Question
Question Posted 12/04/15:
Dear Expert,
We are trying to preprocess the DTI data downloaded from ANDI and we are following the steps suggested here: http://web.stanford.edu/group/vista/cgi-bin/wiki/index.php/Preprocessing_DTI_Data_Using_FSL
The 3rd step is "calculate the Tensors (~5min)". Actually this is bottleneck for us since we couldn't extract bval and bvec files from the data we downloaded from ADNI (using dcm2nii). But we are wondering if we really need this step or not. Is the data provided by ADNI already the tensor that mentioned in the Stanford website?
Thanks in advanced
Dear Expert,
We are trying to preprocess the DTI data downloaded from ANDI and we are following the steps suggested here: http://web.stanford.edu/group/vista/cgi-bin/wiki/index.php/Preprocessing_DTI_Data_Using_FSL
The 3rd step is "calculate the Tensors (~5min)". Actually this is bottleneck for us since we couldn't extract bval and bvec files from the data we downloaded from ADNI (using dcm2nii). But we are wondering if we really need this step or not. Is the data provided by ADNI already the tensor that mentioned in the Stanford website?
Thanks in advanced
Response posted 12/11/15 by Rob Reid:
Hi,
Usually dcm2nii is the best route for getting .bval and .bvec files. If it did not produce them it means it did not recognize the DICOM files as a diffusion sequence. The most common cause we see in our lab is failing to download one or more of the slices. In that case dcm2nii cannot make a 4 dimensional image and will instead stack the heads on top of each other. If you are sure you have the right number of slices, check the Series Description with a DICOM reading tool like dcmdump or dicombrowser.
I hope this helps.
Usually dcm2nii is the best route for getting .bval and .bvec files. If it did not produce them it means it did not recognize the DICOM files as a diffusion sequence. The most common cause we see in our lab is failing to download one or more of the slices. In that case dcm2nii cannot make a 4 dimensional image and will instead stack the heads on top of each other. If you are sure you have the right number of slices, check the Series Description with a DICOM reading tool like dcmdump or dicombrowser.
I hope this helps.
