Hi,
Sorry, I think there's been some confusion here.
Let me start again...
The asym time (TA) determines how to convert phase difference into
field, in rad/s. The formula for this is:
FIELD = (PHASE DIFF) / (ASYM TIME) = (Phi1 - Phi2 ) / TA
So for a pair of gradient-echo scans with different TEs the appropriate
asym time would be the difference in TEs (i.e. TA = TE2 - TE1) as you
said.
Now I think what is confusing is what is meant by echo time when you are
not doing a standard spin-echo (eg. asymmetric spin-echo).
In this case it depends on how the sequence is set up.
Consider the following sequences ...
Case 1:
90 degree pulse - wait TE/2 - 180 degree pulse - wait (TE/2 + TA) -
centre of readout
In this situation the extra phase (compared to the symmetric spin-echo)
is due to
the extra TA wait, and hence the field is given by the above
relationship.
Case 2:
90 degree pulse - wait (TE/2 + TA) - 180 degree pulse - wait TE/2 -
centre of readout
This is like case 1, just with the opposite sign.
Case 3:
90 degree pulse - wait (TE/2 + TA) - 180 degree pulse - wait (TE/2 -
TA) - centre of readout
In this case the phase difference with the symmetric spin-echo is twice
that of
the above cases so that the correct setting for asym time in the
software is
now (2*TA).
I hope this is clearer than my previous efforts and I'm sorry for the
confusion
caused by my previous emails.
All the best,
Mark
On Tuesday, April 27, 2004, at 05:09 am, Minzhi Gui wrote:
> Hi, Dr Jenkinson
> So, if the gradient echo as you said "the asym time is simply the
> factor required to convert the phase difference in the scans to a
> rad/s field value", does that mean the asym time is the time causing
> the phase difference? if this is the case, I think, for gradient echo,
> the asym time should be (TE2-TE1), where TE2> TE1, since the phase
> accumulation difference is caused by the different TE. However, if I
> use this rule to your spin-echo case, the asym time should be twice of
> the time 180 degree pulse shifted, but actually this is not the case.
> Now, I am a little confused, could you please tell me why the time
> cause for your spin echo scan is only the time 180 degree pulse
> shifted? what should be the corresponding "asym time" for gradient
> echo sequence when using youe fugue?
>
> Maybe it's a stupid question, but I will really appreciate your help!
> Minzhi
>
>
>
> ----- Original Message -----
> From: Mark Jenkinson <[log in to unmask]>
> Date: Sunday, April 25, 2004 5:13 pm
> Subject: Re: [FSL] Using gradient echo sequence, which time should be
> the "asym time" in FUGUE ?
>
>> Hi,
>>
>> I've just responded to the question about using gradient-echo vs
>> spin-echo.
>> Unfortunately it is not easy or optimal to use our existing software
>> for this.
>> It is much better to use a spin-echo sequence.
>>
>> As for the asym time, it is the timing difference between the 180
>> degreepulses in the two spin-echo scans. The echo time (TE) is
>> kept the same
>> for both scans, but the 180 degree pulse is set at TE/2 for the
>> symmetric
>> scan (normal spin-echo) and set to (TE/2 + asym time) for the
>> asymmetricscan. If you do want to use gradient echo, then the
>> asym time is simply
>> the factor required to convert the phase difference in the scans
>> to a
>> rad/s field value. However, for gradient echo scans, be aware
>> that the
>> field map itself will be distorted and needs correction before
>> applyingit to the EPI.
>>
>> All the best,
>> Mark
>>
>>
>> On Wednesday, April 14, 2004, at 03:27 am, Minzhi Gui wrote:
>>
>>> Hi,
>>> Did anyone use the gradient echo sequence to undistort the
>> images> before, by using Prelude & FUGUE? I'm now wondering what
>> is the
>>> corresponding "asym time" for gradient echo sequence, is it
>> (TE2-TE1)
>>> or (TE2-TE1)/2 or somethinf else?
>>>
>>> Thanks
>>>
>>> Minzhi
>>
