published_or_final_version / Medicine / Master / Doctor of Medicine

Homozygous α0-thalassemia is a serious autosomal recessive disorder with poor fetal outcome and severe maternal complications. Conventionally, prenatal diagnosis is performed by an invasive test. A non-invasive approach using serial ultrasonography can effectively reduce the need for invasive tests in unaffected pregnancies. For two-dimensional ultrasound prediction, a total of 777 at-risk fetuses were studied from 12 to 20 weeks between 1995 and 2006. At 12–15 weeks’ gestation, the highest sensitivity (98.3%) was achieved by the combination of fetal cardiothoracic ratio (CTR) and/or middle cerebral artery peak systolic velocity (MCA-PSV) at a false-positive rate of 15.8%. At 16–20 weeks’ gestation, the sensitivity of CTR was 100.0%, but the false-positive rate was 5.2%. In contrast, the false-positive rate of MCA-PSV alone was 1.4% and that of the combination of CTR and MCA-PSV was 0%, although their sensitivities were less than 65%. In a cross-sectional retrospective study of 546 samples at-risk and control (268 fetal and 278 neonatal cord blood), the degree of anemia was only mild in 27.5% of the affected fetuses (see chapter 3 for definition of mild anemia). Because MCA-PSV is not very predictive of mild anemia, this may be one of the reasons why MCA-PSV is not very sensitive in predicting an affected pregnancy. A total of 832 at-risk pregnancies were studied using same noninvasive approach at Maternal and Neonatal Hospital of Guangzhou (MNH) and Tsan Yuk Hospital (TYH). The overall sensitivity and specificity of the noninvasive approach was 100% and 95.6% respectively. At MNH, the need for an invasive test was reduced by 78.6%, and all the affected pregnancies were diagnosed before 24 weeks’ gestation. After adequate training and monitoring the quality of the subsequent ultrasound examinations, the results achieved at MNH were comparable to TYH, with at-risk pregnancies including the affected ones being seen at a more advanced gestation at MNH. In a retrospective review of 361 women at risk of carrying an affected fetus, 311 (86.2%) opted for the non-invasive approach using CTR and/or placenta. The cost saving of this non-invasive approach was relatively small (HK$ 2,651) in comparison to the cost of the whole prenatal screening program. On the other hand, the non-invasive approach was more expensive than the direct invasive approach for low MCV couples, as well as couples discordant for α-thalassemia and β-thalassemia. ages. These results support the adoption of non-invasive approach in which routine invasive test or karyotyping is no longer performed. A total of 106 at-risk pregnancies and normal controls were prospectively studied using three-dimensional ultrasonography. Placental volume (PV) at 11-14 weeks, and PV/CRL quotient at 9-14 weeks’ gestation of affected pregnancies were significantly greater than unaffected pregnancies (P<0.05). Using a cut-off point of 1.2ml/mm for PV/CRL quotient to predict an affected pregnancy, the sensitivity, and specificity was 96.2%, and 100.0% respectively. / published_or_final_version / Obstetrics and Gynaecology / Master / Doctor of Medicine

Traditional open interventions have been progressively replaced with minimally invasive techniques. Most notably, direct visual feedback is transitioned into indirect, image-based feedback, leading to the wide use of image-guided interventions (IGIs). One essential process of all IGIs is to align some 3D data with 2D images of patient through a procedure called 3D-2D registration during interventions to provide better guidance and richer information. When the 3D data is unavailable, a realistic 3D patient-speci_c model needs to be constructed from a few 2D images. The dominating methods that use only image intensity have narrow convergence range and are not robust to foreign objects presented in 2D images but not existed in 3D data. Feature-based methods partly addressed these problems, but most of them heavily rely on a set of \best" paired correspondences and requires clean image features. Moreover, the optimization procedures used in both kinds of methods are not e_cient. In this dissertation, two topics have been studied and novel algorithms proposed, namely, contour extraction from X-ray images and feature-based rigid/deformable 3D-2D registration. Inspired by biological and neuropsychological characteristics of primary visual cortex (V1), a contour detector is proposed for simultaneously extracting edges and lines in images. The synergy of V1 neurons is mimicked using phase congruency and tensor voting. Evaluations and comparisons showed that the proposed method outperformed several commonly used methods and the results are consistent with human perception. Moreover, the cumbersome \_ne-tuning" of parameter values is not always necessary in the proposed method. An extensible feature-based 3D-2D registration framework is proposed by rigorously formulating the registration as a probability density estimation problem and solving it via a generalized expectation maximization algorithm. It optimizes the transformation directly and treats correspondences as nuisance parameters. This is signi_cantly di_erent from almost all feature-based method in the literature that _rst single out a set of \best" correspondences and then estimate a transformation associated with it. This property makes the proposed algorithm not rely on paired correspondences and thus inherently robust to outliers. The framework can be adapted as a point-based method with the major advantages of 1) independency on paired correspondences, 2) accurate registration using a single image, and 3) robustness to the initialization and a large amount of outliers. Extended to a contour-based method, it di_ers from other contour-based methods mainly in that 1) it does not rely on correspondences and 2) it incorporates gradient information via a statistical model instead of a weighting function. Tuning into model-based deformable registration and surface reconstruction, our method solves the problem using the maximum penalized likelihood estimation. Unlike almost all other methods that handle the registration and deformation separately and optimized them sequentially, our method optimizes them simultaneously. The framework was evaluated in two example clinical applications and a simulation study for point-based, contour-based and surface reconstruction, respectively. Experiments showed its sub-degree and sub-millimeter registration accuracy and superiority to the state-of-the-art methods. It is expected that our algorithms, when thoroughly validated, can be used as valuable tools for image-guided interventions. / published_or_final_version / Orthopaedics and Traumatology / Doctoral / Doctor of Philosophy

