Ultrasound Signal Processing for Quantitative Microvascular Imaging

点击翻译按钮获取中文摘要

This Discovery Grant will support our research developing improved methods to use ultrasound to image blood flow in networks of small vessels. This research is important because current medical imaging technologies are not sufficiently effective for measuring the subtle changes in blood flow that can be the earliest evidence that a patient is or is not being helped by treatments for conditions such as cancer, cardiovascular disease, and diabetes. Our research will improve three important ultrasound technologies used for vascular imaging, namely power Doppler, ultrafast imaging, and contrast-enhanced ultrasound. Power Doppler is an inexpensive, widely available technology for imaging vessel networks, but the images are strongly affected by settings chosen by the operator of the ultrasound system. Our innovation is to automate the selection of some of those settings, which will improve the accuracy of comparisons among images of different patients or among images of an individual patient that are acquired on different days. Ultrafast imaging is a state-of-the-art Doppler method that improves imaging of the smallest vessels by allowing more measurements to be used to produce each image. Our lab is developing a new ultrafast technique named "spread-spectrum Doppler" that will maximize the number of blood-flow measurements made without reducing image quality. We also plan to combine spread-spectrum Doppler with our automated method for choosing scanner settings, which should increase the effectiveness of both techniques. Contrast-enhanced ultrasound uses microscopic bubbles to increase the signal strength from small vessels above the levels that are possible with Doppler. Our lab is developing new methods to statistically analyze contrast-enhanced ultrasound images with the goal of increasing the technique's sensitivity to changes in how well organized or disorganized a vessel network is, which is valuable information to obtain about vessel networks in cancer tumours and other diseased tissues. Our work is also expected to yield new insights into the physical relationships between the structure of blood vessel networks and the information contained in ultrasound images of those vessels that will be valuable to other medical imaging researchers. This research will provide training opportunities for undergraduate and graduate students in biomedical engineering, electrical engineering, and medical physics that will help prepare those students for industry or academic careers designing and developing medical devices.