Search

Segmentation, recognition and 3D reconstruction of objects have been cutting-edge research topics, which have many applications ranging from environmental and medical to geographical applications as well as intelligent transportation. In this dissertation, I focus on the study of segmentation, recognition and 3D reconstruction of objects using LiDAR data/MRI. Three main works are that (I). Feature extraction algorithm based on sparse LiDAR data. A novel method has been proposed for feature extraction from sparse LiDAR data. The algorithm and the related principles have been described. Also, I have tested and discussed the choices and roles of parameters. By using correlation of neighboring points directly, statistic distribution of normal vectors at each point has been effectively used to determine the category of the selected point. (II). Segmentation and 3D reconstruction of objects based on LiDAR/MRI. The proposed method includes that the 3D LiDAR data are layered, that different categories are segmented, and that 3D canopy surfaces of individual tree crowns and clusters of trees are reconstructed from LiDAR point data based on a region active contour model. The proposed method allows for delineations of 3D forest canopy naturally from the contours of raw LiDAR point clouds. The proposed model is suitable not only for a series of ideal cone shapes, but also for other kinds of 3D shapes as well as other kinds dataset such as MRI. (III). Novel algorithms for recognition of objects based on LiDAR/MRI. Aimed to the sparse LiDAR data, the feature extraction algorithm has been proposed and applied to classify the building and trees. More importantly, the novel algorithms based on level set methods have been provided and employed to recognize not only the buildings and trees, the different trees (e.g. Oak trees and Douglas firs), but also the subthalamus nuclei (STNs). By using the novel algorithms based on level set method, a 3D model of the subthalamus nuclei (STNs) in the brain has been successfully reconstructed based on the statistical data of previous investigations of an anatomy atlas as reference. The 3D rendering of the subthalamic nuclei and the skull directly from MR imaging is also utilized to determine the 3D coordinates of the STNs in the brain. In summary, the novel methods and algorithms of segmentation, recognition and 3D reconstruction of objects have been proposed. The related experiments have been done to test and confirm the validation of the proposed methods. The experimental results also demonstrate the accuracy, efficiency and effectiveness of the proposed methods. A framework for segmentation, recognition and 3D reconstruction of objects has been established, which has been applied to many research areas.

Parallel magnetic resonance imaging offers a framework for acceleration of conventional MRI encoding using an array of receiver coils with spatially-varying sensitivities. Novel encoding and reconstruction techniques for parallel MRI are investigated in this dissertation. The main goal is to improve the actual reconstruction methods and to develop new approaches for massively parallel MRI systems that take advantage of the higher information content provided by the large number of small receivers. A generalized forward model and inverse reconstruction with regularization for parallel MRI with arbitrary k-space sub-sampling is developed. Regularization methods using the singular value decomposition of the encoding matrix and pre-conditioning of the forward model are proposed to desensitize the solution from data noise and model errors. Variable density k-space sub-sampling is presented to improve the reconstruction with the common uniform sub-sampling. A novel method for massively parallel MRI systems named Superresolution Sensitivity Encoding (SURE-SENSE) is proposed where acceleration is performed by acquiring the low spatial resolution representation of the object being imaged and the stronger sensitivity variation from small receiver coils is used to perform intra-pixel reconstruction. SURE-SENSE compares favorably the performance of standard SENSE reconstruction for low spatial resolution imaging such as spectroscopic imaging. The methods developed in this dissertation are applied to Proton Echo Planar Spectroscopic Imaging (PEPSI) for metabolic imaging in human brain with high spatial and spectral resolution in clinically feasible acquisition times. The contributions presented in this dissertation are expected to provide methods that substantially enhance the utility of parallel MRI for clinical research and to offer a framework for fast MRSI of human brain with high spatial and spectral resolution.

A number of chronic disease and behavioural factors are recognised to increase the risk of cardiovascular and cerebrovascular diseases. These putative vascular risk factors have increasingly been recognised to increase the risk of cognitive impairment in the absence of clinically manifest ischemic events. Their relationship to structural brain changes has received limited attention. In this dissertation, I used high resolution magnetic resonance image (MRI) to examine two structural features of the brain, regional gray matter (GM) volumes and silent lacunar infarcts, and determined their association with vascular risk factors. I related these to cognitive function in both cross-sectional and longitudinal studies. The work was based on the data of three waves in two healthy cohorts drawn from the PATH Through Life Study, which is a population-based longitudinal study of ageing comprising 3 cohorts aged 20-24, 40-44, and 60-64 years, with about 2500 participants in each cohort. Random subsamples of Wave 1 of the cohort aged 60-64 years (N = 478) and Wave 2 of the 40+ cohort (aged 44-48 years) (N = 411) were examined cross-sectionally for the MRI sub-study. The MRI cohort aged 60-64 years was re-examined 4 years later in Wave 2. These studies showed that vascular risk factors are associated with lower regional GM volumes and this association varies at different ages. In adults aged 44-48 years, individual risk factors did not show a significant relationship with GM volumes, but the Framingham risk score was associated with less GM volumes in a number of brain regions, suggesting an additive effect of the risk factors. In the 60+ cohort, hypertension was independently associated with less regional GM volumes in bilateral medial frontal, right superior frontal, left superior temporal and precentral gyri. The same cohort, when examined in Wave 2, showed the negative association of hypertension with gray matter volumes to be more widespread. These associations were observed in men but not in women in either wave. Sex dimorphism was observed in the younger cohort as well, with greater GM volumes in temporal and occipital cortices, midbrain and cerebellum in men, while less GM volumes in cingulate and parietal cortices in comparison with women. Lacunar infarcts were present in 7.8 % of the 60+ cohort, and asymptomatic new lacunar lesions developed in 0.4 % per year in this group. The prevalence of lacunar infarcts was correlated with hypertension and a steeper decline in mental speed. These series of studies indicate the relationship of vascular risk factors with changes in brain structure and cognitive function in healthy middle-aged adults. It is suggested that modifying these vascular risk factors may protect the brain from silent lesions and cognitive impairment, and that intervention should begin early in life to have a major impact.

With the values of 6Li T1 measured and the literature values of J(13C-6Li) for these compounds, three new 13C NMR techniques are developed for the analysis of organolithium compounds. Modifications to the spectrometer are discussed, as well as calibrations of the 6Li decoupler channel needed to set up these new experiments. The theoretical development of each technique is presented, as well as data from their verification, using organolithium compounds of known structure. Once qualified, the new experimental techniques are used to analyze a series of alkyllithium / lithium alkoxide mixed aggregates in solution, where structures and values of J(13C-6Li) may not be known. The combination of Ti relaxation measurements and 13C{1H, 6Li} triple resonance techniques serves as a means of determining the structure of organolithium aggregates in solution.