A Hybrid Approach to Motion Correction of Magnetic Resonance Imaging of Human Kidney - Self Navigation Techniques and Image Post Processing

Recent years have opened up the possibility to convert MRI data into reliable quantitative measures. Dynamic contrast enhanced MRI has proven to be a very promising tool for detailed spatial assessment of kidney function. In addition to explore optimal im age acquisition protocols, an important factor is how the acquired image data are analyzed. Recent advances in imaging technology allow fast, high-resolution recordings producing large collections of data in both space and time (3D, 3D+time). With these k ind of data, conventional visual inspection, manual measurements and delineations are very time consuming, or not feasible, and dependent on the observer. Processing of images from abdominal organs like the kidneys can be a very challenging task, since re cordings are hampered by organ motion, pulsations and respiration effects that might cause noisy data and artifacts. Therefore, methods for noise suppression, artifact detection and geometric correction can be important to apply, before segmentation, or v oxel-wise time course analysis is performed. If proper realignments and restorations can be done - or if noise or movement artifacts are negligible - image-derived parameters can be used for modeling renal function. In this project we aim at investigat ing approaches to motion suppression of DCE-MRI of kidney and to evaluate the outcome. Thereby, the overall goal is approached from two complementary directions (reflected also by the expertise of the bilateral partners): motion correction using image acq uisition techniques will be performed by the partner at the University of Heidelberg whereas motion correction based on computational post-processing will be investigated by the partner at the University of Bergen. To reach our goals we identified the fol lowing goals: -Self-Navigating Image Acquisition -Non-rigid Image Motion Correction -Comparison of the approaches -Combination of Image Acquisition based motion suppression and post processing approaches