DBP:Harvard:Software:Bugs:Solved:DICOM 3-21-05

DTI scans acquired in DICOM are not reconstructed properly and cannot be viewed in Slicer

4-13-05 Test of new script

• The new script after Stephan's code modification is now working properly at least for caseD00779 and caseD00735, which had showed flipped images after reconstruction before.

1. run /projects/schiz/software/scripts/lsdi2dwi-all.pl <dir> <case#>

2. open FA images (35 slices) in slicer24 with setting of 0.8593 x 0.8593 x 5, orientation : coronal PA, scalar type : short, Little Endian : yes (*)

• The FA values shown on the screen in accordance to the pointing of mouse arrow seems to be relevant, e.g. corpus callosum area - 700-900, caudate nuc. 200-250, CSF (3rd ventricle) 0-50.
• Please note that the Little Endian should be yes for adequate image loading.

3-28-05 Update from Sylvain

• Raul and Sylvain have looked at it and it seems that this is a problem only Stephan Maier (or somebody in his group) can solve.

3-24-05 UPDATE: PROBLEM

• Three cases were tested for reconstruction of dicom type DTI image using lsdi2dwi.pl & lsdi2dwi-all.pl. This worked well for caseD00764, but for caseD00735 and caseD00779, the resulting images were flipped.

REMINDER***: D files created by lsdi2dwi.pl, and DWI by lsdi2dwi-all.pl, are identical!

• It should be possible to load DWI files into Slicer and "Convert," just as one would for D files

3-23-05: Solution

• Sylvain has provided a solution to this problem, which involves the following command:

• /projects/schiz/software/scripts/lsdi2dwi-all.pl <dir> <case#>

• This script will reconstruct all x.IMA files into ADC, FAI, DWI, etc. with the case# prefix
  • Note: DWI files are the same as "D" files
• A couple reminders:
  • If you wish to reconstruct images from the current working directory, be sure to put a "." for <dir>
  • Don't forget to include the case# in the command line