(A) study of fluctuating flow in NMR imaging:application to arteriography and diffusion imaging

Abstract

The pulsatile nature of arterial blood flow is exploited for NMR arteriography by using a spectral decomposition and the imaging artifacts arising from the fluctuation flow and motion are reduced in NMR diffusion imaging by using the LPR imaging instead of the Fourier imaging. In abdominal NMR imaging, the respiratory motion effects whcih also result in the image artifacts such as the ghosts and blurs are analyzed and their correction method is suggested. Fourier transformation of a series of projection images in the temporal derection along the cardiac cycle results in spectral images where the arteries are included in the images of harmonic components while the stationaly tissues and veins are collected into the image corresponding to the dc component. From the magnitude of the spectral images an arteriogram can be obtained by summation of the images corresponding to the harmonic components excluding the dc component. This principle is applied to Fourier imaging in a cine mode data acquisition as well as line scan imaging. Since this technique does not require the flow-encoding gradients. this is free from eddy current artifacts which have been one of the major obstacles to the projection angiography using the flowencoding gradient. Diffusion imaging is another imaging modality which is extremely sensitive to flow or motion of the object due to the strong diffusion gradient employed to obtain sufficient attenuation by the diffusion effect. In diffusion imaging techniques based on Fourier imaging, fluctruating flow and irregular motion of the object result in the severe flow artifacts smearing in the phase-encoding direction. These flow artifacts have been reduced by use of LPR imaging method instead of the conventional Fourier imaging. This improvement is much owed to the fact the flow-induced phase shift does not affect the image quality in the back-projection reconstruction since in this method only the magnitude of the projection data is used. Furthermore,...