Efficient Interpretation of 3D Vascular Image Data

点击翻译按钮获取中文摘要

DESCRIPTION (Provided by Applicant): Recent developments in Computed Tomography have resulted in the capability to image blood vessels in three dimensions and less invasively than conventional angiography. However, these developments have also resulted in the potential to generate thousands of images per patient study for radiological interpretation. As a result, it has become increasingly difficult to read each image separately and to reach an accurate diagnosis in a reasonable amount of time. Therefore, we propose to develop and validate technology that changes radiological interpretation of volumetric CT vascular image data from visual inspection of individual cross-sections to a paradigm that combines highly efficient, ergonomic, and interactive volumetric visualization and quantitative analysis. To this end, over the first three years of this project, we will develop a system of hardware and software specifically designed for this task. It will consist of a large-area high-resolution display, a set of human-computer interfaces specifically designed for facilitating interaction with large volumetric vascular data sets, and intelligent software capable of guiding the required interactions and generating blood-vessel-specific visualizations and quantitative results with minimal effort. We will conduct a clinical pilot study in the fourth year of the proposed work, during which radiologists will compare the use of a prototype system with conventional image-by-image reading for accuracy and efficiency. We will focus our analyses on (1) aortoiliac aneurysm, and (2) lower extremity occlusive disease, as test cases for the new technology. Upon completion of these studies, we expect that our developments will be easily adaptable to other applications both within and outside of vascular imaging, as well as to other imaging modalities such as MRI. Our overall goal is to change and validate the way crosssectional images are interpreted in general, thus resulting in improved accuracy and efficiency in the assessment of increasingly large volumes of medical image data.