Thaks!
All of the datafiles except the slice profile are the ones which came
with POSSUM. There the object file has resolution of 1mm x 1mm x 1mm
voxels. The b0 fields are having the same resolution as well as
activation volume.
The used pulse variables are following (from pulse.readme):
1. SeqType = epi
2. TE = +3.200000e-02(s)
3. TR = +3.692000e+00(s)
4. TRslc = +7.100000e-02(s)
5. Nread = +52
6. Nphase = +52
7. dread = +3.000000e-03(m)
8. dphase = +3.000000e-03(m)
9. maxG = +5.500000e-02(T/m)
10. RiseT = +2.200000e-04(s)
11. BWrec = +1.600000e+05(Hz)
12. Nvol = +7
13. Nslc = +52
14. SlcThk = +3.000000e-03(m)
15. SlcDir = z+
16. Gap = +0.000000e+00(m)
17. zstart = +0.000000e+00(m)
18. FlipAngle = +9.000000e+01(degrees)
19. PhaseDir = y+
20. ReadDir = x+
21. kspace coverage (phase) = +1.000000e+02(%)
22. first non-zero phase line = +1
and from pulse command (from possum.log)
--te=0.032 --tr=3.692 --trslc=0.071 --nx=52 --ny=52 --dx=0.003
--dy=0.003 --maxG=0.055 --riset=0.00022 --bw=160000 --numvol=7
--numslc=52 --slcthk=0.003 --zstart=0.0 --seq=epi --slcdir=z+
--readdir=x+ --phasedir=y+ --gap=0.0 -v --cover=100.0 --angle=90.0
Do you need more information?
Mark Jenkinson wrote:
> Dear Juha,
>
> We really need more information than this.
> What values have you put into your slice profile and
> what object voxel sizes are you using in the simulator?
> The size of the object voxels (that is, the voxels used to
> represent the object - not the voxels that appear in the
> output image) is critical in determining how the RF field
> is applied, as each object voxel sees a single value of
> the RF field. Therefore if your slice is very narrow but
> your object voxels are not then the simulation will not
> work correctly.
>
> Please give us some more information and we will try
> to help.
>
> All the best,
> Mark
>
>
> On 8 Nov 2010, at 11:55, Juha Pajula wrote:
>
>> Hi!
>> Could You help me once again with the slcprofile of POSSUM simulator?
>>
>> I got instructions from you how to define our own slice profile for POSSUM simulator.
>>
>> I created one according to the instructions and verified it by our fMRI technician. I have tested it with our own data as well as the default data from POSSUM. It won't work. The same simulation with the possum defaults works, but when I change the slcprofile to our own the output is a mess.
>>
>> Earlier instructions were following:
>>
>> "The slice profile is simply truncated at a convenient point where
>> the amplitude is less than 0.3% of the maximum - since it needs
>> to be truncated at some arbitrary point. The frequency
>> difference is normalised as a percentage of the Larmor frequency. "
>>
>> Should the profile have some defined length? Our profile has really narrow band. Can that cause the problem?
>>
>> Results are looking like this (all data is possum default, except the slcprofile):
>> http://www.cs.tut.fi/~pajula2/possum.png
>>
>>
>> Waiting your answer,
>> --
>> Juha Pajula,
>> Research Assistant,
>> Methods and Models for Biological Signals and Images group of Signal
>> Processing department in Tampere University of Technology,
>> Finland
>>
--
Juha Pajula,
Research Assistant,
Methods and Models for Biological Signals and Images group of Signal
Processing department in Tampere University of Technology,
Finland
|