Archived Weekly Agendas for Fall 2009
From Utah Center for Neuroimage Analysis
Contents |
August 24th
Introduction by Guido and a short presentation of the computation environment by Sylvain
August 31st
Sam Preston - Overview of software being developed by Sarang's lab
Sam Gerber - MICCAI Practice talk
Title: On the Manifold Structure of the Space of Brain Images
Abstract: This paper investigates an approach to model the space of brain images through a low-dimensional manifold. A data driven method to learn a manifold from a collections of brain images is proposed. We hypothesize that the space spanned by a set of brain images can be captured, to some approximation, by a low-dimensional manifold, i.e. a parametrization of the set of images. The approach builds on recent advances in manifold learning that allow to uncover nonlinear trends in data.We combine this manifold learning with distance measures between images that capture shape, in order to learn the underlying structure of a database of brain images. The proposed method is generative. New images can be created from the manifold parametrization and existing images can be projected onto the manifold. By measuring projection distance of a held out set of brain images we evaluate the fit of the proposed manifold model to the data and we can compute statistical properties of the data using this manifold structure. We demonstrate this technology on a database of 436 MR brain images.
Paper: http://www.cs.utah.edu/~sgerber/research/brainmanifold.pdf
September 7th
No imaging lunch - Labor Day
September 14th
Manasi Datar - Presentation of MICCAI poster.
Title: Particle Based Shape Regression of Open Surfaces with Applications to Developmental Neuroimaging
Abstract: Shape regression promises to be an important tool to study the relationship between anatomy and underlying clinical or biological parameters, such as age. In this paper we propose a new method to building shape models that incorporates regression analysis in the process of optimizing correspondences on a set of open surfaces. The statistical significance of the dependence is evaluated using permutation tests designed to estimate the likelihood of achieving the observed statistics under numerous rearrangements of the shape parameters with respect to the explanatory variable. We demonstrate the method on synthetic data and provide a new results on clinical MRI data related to early development of the human head.
Sylvain - MICCAI practice talk
Title: Constrained Data Decomposition and Regression for Analyzing Healthy Aging from Fiber Tract Diffusion Properties
Abstract: It has been shown that brain structures in normal aging undergoe significant changes attributed to neurodevelopmental and neu- rodegeneration processes as a lifelong, dynamic process. Modeling changes in healthy aging will be necessary to explain differences to neurode- generative patterns observed in mental illness and neurological disease. Driving application is the analysis of brain white matter properties as a function of age across adulthood, given a database of diffusion tensor images (DTI) of 86 subjects well-balanced across adulthood. We present a methodology based on constrained PCA (CPCA) for fitting age-related changes of white matter diffusion of fiber tracts. It is shown that CPCA applied to tract functions of diffusion isolates population noise and re- tains age as a smooth change over time, well represented by the first principal mode. Age regression on tract functions reveals a nonlinear trajectory but also age-related changes varying locally along tracts. Four tracts with four different tensor-derived scalar diffusion measures were analyzed, and leave-one-out validation of data compression is shown.
September 21st
Tom Fletcher Why averaging works great in 2D but not so great in 3D
Let's say I give you several vectors that came from a multivariate Gaussian distribution. How would you estimate the mean of this distribution? Well, you would probably average the vectors. You might be shocked to hear that this is the best estimate you can get in 2D but not in 3D or higher, where "best" means minimizing mean squared error to the true mean. If you don't believe me, come to Image Lunch and find out why!
