 |
 |
Hi Herve, you should never use 2 nex, otherwise you loose the ability to qc your images. The b0 mapping shouldd be done with alternate phase encoding so that you can use one of the available tools to correct for distortions. We have elected to acquire smaller b values and repeat each volume with opposing phase encoding in order to correct for distortions. Our experience is that 2 mm isotropic is really all you can go for in order to maintain reasonable snr in the center of the brain with 32 channels. This is particularly true for high b values. Another problem is that with more coils the signal from csf is very bright and this creates ringing in brain tissue, however this problem may be highly dependent on what acceleration you use and the details of the recon (my experience is on a Siemens system). Have you considered multi-band? If GE has a product to accommodate that, it could cut your time of acquisition in half and allow you to collect inverse polarity volumes for better distortion correction. Stefano Sent from my BlackBerry 10 smartphone. Original Message From: Herve Lemaitre Sent: Friday, January 29, 2016 5:36 AM To: [log in to unmask] Reply To: FSL - FMRIB's Software Library Subject: [FSL] DTI sequence 3T GE Hi, We are setting up a new dti sequence on a 3T GE scanner. We are trying to match the recent recommendations concerning dti parameters and our limitation in terms of time for scanning pediatric patients (around 10 mins). We ended-up with this set-up: 4 b=0 , 71 directions using (http://www.emmanuelcaruyer.com/q-space-sampling.php) on three shells: 1000, 2000 and 3000 b-value? 2 x 2 x 2 mm resolution? 12 or 32 channel head coils? 1 or 2 NEX? B0 mapping to correct for the distorsions Do you think it is reasonable to use this kind of acquisition or do we still need to stick to a more classical sequence? Thanks for your answers. Hervé Lemaître