Magnetic Resonance Imaging is a challenging imaging technique. Nowadays, the goal is to reduce the acquisition scan-time for a given quality of reconstruction. The applications are for instance cardiac dynamic MRI, where a 2 or 3 D image of the heart is acquired trought time.
Sophisticated algorithms for reconstruction are beeing developped, that need simulation phantom for first validation.
The idea of this project would be now to design a software phantom for cardiac MRI. We could first focus on the 2D static case. We should build aphantom that is inspired from the NCAT phantom (http://www.bme.unc.edu/~wsegars/M5-3.pdf), such that we can analytically compute the MRI response.
Later, the idea is to generalize this technique to non-homogeneous coil sensitivities that may be involved in the MRI measurement process.
Further, the goal is to extend this software phantom to 3D and integrate a motion model for dynamic simulations.