Partial Fourier Data

[tknopp]

Hi @aTrotier,

I was wondering, if you did measure some partial Fourier data with a Bruker scanner, so that we can try reconstructing it. I have seen, that you included functions to read out the information, but if I understand correctly, we can't reconstruct right now.

[aTrotier]

I might have that somewhere or I can acquire a new dataset according to your needs.

I think the reconstruction works (CG-sense or CS) but we don't use the symmetry property of k-space.

Do you want to implement a more simple PF algorithm (homodyne / POCS) ?

[tknopp]

yes, we want to implement a simple direct reconstruction. But at the end I am also interested if there are any opportunities in CS based reconstruction when using partial Fourier. I simply do not know.

[tknopp]

@migrosser

[aTrotier]

The POCS algorithm is the most common one used on all the scanner. Not sure if we should let the user pass the low-res phase as in params["phase_map"] or directly estimate it in the reconstruction but we have less control over the smoothing/edge transition...

Anyway, maybe we should take a look about using the MRD label for calibration lines

[aTrotier]

Do you need a PV6.0.1, PV360.3.5 dataset or both ?

[tknopp]

I think the new PV version would be sufficient. Important would be to include a reconstruction, so that we can compare.

[aTrotier]

I think in pv6.0.1, the partial Fourier undersampling was only reconstructed with zero filling algorithm. Now pocs is available. I'll acquire a dataset next week. Le ven. 1 déc. 2023 à 21:26, Tobias Knopp ***@***.***> a écrit :

…

I think the new PV version would be sufficient. Important would be to include a reconstruction, so that we can compare. — Reply to this email directly, view it on GitHub <#164 (reply in thread)>, or unsubscribe <https://github.com/notifications/unsubscribe-auth/AC5P7O5IXWTF532GMTHN373YHI4PVAVCNFSM6AAAAABAA53W3OVHI2DSMVQWIX3LMV43SRDJONRXK43TNFXW4Q3PNVWWK3TUHM3TOMZTHA4TE> . You are receiving this because you were mentioned.Message ID: <MagneticResonanceImaging/MRIReco.jl/repo-discussions/164/comments/7733892 @github.com>

[tknopp]

interesting. So do you have both versions installed and could potentially convert? If that is not to much of an effort, both versions would of course be helpful.

[aTrotier]

We have a 4.7T with pv6.0.1 and a 7t with pv360. I don't think I can transfert dataset between both version. Le sam. 2 déc. 2023 à 09:07, Tobias Knopp ***@***.***> a écrit :

…

interesting. So do you have both versions installed and could potentially convert? If that is not to much of an effort, both versions would of course be helpful. — Reply to this email directly, view it on GitHub <#164 (reply in thread)>, or unsubscribe <https://github.com/notifications/unsubscribe-auth/AC5P7O2ZP3A5F7XEBA63XC3YHLOVRAVCNFSM6AAAAABAA53W3OVHI2DSMVQWIX3LMV43SRDJONRXK43TNFXW4Q3PNVWWK3TUHM3TOMZWG4ZTO> . You are receiving this because you were mentioned.Message ID: <MagneticResonanceImaging/MRIReco.jl/repo-discussions/164/comments/7736737 @github.com>

[aTrotier]

The bruker shared dataset has a PF acquisition reconstructed with POCS I think