A complex and challenging problem in flow study is to obtain quantitative flow information in opaque systems, for example, blood flow in biological systems and flow channels in chemical reactors. In this regard, MRI is superior to the conventional optical flow imaging or ultrasonic Doppler imaging. However, for high speed flows, complex flow behaviors and turbulences make it difficult to image and analyze the flows. In MR flow imaging, MR tagging technique has demonstrated its ability to simultaneously visualize motion in a sequence of images. Moreover, a quantification method, namely HARmonic Phase (HARP) analysis, can extract a dense velocity field from tagged MR image sequence with minimal manual intervention. In this work, we developed and validated two new MRI methods for quantification of very rapid flows. First, HARP was integrated with a fast MRI imaging method called SEA (Single Echo Acquisition) to image and analyze high velocity flows. Second, an improved HARP method was developed to deal with tag fading and data noise in the raw MRI data. Specifically, a regularization method that incorporates the law of flow dynamics in the HARP analysis was developed. Finally, the methods were validated using results from the computational fluid dynamics (CFD) and the conventional optimal flow imaging based on particle image velocimetry (PIV). The results demonstrated the improvement from the quantification using solely the conventional HARP method.

abstract: Magnetic resonance flow imaging techniques provide quantitative and qualitative information that can be attributed to flow related clinical pathologies. Clinical use of MR flow quantification requires fast acquisition and reconstruction schemes, and minimization of post processing errors. The purpose of this work is to provide improvements to the post processing of volumetric phase contrast MRI (PCMRI) data, identify a source of flow bias for cine PCMRI that has not been previously reported in the literature, and investigate a dynamic approach to image bulk cerebrospinal fluid (CSF) drainage in ventricular shunts. The proposed improvements are implemented as three research projects. In the first project, the improvements to post processing are made by proposing a new approach to estimating noise statistics for a single spiral acquisition, and using the estimated noise statistics to generate a mask distinguishing flow regions from background noise and static tissue in an image volume. The mask is applied towards reducing the computation time of phase unwrapping. The proposed noise estimation is shown to have comparable noise statistics as that of a vendor specific noise dynamic scan, with the added advantage of reduced scan time. The sparse flow region subset of the image volume is shown to speed up phase unwrapping for multidirectional velocity encoded 3D PCMRI scans. The second research project explores the extent of bias in cine PCMRI based flow estimates is investigated for CSF flow in the cerebral aqueduct. The dependance of the bias on spatial and temporal velocity gradient components is described. A critical velocity threshold is presented to prospectively determine the extent of bias as a function of scan acquisition parameters. Phase contrast MR imaging is not sensitive to measure bulk CSF drainage. A dynamic approach using a CSF label is investigated in the third project to detect bulk flow in a ventricular shunt. The proposed approach uses a preparatory pulse to label CSF signal and a variable delay between the preparatory pulse and data acquisition enables tracking of the CSF bulk flow. / Dissertation/Thesis / Doctoral Dissertation Biomedical Engineering 2017

No description available.

No description available.