abstract / Electrical and Electronic Engineering / Doctoral / Doctor of Philosophy

As compared with other imaging techniques such as computed tomography (CT), magnetic resonance imaging (MRI) provides distinctive advantages with better contrast and resolution in imaging brain anatomy and function in vivo. As compared with electrophysiological and histological tracing techniques, MRI enables longitudinal investigation with higher efficiency, lower labor cost and less possibility of sampling error. The major objective of this doctoral work is to utilize cellular and molecular MRI to investigate normal brain functions and injuries in vivo. The results successfully demonstrated MRI as an efficient and sensitive tool for providing comprehensive assessment of brain injuries for promoting accurate prognosis and timely intervention, and for studying fundamental questions with regard to cortical adaptations to challenges in the young adulthood. Firstly, diffusion tensor imaging (DTI) and T2-weighted imaging were employed to characterize longitudinal neuronal and axonal changes of pyramidal tract (PY), a critical part of corticospinal tract, following experimental intracerebral hemorrhage (ICH). Combining DTI with T2-weighted imaging results, ipsilateral PY injuries following ICH were diagnosed as four stages. Quantitative analysis revealed transient diffusivity decreases in PY both contralateral and ipsilateral to the primary hemorrhagic site. Evolution of the ipsilateral DTI parameters correlated with histological findings and indicated evolving and complex pathological processes underlying monotonic FA decrease. These results demonstrated multi-parametric DTI as a valuable imaging tool for non-invasive and longitudinal monitoring of secondary PY injuries. Secondly, DTI and manganese-enhanced MRI (MEMRI) were utilized to detect neuronal changes of substantia nigra (SN) following experimental ICH in rodents. DTI revealed early changes in SN both contralateral and ipsilateral to the primary hemorrhagic site. Evolution of the ipsilateral parameters correlated with the histological results. MEMRI provided insights into the cellular phenotype changes at the late stage. DTI can serve as a valuable imaging tool for non-invasive early detection and longitudinal monitoring of secondary SN injuries, while MEMRI could complementally provide information regarding the late stage inflammation process. Multi-parametric MRI could facilitate clinical and preclinical investigations of SN injuries for exploring disease mechanisms and developing new therapeutic strategies. Thirdly, MEMRI was performed to characterize the interhemispheric interactions in normal and monocularly deprived rodent visual brain. Characteristic transcallosal manganese labeling was observed in the normal group in a manner consistent with previous histological findings. Significant decrease of such labeling was observed in rats with left or right eyelid suturing, or with left eye enucleation, but not in rats with right eye enucleation. These results demonstrated MEMRI as an efficient tool for investigating interhemispheric interactions both anatomically and functionally. These results also indicated that the adult brain recruits different mechanisms for its adaptations to eyelid suturing and enucleation, thus shedding light on our understanding of the transcallosal interhemispheric excitation and inhibition. Lastly, new paradigms other than pressure injection for intracortical manganese administration in MEMRI were introduced to minimize the neuro-toxicity of manganese and maximize the sensitivity of MEMRI for studying cortical functional changes. Transmeningeal diffusion, osmotic pump-based infusion, and intranasal instillation were demonstrated to be successful in tracing interhemispheric connections and detecting stress-related cortical and subcortical changes. / published_or_final_version / Electrical and Electronic Engineering / Doctoral / Doctor of Philosophy

