Saving Images

You may want to save an image for one of three reasons:

The image pixel values have been altered, for example using the ROI Masking tool.
The image has been imported and you do not want to go through the import procedure again.
You want to change the image format, for example from Analyze^TM to NIFTI-1 format.

To save an image, select "Save As ..." from the File menu: file_save_as

You will first be prompted to select a file format for the saved image from image format chooser which will pop up:

save_as_chooser

Note: the Analyze^TM format does not permit the patient- and scan-specific information to be saved. The Analyze image header contains only the most rudimentary information about the image.

Note: The DICOM standard only allow images to be stored with integer-valued pixels (byte, short or colour images). In order to maintain interoperability, if you save an image with a floating-point datatype, the image intensity values are scaled to fit into a short (2-byte) format. The scaling will be done automatically, and information can be lost in this way. However, the original floating-point data are retained using private DICOM elements. The two private elements used are:

<0x0caf, 0x1011> VR=Short Text VM=1 Xinapse Systems Pixel Data Type
<0x0caf, 0x1021> VR=Other Float VM=1 Xinapse Systems Floating-Point Pixel data

If you wish to use floating-point images on other DICOM-compliant equipment, it is better to convert images to short int format before saving in DICOM format, perhaps using the Image Algebra tool to rescale them.

In addition, it is compulsory to store certain information in the image header, such as the patient name. If Jim cannot determine this information automatically, you will be prompted to supply it. However, you can configure how Jim behaves when it is trying to save an image in DICOM format, and the compulsory information is not known. This is done in the user Preferences (DICOM Setting tab).

You now have the opportunity to change the data type for the saved image:

save_as_datatype_chooser

Simply press the button to keep the same data type, or select a new data type from the menu. If you choose a new datatype that has a smaller dynamic range than the original image, you will receive the following warning:

save_as_clip_rescale

For example, if the original image has 16-bit short pixel values, these have a potential range of -37268..+32767. Selecting a unsigned 8-bit byte data (range 0..255) for the saved image would mean that not all pixel values can be stored. The options are:

Clip. Any pixel intensities outside the dynamic range of the new image are set to the extreme values of the dynamic range. In the example above, a pixel intensity above 255 would have an intensity of 255 in the output image; a pixel intensity below 0 would have an intensity of 0 in the output image.
Rescale. All pixel intensities are scaled with a constant scaling factor such that they all fit within the dynamic range of the saved image.

Next you will see a File Chooser to select the directory and file name for the saved image.

Note: if the image pixel values are altered by image editing operations, on unloading the image, you will always be prompted to save the image. At this stage you may chose either to discard the image edits, or to cancel whatever prompted the image unload.

Jim Home