Candidate: Yat Shun Yiu
Title: Ultrasound Imaging Innovations for Visualization and Quantification of Vascular Biomarkers
Date: August 7, 2019
Time: 10:00 AM
Place: EIT 3142
Supervisor(s): Yu, Alfred C.H.
Abstract:
More than 25% of the adult population has carotid atherosclerosis that could lead to major cerebrovascular diseases such as stroke and dementia, which are leading causes of death and serious long-term disability worldwide. Accordingly, the prevention and management of carotid atherosclerosis may play a significant role in reducing the morbidity and mortality of stroke and dementia. Current carotid atherosclerotic risk assessments rely on traditional cardiovascular risk factors that cannot accurately assess individual risk. It has been posited that this risk assessment performance may be improved by incorporating vascular biomarkers that have direct correlation to the initiation and progression of atherosclerosis. Nonetheless, it is not trivial to measure these vascular biomarkers routinely and consistently as current imaging techniques have limited spatial and temporal resolutions to derive these vascular biomarkers.
The goal of this thesis is to devise a new non-invasive imaging framework that can consistently quantify and visualize vascular biomarkers known to be directly related to atherosclerosis. Its overall hypothesis is that vascular biomarkers can be quantified and visualized using high-frame-rate ultrasound (HiFRUS) imaging that can generate images at more than 1000 frames per second. The corresponding sub-millisecond time resolution of HiFRUS can consistently track the spatiotemporal dynamics of blood flow in arteries, hence enabling the derivation of atherosclerotic vascular biomarkers. Based on such premise, this thesis addresses a series of technical hurdles in developing HiFRUS technology and presents new algorithms to quantify atherosclerotic vascular biomarkers. Five specific modules have been completed, and they collectively form the new imaging framework.
In the first module, a live imaging platform was devised to address the limitations of existing ultrasound scanners in executing HiFRUS imaging, thereby facilitating the development of new HiFRUS algorithms. This open platform allows the direct implementation of custom HiFRUS algorithms with real-time performance to deliver live imaging capacity. Such a feature is crucial when conducting in vivo experimentation as the live feedback provides navigation when assessing, for instance, complex flow dynamics in blood vessels. The devised hardware plays a critical role in enabling experimental implementation of new HiFRUS algorithms developed in this thesis.
The second module of this thesis reports the design protocol of a spiral flow phantom that serves as a performance evaluation tool for flow imaging innovations. This spiral phantom is specifically designed to generate omnidirectional flow, enabling a comprehensive performance assessment of novel flow estimators and visualization techniques. This phantom serves to calibrate the performance of a multi-angle vector flow estimator that is one of the core components of the fourth thesis module on motion-resistant imaging technique design.
The third thesis module presents a new HiFRUS imaging algorithm called Doppler ultrasound bandwidth imaging (DUBI) that focus on measuring Doppler bandwidth to quantify and visualize unstable arterial flow, which is a contributing factor to the initiation and progression of atherosclerotic lesions. Results demonstrated that DUBI can effectively detect and visualize the differences in Doppler bandwidth between stable flow and unstable flow. Receiver operating characteristic analysis also showed that DUBI has better sensitivity and specificity in identifying unstable flow than conventional method (0.72 & 0.83 vs 0.68 & 0.66). This algorithm allows the overall imaging framework to obtain new diagnostic information relevant to atherosclerosis.
In the fourth thesis module, a motion-resistant microvascular imaging (MRMVI) mode was devised to consistently map microvessels by suppressing the artifact caused by tissue motions such as arterial distension. Results showed that MRMVI can consistently map 50 µm diameter micro-flow channels in the presence of tissue motions, with the help of ultrasound contrast agent. The efficacy of MRMVI was further demonstrated by mapping an in vivo microvessel network with much finer detail than conventional power Doppler technique. This finding suggests that MRMVI can potentially be used to map intraplaque microvasculature to identify plaque neovascularization that is a prominent feature in advanced atherosclerotic plaque and reportedly is a predictor of future cerebrovascular events.
In the fifth thesis module, an instantaneous volumetric flow rate estimator (IVFRE) was developed to measure cerebral blood flow (CBF), since atherosclerosis is suggested to place extra burden onto the already declining CBF in older adults and may lead to mild cognitive impairment and, potentially, dementia. The IVFRE algorithm first made use of the change in correlation of flow speckle over time to derive the flow flux and then compute the final blood flow rate by integrating the flow flux across the cross section of the target vessel. Results from pilot study demonstrated that IVFRE can measure blood flow rate consistently under different in vivo scenarios. IVFRE addressed the need of a tool to routinely quantify CBF for atherosclerosis evaluation and it can potentially facilitate the identification of substantial CBF reduction in patients with carotid plaque before the development of cognitive impairment and dementia.
Overall, this thesis has reported a series of engineering innovations to achieve a live HiFRUS imaging framework that can assess new vascular biomarkers for risk stratification of carotid atherosclerosis and evaluation of the impact of atherosclerosis on CBF. It is anticipated that this framework can improve the atherosclerotic risk stratification process to more timely identify at-risk individuals. Furthermore, this framework may be used to derive new insights on the pathophysiology of carotid atherosclerosis that would aid future disease management and personalization of treatment strategies.