History of AFNI updates  

|
March 03, 2003 11:53AM
Hello,
We have encountered a problem likely due to our scanner, nonetheless for converting data it means making a change to 'to3d'. We can make a change to the source code to work around this, but wanted to bring up the issue in case other labs had the same problem.

In converting we get, for example:

++ Setting environment variable AFNI_SLICE_SPACING_IS_GAP ++
++ to YES will make the center-to-center slice distance ++
++ be set to Slice_Spacing+Slice_Thickness= 6.000. ++
++ This is against the DICOM standard [attribute (0018,0088) ++
++ is defined as the center-to-center spacing between slices, ++
++ NOT as the edge-to-edge gap between slices], but it seems ++
++ to be necessary for some GE scanners. ++
++ ++
++ This correction has been made on this data: dz= 6.000. ++


For this dataset, the slice thickness dicom field is not equivalent to the center to center slice distance field. (eg., sp=2.9999998, th=3).

so, in mri_read_dicom.c, the line (444):

if( sp < th ) dz = sp+th ; /* the stupid GE way again */

sets dz to sp+th.

this is not actually desirable for us, because sp should equal th.

We will implement a fix to this situation (ie. disable this line in the case sp and th are very close).

If other labs have a similar problem, maybe adding a flag to disable this line would be worthwhile.

George
Subject Author Posted

slice thickness, conversion

George Tourtellot March 03, 2003 11:53AM

Re: slice thickness, conversion

bob cox March 03, 2003 04:37PM

Re: slice thickness, conversion

bob cox March 03, 2003 05:17PM

Re: slice thickness, conversion

George Tourtellot March 04, 2003 10:45AM