Archived Weekly Agendas for Spring 2011
From Utah Center for Neuroimage Analysis
Contents |
Dates
January 17
Martin Luther King Jr. Day
January 24
Speaker: Gopal Veni, Bo Wang
Presentation Title: Review of 2011 NA-MIC Project Week and all hands meeting (AHM)
Description/Abstract:
Gopal Veni and Bo Wang went to 2011 NA-MIC Project Week and all hands meeting (January 10-14, 2011, in Salt Lake City), they will introduce their NA-MIC projects.
Supplemental material:
2011 Winter NA-MIC Project Week Project list
2011 Winter NA-MIC all hands meeting (AHM) AHM Agenda
January 31
Speaker: Fangxiang Jiao
Presentation Title: The review of uncertainty analysis and visualization of Diffusion Tensor Images
Description/Abstract:
Despite the promising results presented by DT-MRI technique, it suffers from a number of sources of uncertainties. The choice of imaging parameters ( such as b value, the number of diffusion weighted directions and the strength of B0 magnetic field ), the registration of diffusion-weighted MR images, the tensor fitting process, the fiber tracking process all bring in some level of uncertainties. In addition, the background noise, imaging artifacts and the motion effects can make the situation even worse. Therefore uncertainty analysis and visualization remains challenging. In this talk, I will go by the DTI processing pipeline and review the researches on uncertainty analysis and visualization.
You can find Fangxiang's slides here: /home/sci/fjiao/Presentations
February 7
Speaker: John Lewis
Presentation Title: Size (always) matters: Fiber length and connectivity in development.
Description/Abstract:
Across species, increases in white matter volume outpace increases in gray-matter volume, but increases in gray-matter volume outpace increases in the size of the corpus callosum. I have hypothesized that this hyposcaling of the corpus callosum stems from the interaction of the conduction delays and cellular costs of the long-distance fibers with normal developmental mechanisms. According to this hypothesis, brains which are larger during development should retain fewer long-distance fibers. Autism provides for a strong test of the hypothesized impact of fiber length during development. Children with autism are known to have enlarged brains during the first years of life. This is predicted to lead to decreased long-distance connectivity. I will present results from neural networks that modeled the typical and the autistic growth patterns, and then describe methods to assess this relation using tractography.
February 14
Speaker: Yanfei Mao
Presentation Title: Feasibility Study of Segmented-Parallel-Hole Collimator for Stationary Cardiac SPECT
Description/Abstract:
A stationary cardiac SPECT system has a benefit of acquiring temporally consistent projections. The most challenging issue in building a stationary system is to provide sufficient projection view-angles. A 2-detector, multi-segment collimator system with 14 view-angles over 180˚ in the transaxial direction and 3 view-angles in the axial directions was designed, where the two detectors are configured 90˚ apart in an L-shape. We applied the parallel-beam imaging geometry and used segmented parallel-hole collimator to acquire SPECT data. To improve the system condition due to data truncation, we measured more rays within the field-of-view (FOV) of the detector by using a relatively small detector bin-size. In image reconstruction, we used the maximum-likelihood expectation-maximization (ML-EM) algorithm. The criterion for evaluating the system is the summed pixel-to-pixel distance that measures the discrepancy between the 3D gold-standard image and the reconstructed 3D region of interest (ROI) with truncated data. Effects of limited number of view-angles, data truncation, varying body habitus, attenuation, and noise were considered in the system design. As a result, our segmented-parallel-beam stationary cardiac SPECT system is able to acquire sufficient data for cardiac imaging and has a high sensitivity gain.
February 21
Presidents’ Day holiday
February 28
Speaker: Clifton Brooks
Presentation Title: Robo-Renoir: Toward Software which Paints Impressionist Paintings.
Description/Abstract:
Painterly Rendering, a sub-field of Nonphotorealistic Rendering, refers to systems which apply stroke based rendering methods to generate synthetic paintings from reference images. This talk articulates various interpretations of the design problem and summarizes a history of methods employed by past systems. Although primarily thought of as an exercise in image synthesis, painterly rendering relies on image analysis techniques from general image processing. This discussion emphasizes the image analysis phase partly because it confronts most of the ambiguity in painterly rendering, but also because the techniques relate to image processing in general.
