Sparse-view C-arm Cone-beam CT for Guidance of Brain Aneurysm Coiling

点击翻译按钮获取中文摘要

DESCRIPTION (provided by applicant): In this project we propose to develop novel methods for improved image quality and dose reduction in c-arm cone-beam computed tomography (CBCT) for the application of guiding coiling for brain aneurysms. C-arm CBCT has become an indispensable technology in interventional neuroradiology, as it forms the basis for the volume-rendered images that are used to guide procedures such as endovascular coiling. Coiling is a surgical technique by which a microcatheter is guided from the femoral artery to the brain and used to insert small platinum coils into the target aneurysm which, ideally, promote clotting, seal off the aneurysm from blood flow, and prevent further stress and aneurysm rupture. CBCT is necessary to obtain the three dimensional information which: (1) informs placement of the microcatheter; (2) assesses the suitability of the aneurysm to coiling, which depends on the target aneurysm's dome-to-neck ratio, as well as its relationship to surrounding vasculature; and (3) verifies the placement and packing of the coils within the aneurysm and across the aneurysm neck. Therefore, the entire procedure depends crucially on high-quality CBCT images. However, a stringent and necessary emphasis on image quality in these procedures has led to a sacrifice in patient and physician safety in terms of radiation dose. We propose that by developing advanced optimization-based CBCT reconstruction algorithms, novel sparse-view scans can be used which deposit less radiation dose without a corresponding sacrifice of image quality. Further, we seek to obtain not only equivalent results to the current standard algorithms, but to achieve an objective improvement in image quality. Therefore, the specific aims of this project are: (1) to design novel data pre-processing methods appropriate to sparse-view data acquisition, (2) to develop iterative reconstruction methods for sparse-view c-arm CBCT data, and (3) to develop and apply objective assessment of image quality to evaluate sparse-view reconstruction. The first specific aim will address the fact that the c-arm systems used in interventional neuroradiology are susceptible to various miscalibrations, especially uncertainties in scanner geometry, such as gravitational sag, which vary from view to view. These geometric deviations must be accurately and precisely determined for optimal image quality in the reconstructed image, especially for non-conventional scans like sparse-view scans. The second specific aim will develop novel optimization-based image reconstruction algorithms for sparse-view CBCT angiography data. Lastly, the third specific aim will be to develop objective, task-specific figures of merit so that the image quality implications of the results in the first two ais can be assessed. We believe that the project is of high scientific and clinical significance in the sense that the novel formalism for image processing, reconstruction, and evaluation we propose can provide a framework for obtaining low-dose CT images in interventional neuroradiology, thereby sparing harmful excess radiation dose while ensuring a stringent level of image quality.