SEMANTIC CLASSIFICATION OF PICTURES & WORDS: FMRI OF REPETITION PRIMING: MEMORY

点击翻译按钮获取中文摘要

Introduction: This is an interdisciplinary bioengineering project to develop and evaluate dynamic magnetic resonance imaging (MRI) techniques for the clinical management and surgical planning of pediatric patients suffering from deteriorating renal function. The long term goal is to replace the existing renal tests, which involve significant costs and risks to patients, with a comprehensive, cost-effective, noninvasive examination. Methods: 1) A model of the kidney developed to predict the velocity of the contrast media in the late proximal tubule and thin descending loop of Henle. Assuming that the initial velocity of the filtrate is directly related to the pressure (both osmotic and hydrostatic) gradients between the glomerulus and the initial proximal convoluted tubule, velocities in different parts of the nephron can be extrapolated, in principle, if the relevant spatially dependent pressures are known. Using lumped parameter element design theory, the goal would be to predict these relevant pressure gradients given a priori knowledge of glomerulus and collecting system pressures, fluid viscosity, diffusion constants, and tubule geometry (area and length). 2) A model of the kidney may be developed so that absolute Gd-DTPA concentrations can be ascertained. This would give a direct measure of regional concentrating ability similar to the nuclear medicine studies that are currently used. Since, for a given sequence (e.g., GRASS), the signal intensity is a complicated function T1, T2, T2* and Gd-DTPA concentration, many factors must be considered. Conclusions: During the next year we will compare interleaved spiral, short-TR GRE, single-shot EPI pulse sequences in terms of spatial resolution, temporal resolution, and sensitivity to inhomogeneities and motion. The ultimate choice of dynamic MRI pulse sequence will also be constrained by the minimum temporal and spatial resolution requirements needed to visualize renal pathology. The data analysis and image processing algorithms will focus on techniques such as correlation imaging that emphasize the detection and visualization of regional renal defects rather than absolute quantitation of kidney function.