Standard Bruker datasets for PV360 versions

[aTrotier]

Hi,
Just for information :

During a meeting with the bruker team and the SAIN (Small Animal Imaging Network), we have raise the issue about the multiple version of PV360 released every 6 month or so. I asked bruker to release standard datasets to help the developers to validate their tools here.

A page is now online with multiple datasets and we can give feedback if we need more specific acquisition :
https://www.bruker.com/protected/en/services/communities/pci-community/paravision-versions/paravision-standard-datasets.html

For now, I will validate the reconstruction locally
Aurélien

[aTrotier]

Already have some errors with jcampdx files for some of them...

[aTrotier]

solved by : #202

[aTrotier]

Next error seems to be due to Zero-filling or partial fourier in the read direction. Under investigation !

[aTrotier]

The matrix is 384x384 with PF / ZF which reduced it to 200x288

The trajectory during acq conversion is tr = trajectory(Float32, "Cartesian", f.params["encodedSize"][2],f.params["encodedSize"][1]) -> 384 x 384 but the corresponding subsampling does not match

[aTrotier]

I try to find a solution about the PF + ZF.

In the ismrmrd format they are using additional fields in the encoding part : https://ismrmrd.readthedocs.io/en/latest/_downloads/791fd76184e16aadf992d3b4932aad8c/ismrmrd_example.xml

<encoding>
<trajectory>cartesian</trajectory>
<encodedSpace>
<matrixSize>
<x>256</x>
<y>140</y>
<z>80</z>
</matrixSize>
<fieldOfView_mm>
<x>600</x>
<y>328.153125</y>
<z>160</z>
</fieldOfView_mm>
</encodedSpace>
<reconSpace>
<matrixSize>
<x>128</x>
<y>116</y>
<z>64</z>
</matrixSize>
<fieldOfView_mm>
<x>300</x>
<y>271.875</y>
<z>128</z>
</fieldOfView_mm>
</reconSpace>
<encodingLimits>
<kspace_encoding_step_1>
<minimum>0</minimum>
<maximum>83</maximum>
<center>28</center>
</kspace_encoding_step_1>
<kspace_encoding_step_2>
<minimum>0</minimum>
<maximum>45</maximum>
<center>20</center>
</kspace_encoding_step_2>
<slice>
<minimum>0</minimum>
<maximum>0</maximum>
<center>0</center>
</slice>
<set>
<minimum>0</minimum>
<maximum>0</maximum>
<center>0</center>
</set>
</encodingLimits>
<parallelImaging>
<accelerationFactor>
<kspace_encoding_step_1>1</kspace_encoding_step_1>
<kspace_encoding_step_2>1</kspace_encoding_step_2>
</accelerationFactor>
<calibrationMode>other</calibrationMode>
</parallelImaging>
</encoding>

Should we fixed some fields in the params to help the subsampleIndices ?