Variable bandwidth filtering for magnetic resonance imaging with pure phase encoding

Abstract

Magnetic resonance imaging with pure phase encoding (sometimes known as single point or constant time imaging) has many desirable advantages, but is usually time consuming in comparison to frequency encoding methods. In single point imaging the maximum signal bandwidth is proportional to both the phase-encoding gradient amplitude and the object size. It is usual practice to set the acquisition filter bandwidth to the maximum value expected during a measurement. Hence the filtering employed in this kind of measurement is not optimal for the low frequency k-space points. An optimal way to set the filter bandwidth is presented in this study. By reducing the filter bandwidth to match the point sampled in k-space, the inherent SNR is improved and this, in turn, may be used to reduce the number of signal averages required for acceptable SNR. The variable bandwidth filter offers a theoretical SNR increase of 41%. This paper shows the results of its application and comparison with fixed low-pass filtering. Practical measurements show a gain of 20% in SNR, which would translate into a 31% reduction in averaging time required for a single image without any detrimental effects on the image quality.