Using M-Lines for Theoretically Exact Image Reconstruction from Cone-beam Data of an Ellipse-line-ellipse Trajectory

Abstract: CT-like 3D imaging with angiographic C-arm devices has been proved to be a valuable tool to assist angiographic therapeutic procedure for planning, guidance and assessment. The field-of-view (FOV) is determined by the given detector size. In some cases the FOV might not cover the region-of-interest requested by clinical applications such as interventions of spine, descending aorta, peripheral arteries, etc..

Various source trajectories for extended acial FOV could be implemented on a multii-axis angiographic C-arm system. The ellipse-line-ellipse (ELE) source trajcectory is complete, allowing exact and stable image reconstruction, and has the potential of achieving a large axial FOV per rotation.

This thesis focuses on implementing differentiated backprojection reconstruction method followed by inverse Hilbert transform (DBP-HT) using ELE trajectory for long-object cone-beam imaging. To be specified, we implement reconstruction method based on M-lines for cylindrical simulated phantom. Emphasis is placed on accuracy. Phantom is designed to satisfy long-object scanning requirement. Matlab is used to simulate the entire process from forward cone-beam projection acquisition to image reconstruction.