To call the POSSUM GUI, type Possum in a terminal (type Possum_gui on Mac), or run fsl and press the POSSUM button.
Now set the filename of the 4D input image (example in the GUI is $FSLDIR/data/possum/brain.nii.gz). You can view the image by pressing the Preview image tab (you will see an image of only one tissue type, e.g. grey in the case of the example given in the GUI).
Now choose the tab Pulse sequence. Default values are set already. TE is the echo time (in seconds), TR is the repetition time between the RF pulses (in seconds) and TRslice is the time it takes to generate a single slice (in seconds). TRslice is set with the Autoset to be TR divided with the number of slices (Number of Voxels: Z). You can choose to unset the Autoset and to specify your own TRslice.
You can then choose the number of your in-plane image voxels by setting X and Y values in Number of Voxels. The number of slices in your simulation is chosen by setting Z. By changing these values you will notice that the Approximate run time of POSSUM changes. This is because as you are increasing the number of points in your image matrix or increasing the number of slices the number of calculations the program needs to do increases, so the run time increases (or the other way around).
You can also change the size of your in-plane image voxels by setting the X and Y values in Voxel Size. The slice thickness is chosen by setting Z. As you are changing this you will notice that the values in the Field of View are changing. This is because the field of view is equal to the product of the number of voxels and the voxel size. Both voxel sizes and the field of view are in millimetres.
Now specify the Number of Volumes, the size of the Gap (in millimetres) between the slices and the read-out bandwidth BW. Echo spacing is showing the time it takes between the centres of two neighbouring k-space lines.
Now specify gradient directions for the Read gradient, Phase encode gradient, and the Slice select gradient. Direction of the acquisition during the slice selection is set by choosing an appropriate option from the Slice select direction. For scanning from the top to the bottom of the head choose "-" and for the scanning from the bottom to the top of the head choose "+" (this is when the slice selection gradient is Z). The images in both cases will be the same. The differences can occur in case the subject moves. The motion effects (including the spin history effects) if the subject is moving "along" the slice acquisition will be different from when the subject is moving "against" the slice acquisition.
Now specify which area of the brain you want to image by setting the Starting slice position. You can click on the Preview slice prescription in order to see the extent of your chosen field of view (FOV).
Finally, when you finish all of the pulse sequence specifications, you can check if everything is consistent within the pulse sequence by using the button Consistency check.
Now choose the tab B0 field. You can choose between the 1.5T and 3T Field strength. Then you can choose the input file for the MR parameters (default examples are $FSLDIR/data/possum/MRpar_1.5T and $FSLDIR/data/possum/MRpar_3T).
Now choose if you want to have susceptibility induced B0 field inhomogeneities by selecting either "None" or "Custom file". If you choose the "Custom file", specify the base name for your B0 field inhomogeneities basis set (e.g. $FSLDIR/data/possum/b0_ppm_), and choose the appropriate Units for your files.
Now choose the tab Motion. Press button where it says None and choose if you want to have no motion or to specify a custom file (e.g. $FSLDIR/ data/possum/motionRzLarge_0.12s).
Now choose the tab Activation. Press button where it says None and choose if you want to have no activation or to specify a custom file. When specifying the custom file, there are two files that need to be specified: T2* time course (e.g. $FSLDIR/data/possum/activation3Dtimecourse) and T2* spatial modulation (e.g. $FSLDIR/data/possum/activation3D.nii.gz). The second one can be viewed with the Preview image.
Now choose the tab Noise. Press button where it says None and choose if you want to have no noise or to have Thermal (white) noise. When specifying the white noise you can do so by either specifying the SNR or the absolute intensity of the noise (standard deviation).
Now choose the tab Output and specify the name of the simulation directory. This directory will contain all of the input, log and output files.
If you have the SGE environment on your cluster (ask the IT officer about this) you can specify a number of processors POSSUM can run on. This will significantly reduce the time of running. If you do not have the SGE environment, the number of processors should be 1.
When POSSUM setup is complete press the Go button. The setup gets saved in a temporary POSSUM setup file. Then a script (called possumX) is run which uses the setup file and carries out all the simulation steps asked for, starting by creating a POSSUM results directory, and copying the setup file into here, named design.fsf (this setup file can be later loaded back into POSSUM using the Load button). You can create this file also by using the Save option.
Once the script has started running you can Exit the POSSUM GUI. The analysis will continue until completion. Your result at the end is called image_abs.nii.gz and you can see it by opening it in the fslview.
If you need to see these documentation files press the Help button.