Correction of accidental patient motion for online Mr thermometry

Abstract

Magnetic Resonance (MR) temperature mapping can be used to monitor temperature changes in minimally invasive thermal therapies during the procedure. The proton resonance frequency (PRF) shift technique gives an estimate of the relative temperature variation, comparing contrast between dynamically acquired images and reference data sets. However, organ displacements due to physiological activity (heart and respiration) may induce important artifacts on computed temperature maps if no correction is applied. This paper summarizes existing classical methods and presents a new approach for correction of accidental motion in such MR images in order to increase robustness of temperature estimation using the PRF shift. The correction method described in this paper consists of using image registration techniques to estimate motion on anatomical images. © Springer-Verlag Berlin Heidelberg 2004.