones actually belong with this volume: do not drag multiple slice files into rview !
For DICOM it is advisable to arrange slices for different sequences into separate directories/folders. This will avoid the need for rview to check all the slice files in a folder containing many files to see if they are in the same volume as the selected slice.
You should select the image header file and not the raw data file for formats such as analyze or interfile. (the geometry is read first from the header and then the corresponding voxels are read from the datafile [.img file] )
Before loading the data, check the section on file formats to see if your image file format is supported.
It is advisable to check for obvious loading problems when first trying to load your data:

Check the orthogonal views to see if the slices are too thin or thick (organ looks stretched or squashed). (incorrect voxel dimensions) and check the voxel dimensions in mm using 'Scan Info' against what you expect for your data for the field of view/pixelspacing/slicegap and slice thickness: remember to take into account slice thickness and gaps for the Z voxel size.
Check if all the slices are loaded (from the top to the bottom of the volume)
Check if the slices are contiguous (no gaps or discontinuities) to see if any are missing.

Data Orientation and Display Convention:

For DICOM, NIFTI (with quaternion set) and some SPM/analyze format image datasets, rview can read the header and estimate whether the image data slices are primarily axial, coronal or sagittal. rview first loads the data into memory storing the data in a standardized axial, saggittal or coronal way within memory (i.e. reflections and slice order resolved). The Data Orientation transformation within rview is then set to one of three combinations of rotations that map the data (either axial, saggittal or coronal) to a standard reference coordinate system for display in rview.

Examples of the process of loading different acquisition types into rview when the data orientation is read from the header (eg DICOM or NIFTI)

The data Orientation Transformation parameters can be seen in the view tool. If the data has been loaded correctly into rview then the images should be displayed in radiology coordinates with, in the default display, the axial slice shown on the far left, and the left of that image corresponding to the right of the subject.

This shows the display of a NIFTI test dataset displayed in radiology coordinates (right of image is left of subject).

In addition to this there is a set of user controlled slice view and render view transformations that allow the user to view the volume data in different orientations. These are described in more detail in the page on the View Tool.

Floating Images and their Orientation with Respect to the Reference Image

When you load an image into the reference space, this sets the reference Data Orientation (as described above) which determines the view of the Reference Space. When you load a second image file as a floating image area, rview will attempt to read the orientation of that file and then set the 'starting estimate' of the transformation mapping from the reference image space to that floating image, which is used for the display of each floating image. As a result, the images should be displayed correctly with respect to the reference image dataset.

Here an image dataset with coronal acquisition orientation has been loaded into the reference space, and then an
Axial dataset has been loaded into the Floating Image A.

Notes on loading new data in place of old:

1. If you load another floating image, that overwrites an earlier loaded floating image, then the transformation estimate for Image A will, by default, not be updated from the headers of the new image: The registration estimate will be retained and applied to the new dataset. However, a recalculation of the transformation for a floating image from Image Headers can be forced using the Alignment Tool (From Data) option.

2. If you load a new Reference Image, this will update the data orientation and reference view. Thus, the transformation applying to each floating image may need to be updated, unless the new reference image is in the same coordinate system of the first reference image you loaded.