search

kschan0214 / sepia

Matlab GUI pipeline application for quantitative susceptibility mapping (QSM)
MIT License
46 stars 10 forks source link

spatial mismatch #82

Closed lconcha closed 5 months ago

lconcha commented 5 months ago

Hello.

I am trying to compute QSM on rodent data. Images were converted from paravision format using brkraw.

After completion of the pipeline, some derived maps appear to be spatially discordant to the original magnitude image. It may seem like this is due to mask erosion, but it truly seems to be misaligned. Please see a snapshot of the magnitude image and sepia outputs, and notice how the Chi map seems to be shifted with respect to the magnitude image. The last panel in the figure shows sepia configuration used.

Am I doing anything wrong with how the maps are calculated? Your help is much appreciated!

sepia_rat

lconcha commented 5 months ago

Good news. It turns out the phase images were misplaced with respect to the magnitude images before going into sepia. These maps were generated by the scanner (Bruker) using Paravision 7.0. It strikes me as very odd that these maps could be misplaced at this stage, but I repeated the process for generating them and they are now in the right place. In turn, outputs from sepia are looking nice! I will have to chase down the source of this problem, but it is definitely not due to sepia processing.

image

kschan0214 commented 5 months ago

Thanks for the update and good to hear the software is working fine!