March 7
Speaker: Torben Paetz
Presentation Title: Segmentation of Stochastic Images using Stochastic Extensions of the Ambrosio-Tortorelli and the Random Walker Model
Description/Abstract:
We discuss methods based on stochastic PDEs (SPDEs) for the segmentation of images with uncertain gray values resulting from measurement errors and noise. Our approach yields a reliable precision estimate for the segmentation result, and it allows us to quantify the robustness of edges in noisy images and under gray value uncertainty. Our ansatz space for such images identifies gray values with random variables. For their discretization we utilize the generalized polynomial chaos expansion and the generalized spectral decomposition (GSD) method.
This approach is used to derive the stochastic generalization of the Ambrosio-Tortorelli approximation of the Mumford-Shah functional along with the corresponding extension of the Gamma-convergence proof. Furthermore, we present extensions of edge linking methods and show, how it is possible to combine adaptive grids for the spatial dimensions with the problem dependent stochastic subspaces from the GSD method. Moreover, we present the extension of the random walker segmentation for our stochastic images, which is based on an identification of the graph weights with random variables, leading to a stochastic Dirichlet problem. We demonstrate the performance of the methods on artificial data, a data set from a digital camera as well as real medical ultrasound data. A comparison of the intrusive GSD discretization with a stochastic collocation and a Monte Carlo sampling is shown.
March 14
Speaker: Luke Hogrebe
Presentation Title: Trace Driven Registration of Neuron Confocal Microscopy Image Stacks
Description/Abstract:
Research in generating digital 3-D neural circuit reconstructions from histological data has predominantly focused on single sections. Ultimately, however, neurobiologists desire to study the long range connectivity of the brain, which requires tracking neurons across many serial sections. Registration of these sections is complicated by tissue deformation and loss introduced during the cutting and mounting processes. This work investigates a method for registering axonal sections using centerline traces to provide the locations of axons at section boundaries and the angles at which the axons approach the boundaries. This information is used to determine correspondences between two sections followed by stack transformations.
March 21
Spring Break
March 28
Speaker: Matt Wachowiak
Dr. Matt Wachowiak is a USTAR professor at Brain Institute and an associate professor of physiology at the University of Utah. Here is an introduction of his research (see this).
April 4
Speaker: Naveen Nagarajan
Naveen Nagarajan is a postdoctoral fellow in Department of Human Genetics at the University of Utah. He will give a short introduction about the current research projects in his lab.
Speaker: Fangxiang Jiao
Presentation Title: Uncertainty Analysis and Visualization in High Angular Resolution Diffusion Imaging
Description/Abstract:
In this paper, we propose a new and accurate technique for uncertainty analysis and uncertainty visualization based on orientation distribution function (ODF) glyphs, associated with high angular resolution diffusion imaging (HARDI).Our visualization applies volume rendering techniques to an ensemble of 3D ODF glyphs, designated here diffusion shapes to capture their variability due to underlying uncertainty. This rendering elucidates the complex heteroscedastic structural variation in these shapes. Furthermore, we quantify the extent of this variation by measuring the fraction of the volume of these shapes, which is consistent across all noise levels, the certain volume ratio.
Our uncertainty analysis and visualization framework is then applied to synthetic data as well as to HARDI human-brain data to study the impact of various image acquisition parameters and background noise levels on the diffusion shapes.
