Emma Louise Hall1, Mary C. Stephenson1, Darren Price1, Peter G. Morris1
1Sir Peter Mansfield Magnetic Resonance Centre, University of Nottingham, Nottingam, Nottinghamshire, United Kingdom
In magnetic resonance spectroscopy (MRS), the acquisition of data from multi-element receive coils is becoming common. As the number of coil elements increases the need arises for robust reconstruction methods. Here we compare three methods based on signal or signal and noise weighting. We demonstrate this on for proton MRS data acquired using a 32 channel receive coil at 7 Tesla acquired in different brain regions. We find that the visual cortex benefitted the most, whereas medial prefrontal cortex was relatively unaffected.