Functional magnetic resonance imaging (fMRI) is a noninvasive technique that can measure blood oxygenation level dependent (BOLD) signal changes in a large field of view with high spatial resolution. The objective of this dissertation is to explore and integrate novel and noninvasive fMRI methods at 7 Tesla to investigate the auditory and visual functions. First, different fMRI methods and stimulation paradigms were employed to explore some basic auditory functions such as sound pressure level (SPL) dependence in different brain structures, and periodotopy and tonotopy in the inferior colliculus (IC). BOLD signal changes increased significantly with SPL and the dependence was monotonic in the IC and lateral lemniscus (LL). The external cortex of IC (ECIC) had higher BOLD signal change than the central nucleus of IC and LL at high SPLs. This study indicates that sparse temporal sampling that is used to reduce the adverse effects of scanner noise may not be a prerequisite in auditory fMRI studies of the IC. Periodotopy and tonotopy in the IC was investigated using continuous imaging with passband balanced steady state free precession (bSSFP) sequence instead of sparse temporal sampling and echo planner imaging (EPI). The spatial gradients of best amplitude modulation frequency (referred to as periodotopy) and characteristic frequency (referred to as tonotopy) varied across the IC, but were approximately perpendicular at different locations. These findings enhance our understanding of how auditory information is preserved in the midbrain. Second, higher order function of behaviorally relevant sounds response selectivity in subcortical structures was investigated. The IC was found to exhibit a stronger response to forward vocalization than to the temporally inverted one. Moreover, blocking cholinergic projections to the IC by atropine injection was observed to significantly reduce the IC response selectivity to the 22 kHz vocalizations. These findings demonstrate the IC response selectivity to vocalizations and suggest that the cholinergic projection contributes to IC responses selectivity to the 22 kHz vocalization. This study provides further understanding about the higher order auditory processing and may have implications for the neural mechanisms underlying human speech perception Third, BOLD fMRI was applied to measure the brain response to stationary and apparent motion visual stimulation. The response of superior colliculus (SC) was weaker under dim light and saturates at higher intensities. Further, the BOLD signal changes and number of activated voxels were both significantly lower during 164 ˚/s apparent motion stimulation compared to stimuli at slower speeds. The results suggest that the SC was more sensitive to slow moving visual stimuli. This is the first fMRI study to investigate motion responsiveness and stimulus speed dependence in the rat SC. Results from these studies complement current knowledge and demonstrate the sophisticated role of subcortical structures such as IC and SC, which may have strong clinical significance to the field of auditory and visual research. Findings from the animal studies should open up new avenues of research and lay the ground work for future human studies. / published_or_final_version / Electrical and Electronic Engineering / Doctoral / Doctor of Philosophy

Regenerative medicine is an interdisciplinary field which has advanced with the use of biotechnologies related to biomaterials, growth factors, and stem cells to replace or restore damaged cells, tissues, and organs. Among various therapeutic approaches, cell-based therapy is most challenging and exciting for both scientists and clinicians pursuing regenerative medicine. Specifically, stem cells, including mesenchymal stem cells and adipose-derived stem cells, are promising candidate cell types for cell-based therapy because they can differentiate into multiple cell types for tissue regeneration and stimulate other cells through neovascularization or paracrine signaling. Also, for effective treatment using stem cells, the tissue engineered constructs, such as bioactive degradable scaffolds, that provide the physical and chemical cues to guide their differentiation are incorporated with stem cells before implantation. Also, it was previously demonstrated that tissue-engineered matrices can promote tubulogenesis and differentiation of stem cells to vascular cell phenotypes. Hence, during tissue regeneration after stem cell therapy, there are numerous factors that need to be monitored. As a result, imaging-based stem cell tracking is essential to evaluate the distribution of stem cells as well as to monitor proliferation, differentiation, and interaction with the microenvironment. Therefore, there is a need for a stem cell imaging technique that is not only noninvasive, sensitive, and easy to operate, but also capable of quantitatively assessing stem cell behaviors in the long term with high spatial resolution. Therefore, the overall goal of this research is to demonstrate a novel imaging method capable of continuous in vitro assessment of stem cells as prepared with tissue engineered constructs and noninvasive longitudinal in vivo monitoring of stem cell behaviors and tissue regeneration after stem cell implantation. In order to accomplish this, gold nanoparticles are demonstrated as photoacoustic imaging contrasts to label stem cells. In addition, ultrasound and photoacoustic imaging was utilized to monitor stem cells and neovascularization in the injured rat tissue. Therefore, using these methods, tissue regeneration can be promoted and noninvasively monitored, resulting in a better understanding of the tissue repair mechanisms following tissue injury. / text

Imaging systems often make use of macroscopic lenses to manipulate light. Modern microfabrication techniques, however, have opened up a pathway to the development of novel arrayed imaging systems. In such systems, centimeter-scale areas can contain thousands to millions of micro-scale optical elements, presenting exciting opportunities for new imaging applications. We show two such applications in this thesis: pressure sensing in microfluidics and high throughput fluorescence microscopy for high content screening. Conversely, we show that arrayed elements are not always needed for three dimensional light field imaging. / Engineering and Applied Sciences

published_or_final_version / abstract / toc / Medicine / Doctoral / Doctor of Philosophy

No description available.