Evaluation of image quality in magnetic resonance imaging with partially parallel imaging methods

Magnetic resonance imaging using partial parallel imaging acquisition and reconstruction methods offers substantial advantages over current imaging MRI methods, due to its ability to decrease scan time or increase image resolution. It is desirable that the partially parallel imaging magnetic resonance imaging data be expressed quantitatively, in a standardized manner that will be consistent from clinic to clinic. Obstacles to this goal include the fact that the image quality parameters of different partially parallel imaging magnetic resonance imaging methods vary with the specifics of the image encoding schemes implemented on different makes and models of magnetic resonance imaging scanners and radio frequency coils. Additionally metrics used to obtain image quality for standard Fourier Transform MRI methods have not been validated as accurate predictors of image quality when partially parallel imaging magnetic resonance imaging methods are involved. Thus a significant amount of data must be obtained and evaluated in order to provide the basis for developing standards for magnetic resonance imaging protocols which use partially parallel imaging. In the 1st Aim of this project we developed a standard, generic testing methodology, consisting of a phantom and image analysis method, to investigate the accuracy of various signal-to-noise ratio measurements when partially parallel imaging methods are involved. In the 2nd Aim we investigated image quality with regards to partially parallel imaging protocols and determined the most functional method of uniformity quantification from those standard methods available and determined the effect of partially parallel imaging magnetic resonance imaging processes on image uniformity. In the 3rd Aim we determined how different implementations of partially parallel imaging magnetic resonance imaging across manufacturer's platforms changes the relative decreases in signal-to-noise ratio for an assortment of common pulse sequence types. This aim also produced a range of expected signal-to-noise ratio decreases with the implementation of partially parallel imaging in magnetic resonance imaging protocols. Data obtained in the course of these investigations can be used to develop standards for the evaluation of image quality in partially parallel imaging magnetic resonance imaging.