Input Images

Fat/Water Separation requires complex data (real/imaginary or magnitude/phase) in order to work.

Several formats of complex input image data are automatically recognised:

  • Separate real and imaginary components in different (real-valued) image slices.
  • Separate magnitude and phase components in different (real-valued) image slices.
  • Images with a Complex pixel data type (each pixel value is stored as a pair of adjacent 32-bit floating-point values). The program does not work with double-precision (64-bit) complex images.

If real, imaginary, magnitude and phase images are all present, then the program will only use the real and imaginary components of the image.

All slices/echo times can be in a single input image. Alternatively, each echo time can be stored in a separate image. Finally, each complex component and/or echo time can be stored in a separate image. The program is able to automatically detect these different situations.

For images that have a Complex pixel data type, the images data should structured as follows:

  • If there is a single input image, then all echos for the first slice location should be followed by all echos for the second slice, followed by all echos for the third slice, etc.
  • If there are multiple input images, then all echos for the first slice should be in the first image file. All echos for the second slice should be in the second file, followed by all echos for the third slice, etc.

You may find the Slice Extractor, Image Concatenator, and Image Interleaver useful for helping to organise your images into the required format.

For images that have complex components in different image slices (real & imaginary, or magnitude & phase) then the program should be able to determine the organisation of the data automatically, as long as the images are in DICOM format, or the have been converted using Jim's Image Converter from a DICOM source.

Note: you can create images with a Complex pixel data type from images that have real & imaginary or magnitude & phase components in separate images using the To Complex tool.

When you have set the processing options and the input images, press the Button to start quantitative fat/water separation button to start the processing. Processing will likely take several minutes to complete due to the complexity of estimating the B0 inhomogeneity.