 |
 |
Thanks for posting this. It was helpful for me at least to see it explained again. Peace, Matt. On 1/30/15, 11:39 AM, "Ajay Kurani" <[log in to unmask]> wrote: >Dear FSL Experts, > I have cleared up the issue and was incorrect in my assumption >regarding the readout time for partial fourier. Below is an explanation >I received in case it benefits anyone else in the future: > >In your case the actual and effective echo spacing are both 0.46ms. >Hence the effective time for the readout is (0.46*76) ms. > >For example, if parallel imaging was used with a factor of two then >although the echo spacing of the acquisition would still be 0.46 ms, the >k-space interval is twice as big (we skip every other line) so for the >effective echo time (of having not skipped lines in k-space) one would >have to half this to 0.23ms. In this case we would use (0.46*76)/2 ms. > >As another example, consider phase over-sampling. In this case the >k-space interval is reduced in order to achieve the increased field of >view. This means that the effective interval (the k-space interval >corresponding to the final field of view) would be larger than 0.46ms, >e.g. for 12.5% oversampling we would use (1.125*0.46*76) ms. > >If we had both parallel imaging factor two and 12.5% over-sampling we >would combine the effects and enter (0.46*76*1.125)/2 ms. > >However, with partial Fourier even though you acquire less lines of data, >the interval in k-space is not altered, or equivalently the field of view >is not altered (you just stop acquiring data earlier). Therefore the >effective echo spacing is equivalent to the actual echo spacing of the >acquisition hence we use (0.46*76) ms. > >Thanks, >Ajay