Partial Fourier issue

[aTrotier]

I am reconstructing a 3D FLAIR sequence. I have an issue with sense reconstruction.

Diving a little bit it is an issue with how the kspace is filled

raw_imaging = remove_raw_by_flags(raw,"ACQ_IS_NOISE_MEASUREMENT")
acq_imaging = AcquisitionData(raw_imaging)
kspace = kDataCart(acq_imaging)
heatmap(kspace[128,:,:,1,1,1] .|> abs .|> log)

## let's see what happens if create my own kspace
s = raw_imaging.params["encodedSize"]
raw_imaging.params["enc_lim_kspace_encoding_step_1"]
raw_imaging.params["enc_lim_kspace_encoding_step_2"]

nCh = raw_imaging.params["receiverChannels"]

kspace_2 = zeros(ComplexF32,s...)

sR = size(raw_imaging.profiles[1].data)
for (i,p) in enumerate(raw_imaging.profiles)
    ky =  Int(raw_imaging_motion.profiles[i].head.idx.kspace_encode_step_1) +1

    kz =  Int(raw_imaging_motion.profiles[i].head.idx.kspace_encode_step_2) +1

    kspace_2[:,ky,kz] = p.data[:,1]
end
heatmap(kspace_2[128,:,:] .|> abs .|> log)

I get the same results with kDataCart or doing it by hand

The encoding limits gives :

which corresponds to the maximum value filled along ky and kz

[aTrotier]

If I correct the ky indices (not the kz ones) : ky = Int(raw_imaging_motion.profiles[i].head.idx.kspace_encode_step_1) +1 + (s[2] - ky_lim.maximum)

I get :

I don't know if It is an issue with the way the mdh are labeled in the sequence or during the conversion to ISMRMRD

If I try to use the enc_lim center does not store the right value (in blue the center and red the maximum of enc_lim along ky/kz)

[aTrotier]

mapVBVD does not handle the reconstruction size and create a kspace according to enc_lim:

If I take a look at the metadata you can find 2 fields that stores the center lin/partition : twix.image.centerPar and twix.image.centerLin

I will cross link this issue on siemens_to_gadgetron

[aTrotier]

ismrmrd/siemens_to_ismrmrd#142

During conversion the correct center is displayed in the terminal

[aTrotier]

This information is stored (for each profiles) in

julia> raw_imaging.profiles[1].head.idx.user[6] |> Int
91

julia> raw_imaging.profiles[1].head.idx.user[7] |> Int
95