Monte Carlo method for uncertainty propagation in magnetic resonance-based electric properties tomography

This paper investigates the uncertainty propagation in magnetic resonance-based electric properties tomography, a quantitative imaging technique that recovers the electric properties distribution inside a human body at the Larmor frequency. This is a needed step in order to make the quantitative results reliable for in vivo applications. To this aim, the contrast source inversion method is investigated as a promising technique and the uncertainty propagation through the corresponding model is studied by means of the Monte Carlo method. Thanks to the increased quality in the recovered electric properties, the results suggest that TEM coils for parallel transmission could be a preferable choice.