Hello-

Our site just acquired a new Siemens 1.5 scanner last week and I am having some difficulties reading in the Siemens Mosaic epi images. I am running the latest release of afni (8/6- precompiled binaries glibc_21 version) on a caldera platform. The data was collected I-S, with 35 slices, 82 volumes, 220 FOV, 4 mm slice thickness. When I enter the following command

to3d -time:tz 82 35 3000 alt+z Ser*

I receive the following feedback

to3d: 2D slices into 3D datasets for AFNI, by RW Cox
Counting images: total=2870 2D slices
++ DICOM WARNING: file SerieMR-0021-0001.dcm has Window tags; setenv AFNI_DICOM_WINDOW YES to enforce them
++ DICOM NOTICE: 6x6 Siemens Mosaic of 35 64x64 images in file SerieMR-0021-0001.dcm
Reading images: ++ DICOM WARNING: file SerieMR-0021-0036.dcm has Window tags; setenv AFNI_DICOM_WINDOW YES to enforce them
++ DICOM NOTICE: 6x6 Siemens Mosaic of 35 64x64 images in file SerieMR-0021-0036.dcm
.++ DICOM WARNING: no more Window tags messages will be printed
++ DICOM NOTICE: no more Siemens Mosiac messages will be printed
.........................................
Command line TR=3000ms ; Images TR=3s
Making widgets.....


The main problem is that the images in the to3d widget are interleaved. I have tried changing my .afnirc file to AFNI_SIEMENS_INTERLEAVE = YES and I have also tried entering the
‘-sinter’ command directly on the to3d command line. None of these changes (or combination of changes) seem to produce any differences in the way that to3d reads the image files.

I also looked at the first image in the series using

afni –im ‘first_image’

When I do this, the axial slices appear in the spatially correct order from I-S (they are not interleaved) but they appear under the sagittal viewer. This makes me think that they are not interleaved in the file - which is very confusing as they become interleaved after to3d reads them.

The other minor possible problem that I noticed is with the offset for the y-coordinate. The FOV was 220 but for some reason the offset for the y coordinate = 137.5 instead of the expected 110. The x coordinate is correct (110) and the z coordinate corresponds with the value listed for the first slice in the DICOM file. Has anyone else experienced this with the Siemens file? I know that this can be fixed on the command line but I thought it was odd that to3d would get two out of three offsets correct.

I have looked at the many old messages that I could find on this topic both on the afni message board and elsewhere on the net. It seems like there are many interface problems with reading the Mosaic files. Short of using the imcut program or another software to first read the files and put them into something like an analyze format is there a solution? I am not sure if the problems that I am having are due to the fact that this is a newer machine and therefore possibly a newer version of Mosaic? or if I should try something else? or if I am just doing something wrong.

Thanks in advance.

Andy Mayer