April 11
Speaker: Neda Sadeghi
Presentation Title: Statistical Growth Modeling of Longitudinal DT-MRI for Regional Characterization of Early Brain Development
Description/Abstract:
To gain a better understanding of neurodevelopment, clinical imaging increasingly makes use of longitudinal image studies. In this talk I will present a framework to model growth trajectories and to determine significant regional differences in growth pattern characteristics. We use non-linear mixed effect modeling where temporal change is modeled by the Gompertz function. The Gompertz function uses intuitive parameters related to delay, rate of change, and expected asymptotic value; all descriptive measures which can answer clinical questions related to growth. Our proposed framework combines nonlinear modeling of individual trajectories, population analysis, and testing for regional differences. We apply this framework to the study of early brain maturation in white matter regions as measured with diffusion tensor imaging (DTI). Regional differences between anatomical regions of interest that are known to mature differently are analyzed and quantified. I show statistical tests results for radial diffusivity (RD) measurements as these are known to be sensitive to the degree of myelination and axonal structuring.
Speaker: Xiang Hao
Presentation Title: Adaptive Riemannian Metrics for Improved Geodesic Tracking of White Matter
Description/Abstract:
We present a new geodesic approach for studying white matter connectivity from diffusion tensor imaging (DTI). Previous approaches have used the inverse diffusion tensor field as a Riemannian metric and constructed white matter tracts as geodesics on the resulting manifold. These geodesics have the desirable property that they tend to follow the main eigenvectors of the tensors, yet still have the flexibility to deviate from these directions when it results in lower costs. While this makes such methods more robust to noise, it also has the serious drawback that geodesics tend to deviate from the major eigenvectors in high-curvature areas in order to achieve the shortest path. In this paper we formulate a modification of the Riemannian metric that results in geodesics adapted to follow the principal eigendirection of the tensor even in high-curvature regions. We show that this correction can be formulated as a simple scalar field modulation of the metric and that the appropriate variational problem results in a Poisson's equation on the Riemannian manifold. We demonstrate that the proposed method results in improved geodesics using both synthetic and real DTI data.
Supplemental material: Paper
April 18
Speaker: Corentin Bouchaert
Presentation Title: Brain image processing of cocaine-addict mothers
Description/Abstract:
The present work is part of a multidisciplinary, translational research endeavor involving both animal and human studies. This project is focused on the neurobiological characteristics of cocaine-addict mothers, offspring prenatally exposed to cocaine and rodents exposed to cocaine. Here, we are interested only in the mothers. We will present the dataset. We have tried to find some characteristics using Structural MRI images and DTI images. We will show the working progress, what we are looking for exactly and the preliminary results.
Speaker: Laura Dumont
Presentation Title: Longitudinal DTI Analysis
Description/Abstract:
Disorders such as Autism or Schizophrenia have been hypothesized arising during early brain development. Increasingly used in clinical imaging, longitudinal studies offer us a better understanding of brain development and allow us to study normal and high risk children neurodevelopment. Will be presented in this talk two large studies, CONTE (Early Brain Development in Children at High-Risk for schizophrenia) and ACE (Longitudinal study of infant sibling at risk for Autism). Process and results will be discussed.
Speaker: James Fishbaugh
Presentation Title: Estimation of Smooth Growth Trajectories with Controlled Acceleration from Time Series Shape Data
Description/Abstract:
Longitudinal shape analysis often relies on the estimation of a realistic continuous growth scenario from data sparsely distributed in time. In this work, we propose a new type of growth model parameterized by acceleration, whereas standard methods usually control the velocity. This mimics the behavior of biological tissue as a mechanical system driven by external forces. The growth trajectories are estimated as continuous flows of deformations, which are twice differentiable. This differs from piecewise geodesic regression, for which the velocity may be discontinuous. We evaluate our approach on a set of anatomical structures of the same subject, scanned 16 times between 4 and 8 years of age. We show our acceleration based method does not greatly alter the goodness of fit of the estimation compared to piecewise geodesic regression. Leave-several-out experiments show that our method is robust to missing observations, as well as being less sensitive to noise, and is therefore more likely to capture the underlying biological growth.
Slides: PDF
April 25
Speaker: Alfred Inselberg
Information:
Dr. Alfred Inselberg, inventor of parallel coordinates, is a senior fellow at San Diego Supercomputing Center and a professor in computer science and applied mathematics departments at the Tel Aviv University. This is his personal home page: http://www.math.tau.ac.il/~aiisreal/