Foundational approaches for realizing practical, non-intrusive measurements using filtered Rayleigh scattering (FRS) are presented and analyzed for the multi-property quantification of internal flow fields. Validation is challenging in applying computational fluid dynamics (CFD) solutions to real-world scenarios, necessitating benchmark measurements with well-defined uncertainties. The ideal instrument for achieving the required measurements should be non-intrusive and require no particulate or gas seeding. One approach that satisfies these requirements is filtered Rayleigh scattering. FRS is a laser-based optical diagnostic technique that allows for the simultaneous, non-intrusive measurement of three-component velocity, static temperature, and static density everywhere within a two-dimensional plane illuminated by laser light without using any form of flow seeding. The major disadvantage of FRS is that it is very susceptible to signal contamination from particles and surfaces illuminated by the probing laser source. The effects of these contamination sources of the FRS signal are quantified as a function of their intensity relative to the Rayleigh scattered light. As the most significant contributor to Rayleigh scattering contamination, methods for reducing geometric or background contributions were investigated. Structured illumination was applied in cross-correlation Doppler global velocimetry to reduce geometric scattering contributions in image acquisition, demonstrating the removal of background scattering biases in an FRS-similar technique. For multi-property measurements, it is shown that with only an order of magnitude estimate of Mie and geometric scattering, a range of wavenumbers termed the rejection region can be pre-defined such that molecular iodine absorbs the contamination. At the same time, Rayleigh scattered light can pass through. Mie and geometric scattering contributions are reduced to negligible levels within the rejection region, allowing for unbiased temperature and density measurement. Additionally, a method for determining only Doppler shift, desirable due to its increased processing speed and spatial resolution, was developed and shown to be robust to at least one order of magnitude greater Mie and geometric scattering than other methods. The biases associated with sampling a statistical average of the flow using time-averaged FRS were also investigated. The result is that measuring flow properties with the "constant in time" assumption is valid up to a turbulent intensity of 20%, resulting in biases in velocity and temperature greater than 10% of the measurement uncertainties predicted without these contributions. These advancements allow researchers to optimize measurement parameters and predict uncertainties before integrating them into a facility. These methods were implemented in a turbulent, highly distorted internal flow environment with Mie and background scattering present. Measurement uncertainties for vector velocity components, static temperature, and static density are predetermined using a 95% confidence interval on the Monte Carlo simulation results. Derived measurement uncertainties are calculated by propagating the results of the Monte-Carlo simulation. Measurements are compared to reference five-hole probe and particle image velocimetry measurements to assess the validity of the predicted uncertainty bounds. The results from this study show good agreement in the measurement of axial velocity and derived circumferential and radial flow angles when compared to reference measurements. These comparisons typically yield measurements that measure the same value as the five-hole probe data within the pre-defined uncertainty bounds of 9 m/s, 1.0°, and 3.8°, with significant deviations occurring at radii greater than 71% for tangential flow angle and radii greater than 55% for radial flow angle. Compared to facility average measurements, static density and static pressure data collected over the entire plane show RMSD values comparable to predicted measurement uncertainties of 0.043 kg/m^3 and 4.0 kPa, respectively. For the same comparison, temperature measurements show a greater RMSD than the predicted uncertainty of 8.4 K. While additional work remains to identify sources of bias error in some measurements, this work lays the foundation for FRS-based diagnostics to be used as a replacement or supplemental measurement technique in quantifying the state of fluid flow fields. / Doctor of Philosophy / Rayleigh scattering is a process that results from the interaction of light with microscopic particles that, whether we know it or not, we experience every day. When sunlight interacts with air molecules, the light scattered to our eyes is blue. The fact that the sky appears blue indicates a key property of Rayleigh scattering in that it is most efficient for the shortest wavelengths. What isn't apparent is that a whole host of other properties can be extracted from observed scattering by imaging it with a camera and a specialized filter when illuminated by a narrow wavelength laser. The problem is that a few dust particles, small enough to pass through a household air filter, can scatter more light than all the air molecules in a shot glass, with laser light scattering off large surfaces even more intense. The primary focus of this dissertation is to define Foundational approaches for realizing practical, non-intrusive filtered Rayleigh scattering techniques and methods necessary so that the light scattered from air molecules can be measured while avoiding the scattering from particles and surfaces. These approaches enable the measurement of the three-component velocity, temperature, and density of the gas being illuminated without the measurement affecting the flow itself. Because all these properties can be measured simultaneously, Rayleigh scattering provides one of the most comprehensive experimental measurement techniques available to researchers, making it highly desirable in quantifying gaseous flows and validating computational fluid dynamics calculations. Measurements collected with the techniques outlined in this work are validated experimentally using reference measurements in a large-scale internal flow facility, providing the groundwork for future applications of Rayleigh scattering-based diagnostics.

No description available.

In spectral domain optical coherence tomography (SD-OCT), any transverse motion component of a detected obliquely moving sample results in a nonlinear relationship between the Doppler phase shift and the axial sample velocity restricting phase-resolved Doppler OCT. To circumvent the limitation, we propose the lateral resonant Doppler flow quantification in spectral domain OCT, where the scanner movement velocity is matched to the transverse velocity component of the sample motion.