Dear Varina,
Despite I dare to reply, I am not sure I qualify as expert.
1. Standard Philips field mapping protocol acquires very low resolution data and interpolates images on output. Make sure the acquisition is not too coarse. Use fat suppression.
2. DICOM viewer will show second image in Hz, but internally, it is saved as integer. DICOM file has two conversion coefficients, offset and slope to be applied to integers and obtain Hz. I am not aware of any DICOM to NIFTI converter that can apply such scaling. The offset is easy to see: outside of brain must have field 0Hz. After conversion to nifty, this value will be around 1024 or so. Now you need slope.... Probably a good guess is that the min and max frequency is given by delta TE of acquisition. These must correspond to 0 and 4095 ?
Roman
________________________________________
From: FSL - FMRIB's Software Library [[log in to unmask]] on behalf of Varina Wolf [[log in to unmask]]
Sent: Sunday, March 18, 2012 3:14 PM
To: [log in to unmask]
Subject: [FSL] Field Map Question
Dear Experts,
I have Phillips 3T aquired magnitude and field map images in DICOM format, with the questions below:
1. To BET the magnitude image, what parameters do you recommend? The standard ones, leave the skull inplace, likely secondary to the lack of contrast between scalp and brain in this image type.
2. My Field Map image is in Hz, what is the best way to find the appropriate converstion factor to render it in rad/s units?
3. In using the dcmniigui to convert the original field map dicom images, I have tried various iterations to which Nifti type to convert to with the results:
- FSL 4D Nii: one big file with the magnitude and Field Map dumped in
- SPM 3D Nii: two separate files not recognized by fsl programs (.img)
- 4D Nii/hdr: one big file but recognized by at least BET so far.
Any recommendations?
Thanks in advance,
Varina
|