parameters

[LHuysmans]

Hi, I'm trying to use the dixonfix on 6pt dixon images of size 384x384x340 with a spacing of 1.17x1.17x2. I do succeed to running it and get some output, but the result is not very good. So I was wondering how I should choose the different parameters in the config files? I struggle most with spacing (as it is limited by the hardware I use) and landa. If I use a spacing of 2 for training the regression trees I don't succeed in the end to run the dixonfix as it crashes on a memory error. What parameters would you advice for my data? Both during training of the regression trees and for running the dixonfix in the end?
Thank you in advance!

[bglocker]

HI. Apologies for the slow response. It's difficult to comment on the parameters as they may depend on specifics of the data, which I cannot judge. I wouldn't recommend using much lower resosultions than 2mm, so you may need to look for a machine with more RAM to make it work. Ideally, the regression forests should produce a visually reasonable prediction of the swap-free fat and water images. How do the predictions look for you? Can you share any screenshots?

[LHuysmans]

Hi, thank you for the response. The predictions look fine. I've solved what I wanted to do. Thank you for your help.