There is something wrong with these dicoms. First, there is no Siemens "ASCII header", so something has been done to it or it was the result of something non-standard on the scanner. Second, the reason it cannot figure out what the direction cosine is for the slices is that all files have the same image position (tag 20 32). This is the real problem. As far as the dicom is concerned, you have a bunch of slices all in the same location and so it is not possible to compute the direction cosine in the slice direction. I also noticed that the ImageType (tag 8 8) is "DERIVED\SECONDARY" so maybe this file is the result of some non-standard post-processing on your scanner.
doug
Hemant Bokil wrote:
I am trying to convert some dicom slices to mgz but am getting errors. From the errors it appears that mri_convert is looking for some metadata and that this metadata is corrupted. The MRs themselves look fine when I view them in Matlab. The full screen dump is below:
mri_convert -it dicom -ot mgz DICOM_MR_DATA/MR2/MR20000.DICOM MR2 $Id: mri_convert.c,v 1.179.2.2 2011/05/16 20:53:47 greve Exp $ reading from DICOM_MR_DATA/MR2/MR20000.DICOM... Starting DICOMRead2() dcmfile = DICOM_MR_DATA/MR2/MR20000.DICOM dcmdir = DICOM_MR_DATA/MR2 Ref Series No = 3 Found 83 files, checking for dicoms Found 80 dicom files in series. First Sorting Computing Slice Direction Vs: 0 0 0 Vs: -nan -nan -nan Second Sorting Counting frames nframes = 80 nslices = 1 ndcmfiles = 80 PE Dir = ROW (dicom read) TransferSyntaxUID: --1.2.840.10008.1.2-- jpegUID: --1.2.840.10008.1.2.4-- Loading pixel data TR=40.00, TE=8.00, TI=0.00, flip angle=40.00 i_ras = (-1, -0, 0) j_ras = (-0, -1, 0) k_ras = (nan, nan, -nan) Reslicing using trilinear interpolation MRIresample(): error inverting matrix; determinant is -nan, matrix is: -1.094 -0.000 nan nan; -0.000 -1.094 nan nan; 0.000 0.000 -nan -nan; 0.000 0.000 0.000 1.000;
Freesurfer mailing list Freesurfer@nmr.mgh.harvard.edu https://mail.nmr.mgh.harvard.edu/mailman/listinfo/freesurfer