After registering, running your scan and sending all your data sets, log into webdb and click on the session you’ve just completed (n.b.: you cannot access the data before the end time you indicated when reserving the session.) You will be shown the following page that displays the series with a check box next to each series. Select the series you’d like to extract.
Chose an Output Disk. You will notice a pull-down menu next to Disk. It will look something like this:
Any option starting with a capital name and column, in this case "LARMOR:", indicates a disk on a remote machine. All other options indicate disks on the CBI file server. Here the Heeger lab is used as an example, each user will only see a list of disks that they have access to. The list can and will be modified when needed.
Fill in the Output Directory value.
You can specify a directory tree to create nested directories, e.g.: my_data/july1/first_run. Click on the Default button if you want to use the default directory (the accession number).
Organization indicates whether you want the data all in one directory or you want to split it with the data from each series in a different directory.
Unwarp images refers to the gradient non-linearity correction. It creates a folder called "unwarped" in each of the series with the unwarped version of the images. Use this option if you want to register images across different sessions (different subject positioning). For more information see CBI’s Gradient non-linearity correction paper (username and password required).
If you selected any Siemens data (online reconstructed) you will see this section. The Siemens Data Format can be converted to 3D Analyze, NIfTI (either single or dual file) or DICOM. All conversions will include a ".txt" file which is a dump of the DICOM header of the first slice.
The Raw Data Format will be available if you selected raw (offline) data to be reconstructed.
The Version drop-down menu allows you to reconstruct your data using older versions of the code,
or to just copy the raw data file ("Copy only" option). For a list of changes in the reconstruction for different versions, see below.
N.B.: Don’t use versions other than the default unless you are sure you know what you are doing (for example: if you started a study reconstructing images with one version and CBI releases a new version you might want to keep using the initial one).
Recontruction Type: if Corrected is picked and calibration scans are included in the series to be reconstructed, the images are reconstructed using the information in the first calibration scan: for Multi-Echo data, the full iterative non-linear reconstruction will be used; for regular EPI data, B0 unwarping will be performed. If there are calibration scans and Conventional is selected, the calibration scans are processed, but the images are reconstructed without using the information from the calibration scans. If no calibration scans are available the option is ignored.
Output File Type allows you to pick the output format. The options are:
Raw Data File allows you to indicate whether the original raw data is copied to your output directory or simply a link is placed there. The default is Link and very rarely will you want a Copy
Dry Run (mainly for debugging) will generate all the appropriate scripts, but the reconstruction is not actually executed.
Pressing OK will submit the reconstruction job(s) to the cluster. You will receive an email message to let you know the reconstruction has terminated.