ImageJ - MRC CBU Imaging Wiki
location: ImageJ

Accessing scans from the Bruker WBIC 3T system

Usually scans will need to converted into another format to use with software packages (e.g looking at functional scans in SPM or coregistering a 3D scan with a CT). In all these cases it is important that the image orientation is correct (i.e. the package knows which side of the image is left). Some scans are acquired in non-axial orientations and particular care must be taken over these scans. If in doubt ask or perform a phantom scan on an object with a define L_R difference to double check.

For all that follows you will need to know the subject's WBIC number and the process number (scan number) of the scan for that session. The latter can be found on the MRI record sheet in the green folder and is the middle number in a sequence like: 1:1:1. All Bruker images have a number like this. The format is studynumber:processnumber:reconstructionnumber You will also need to put the data you are creating somewhere. Please make sure you use a useful directory name and names for your images - having 10 scans all called m00.img is not helpful. It is also best to write down any processing you do to images so that you can remember for next time or so things can be traced back.

3D scans into Analyze format

This will be used most often for coregistration with PET scans in SPM software. The majority of 3D scans should be acquired using the SPGR_standard protocol which acquires a 3D spoiled grass T1 weighted image with sagittal slices (resolution 1x1x2mm(through slice)).

This next part uses the ImageJ software package which has had plugins added in-house by Guy Williams. To use ImageJ, get a Unix prompt and type:

imagej

This will then (after a delay) bring up the main imagej window. Select the File - Import menu and choose Bruker loader. This will then generate a new window with a list of everything that is in the WBIC archive directory. Click on the subject/session no that you want. You are then asked for the Series number. The series number is the number of the individual scanning series - i.e there is one series for the 3d image (if you have one), one for each EPI session, another for phase maps etc. To know which series number you want you will need to have noted it during the scanning session, or have a copy of the radiographer's sheet where this is recorded. Type the number in the box and the image should load. A 3D file may take a little while but should eventually appear in a new window. Have a look at the scan and check it is the one you want.

If AND ONLY IF it is a sagittal MRI taken using the SPGR_standard protocol then you can use another plugin to rotate this scan into an orientation suitable for SPM software. Select the Image - Rotate menu and choose Sagittal to SPM plugin. This will then run (takes a couple of mins please be patient) and a new scan should appear. This should be a 'nose down' axial scan, with left on the left.

To save this in Analyze format select the file menu and save as Analyze format. This will create a new file in the directory of your choice (it asks you!). Beware that you should type the directory in at the top box and press return - you will then be able to see the contents of that directory.

This file is now in SPM Neurological orientation (ie x inc left to right, y inc post to ant, z inc inf to sup). Please make this VERY clear to whoever you send the images to. All scans should use a cod liver oil marker on the RIGHT hand side of the head. These look like little round bumps on the subject's scalp.

Functional images to Analyze format

When the functional data is reconstructed it must be turned in Analyze format. This can be done in ImageJ (see above). Choose the File - Import - Bruker loader from the top menu. A window will come up with subject/session numbers listed. By default this is a listing for the current CBU WBIC data archive - mounted as /cbu/imagers/wbic_data. This contains data that has been transferred recently.When you have found your data, select the subject/session number with a double click. Then select the scan you wish to visualise (see scanning sheet for details). Then run through the data quickly to check image quality etc when it pops up in the window. This will be a window paging through the slices volume by volume (so 20 slices for first time point followed by 20 slices for second time point etc). If the scanner bugged out - find the first naff image (usually stripy or absent) and record its number (you will need this later). Then select the fMRI time series from the Save as... menu - this writes the file as a 4D Analyze file (x, y, z, number of volumes or timepoints). Write it somewhere sensible and call it a sensible name (i.e. with the subject number and run for that imaging session info). When this image is written close the image before opening a new one or you'll run out of memory eventually.

The data can then be turned from the 4D file into the little timepoint files that SPM loves by running the ana4dto3d program. From the Unix prompt, run ana4dto3d. The format is:

ana4dto3d the4dfile.img

This is a good time to omit the dummy scans, by not unpacking them from the 4D file, using the -d option in ana4dto3d. For example, if you want to omit the first 5 scans:

ana4dto3d -d5 the4dfile.img

You can also unpack the 4D file into a directory other than that containing the 4D file by putting a directory name at the end of the command line (here also omitting 5 dummy scans):

ana4dto3d -d5 the4dfile.img ../the3dfiledirectory

Finally

You will also need to copy some key text files from your subject's directories for further reference.This requires a knowledge of the directory structure on the WBIC disk. Your data will normally be in a directory /cbu/imagers/wbic_data/ [subj:sessno]. From that directory, you will need a copy the text file "subject". Each series is in a separate subdirectory of this directory, e.g 1, 2 etc. For each series that you are interested in, you need a backup of the "imnd" file, and the "acqp" file in the series directory. Finally, the reconstructed data is kept in a subdirectory "pdata" in the series directory. As there may be more than one reconstruction, the data is in fact kept in a subdirectory of the pdata directory, 1, 2 ... You will also need a backup of the "reco" file in the [subj:sessno] / [seriesno] /pdata/ [reconstruction no] directory.

This is best achieved by changing directory to the subject/session no directory in /cbu/imagers/wbic_data, and running the following command from the Unix prompt:

tar cvf /imaging/matthew.brett/000190-0/textfiles.tar subject
*/imnd */acqp */pdata/*/reco

where /imaging/matthew/000190-0 is the directory you are storing your subject's data in. Not that in this case the subject session directory was /mridata/wbic_cbu/000190-m00.2i1 - where m00 means the first session ever for this subject, and noone seems to know what .2i1 means.We will try and automate this better, um, sometime.

This page created by Emma Williams, edited for the CBU by MatthewBrett. Last updated 20/11/00 (MyB)

CbuImaging: ImageJ (last edited 2013-03-07 21:24:04 by localhost)