Advanced Strategies for Image Quality Improvement and Dose Reduction in CT

点击翻译按钮获取中文摘要

X-ray computed tomography (CT) has proven to be a critical and highly valued tool in the advancement of human health. These systems are employed in every facet of medicine from screening for cancer, to assessing injuries, and most recently guiding treatments such as radiation therapy and surgery for cancer. Existing CT technology is not fully utilizing the capabilities of the underlying physics to maximize image quality (e.g. contrast-to-noise) and minimize the radiation dose to the patient. The proposed program of research builds upon 20+ years of experience in X-ray imaging methods and system development to develop next-generation CT methods that leverage advances in computational horsepower, mechatronics, and nanotechnology to bring a safer, higher performing level of CT imaging to medicine. The program of research focuses on two major topics (i) correcting unwanted X-ray scatter signal in the detector, and (ii) the development of a novel fluence-field modulated CT (FFMCT) method. Scattered X-rays reaching the detector have a significant negative impact on image quality and this is exacerbated in cone beam CT (CBCT) due to the volume of tissue irradiated and lack of a rejection method. Accurate modeling of the physics underlying scatter with Monte Carlo (MC) simulations allows us to better understand and correct for scatter, leading to improved images and patient dose management. At present, MC simulations are computationally intensive and time-consuming. This project will develop more efficient computational methods for MC-based scatter correction, leading to correction methods that can be applied in real-time in real-world clinical settings to compensate for spurious signal and improve image quality. FFMCT was developed in my laboratory and has the potential to transform CT technology by changing the operation of the CT scanner depending on the details of the imaging task. It involves dynamically adjusting the X-ray fluence pattern during image acquisition so that the image quality and radiation dose are adjusted for specific regions-of-interest. FFMCT is able to account for known X-ray densities in the object, such as the high density of bone, and reduce dose to sensitive tissues not of interest (e.g. breast tissue in lung screening)—and as a result should improve existing CT image quality. The hybrid approach of using FFMCT and high-speed MC methods will allow alternative CT geometries to be explored, including the use of programmable multisource X-ray emitting arrays for imaging.