No abstract

Diffusion Tensor Imaging (DTI) of the human brain, coupled with tractography techniques, enable the extraction of large-collections of three-dimensional tract pathways per subject. These pathways and pathway bundles represent the connectivity between different brain regions and are critical for the understanding of brain related diseases. A flexible and efficient GPU-based rendering technique for DTI tractography data is presented that addresses common performance bottlenecks and image-quality issues, allowing interactive render rates to be achieved on commodity hardware. An occlusion query-based pathway LoD management system for streamlines/streamtubes/tuboids is introduced that optimizes input geometry, vertex processing, and fragment processing loads, and helps reduce overdraw. The tuboid, a fully-shaded streamtube impostor constructed entirely on the GPU from streamline vertices, is also introduced. Unlike full streamtubes and other impostor constructs, tuboids require little to no preprocessing or extra space over the original streamline data. The supported fragment processing levels of detail range from texture-based draft shading to full raycast normal computation, Phong shading, environment mapping, and curvature-correct text labeling. The presented text labeling technique for tuboids provides adaptive, aesthetically pleasing labels that appear attached to the surface of the tubes. Furthermore, an occlusion query aggregating and scheduling scheme for tuboids is described that reduces the query overhead. Results for a tractography dataset are presented, and demonstrate that LoD-managed tuboids offer benefits over traditional streamtubes both in performance and appearance.

Abstract. We describe an efficient and accurate method for segmenting sets of subcortical structures in 3D MR images of the brain. We first find the approximate position of all the structures using a global Active Appearance Model (AAM). We then refine the shape and position of each structure using a set of individual AAMs trained for each. Finally we produce a detailed segmentation by computing the probability that each voxel belongs to the structure, using regression functions trained for each individual voxel. The models are trained using a large set of labelled images, using a novel variant of 'groupwise' registration to obtain the necessary image correspondences. We evaluate the method on a large dataset, and demonstrate that it achieves results comparable with some of the best published.

No abstract

We address the problem of extracting information from groups of medical images of the same anatomy. We describe an algorithm which simultaneously segments and registers a set of such images, incrementally constructing a model of their structure and the correspondences across the set. The framework explicitly models the fraction of each tissue type, rather than the expected intensity in each voxel, to decouple the model from details of the imaging sequence and modality. When estimating the optimal deformation field, the current image is compared to a reconstructed image, generated from the model tissue fractions and the current estimate of intensity distributions for each tissue type in the current image (i.e. an estimate of how the model would appear given the imaging conditions for that image). We describe the algorithm in detail and present results of applying it to a set of MR images of the brain.