search

SyneRBI / SIRF-Exercises

SIRF Training and demonstration material
http://www.ccpsynerbi.ac.uk
Apache License 2.0
18 stars 21 forks source link

Joint tv #103

Closed ckolbPTB closed 3 years ago

ckolbPTB commented 3 years ago

the only problem left are the coil maps. I use the simulated Shepp-Logan phantom to calculate the coil maps but they don't really fit to the brain data. Because there is no noise and the acquisition model is perfect it does not show up if the algorithm converges enough but it needs to be fixed at some point... jtv_notebook_final

ckolbPTB commented 3 years ago

I added one more notebook to describe how to create a different undersampling pattern which can then be used for the joint-tv notebook. I fixed the coil map problem, added comments at the end for further tasks and Vaggelis also added some more math.

Everything seems to be working, I just find it difficult to show, that the joint reconstruction is better than doing each problem separately, but maybe this is not necessarily the point...

KrisThielemans commented 3 years ago

Thanks a lot. Can we just delete the existing notebook? Please do so.