September 28th
GPU implementation of the SART algorithm for CT reconstruction
by Yongsheng Pan
Computed tomography on the C-arm CT has been extensively studied and widely used in modern society. Although most manufacturers choose the filtered backprojection algorithm ( FBP ) for its accuracy and efficiency, iterative reconstruction methods have a significant potential to provide superior performance for incomplete, noisy projection data. However, iterative methods have a high computational cost, which hinders their practical use. Furthermore, regularization is usually required to reduce the effects of noise. In this paper, we analyze the use of the Simultaneous Algebraic Reconstruction Technique ( SART ) with total variation (TV ) regularization. Additionally, graphics hardware is utilized to increase the speed of SART implementation. NVIDIA’s GPU and Compute Unified Device Architecture ( CUDA ) comprise the core of our computational platform. The results from the FDK algorithm on the 3D Shepp-Logan phantom and real data are provided in this summary. Experimental results of SART from CPU using cone-beam are also provided on 3- D synthetic images. Preliminary results on 3D synthetic and real images using TV regularization and GPU computation are discussed. This work is performed jointly by the Scientific Computing and Imaging ( SCI ) institute in the University of Utah and GE Healthcare Surgery.
October 5th
Review of MICCAI - Ross and Sarang
October 12th
Fall Break, NO SEMINAR
October 19th
Review of MICCAI
October 26th
Yanfei Mao
Title: fMRI Resting-State Connectivity using Non-Negative Matrix Factorization
Abstract: Investigating resting-state connectivity patterns from Functional Magnetic Resonance Imaging (fMRI) data is a challenging task for any analytical technique. In this project, we proposed to use a new technique based on Non-negative Matrix Factorization (NMF) for the analysis of fMRI resting data, and discuss the role which this exploratory technique could take in the scientific inference of resting-state connectivity patterns. We applied NMF to fMRI data acquired at rest in order to identify the functional connective areas of brain, and demonstrate that this is an effective tool for the identification of low-frequency resting-state patterns. In comparison to Independent Component Analysis (ICA), NMF doesn’t rely on the assumptions of signal independency and non-Gaussian noise distribution. We also show the cortical functional networks exhibit high spatial consistency.
November 2nd
Yaniv Gur
SPD Tensors Regularization via Iwasawa Decomposition
Abstract: In this talk I will present an algorithm for regularization of symmetric positive-definite (SPD) tensors (e.g., diffusion tensors). This algorithm is based on fundamentals of differential geometry, and on unique tensor decomposition, the Iwasawa decomposition. As a result of this decomposition, the original elements of the tensors are replaced by a new set of elements. This new set of elements defines the Iwasawa coordinate system which is used here to parameterize the manifold of SPD tensors. In this framework, the manifold is equipped with the natural GL(n)-invariant Riemannian metric, defined with respect to the new coordinate system. Then, using this metric and a functional over the tensor field, we construct a set of Laplace-Beltrami equations for the Iwasawa coordinates. The numerical solution of this set of equations eventually defines a multi-channel regularization process on the manifold of SPD tensors. To demonstrate this algorithm, at the end of the talk I will present regularization results of real, in vivo, DT-MRI datasets, as well as regularized fiber tractography.
November 9th
Bo Wang
Xiang Hao
Jacob Hinkle
November 16th
Liz Jurrus
Detection of Neuron Membranes in Electron Microscopy Images using Auto-Context
Study of nervous systems via the connectome, the map of connectivities of all neurons in that system, is a challenging problem in neuroscience. Towards this goal, neurobiologists are acquiring large electron microscopy datasets. However, the shear volume of these datasets renders manual analysis infeasible. Hence, automated image analysis methods are required for reconstructing the connectome from these very large image collections. Segmentation of neurons in these images, an essential step of the reconstruction pipeline, is challenging because of noise, anisotropic shapes and brightness, and the presence of confounding structures. The method described in this talk uses a series of artificial neural networks (ANNs) in an auto-context framework combined with a feature vector that is composed of image intensities sampled over a stencil neighborhood. Employing auto-context means that several ANNs are applied in series while allowing each ANN to use the classification context provided by the previous network to improve detection accuracy. We develop the method of serial ANNs for auto-context and show that the learned con- text does improve detection over traditional ANNs. We also demonstrate advantages over previous membrane detection methods. The results are a significant step towards an automated system for the reconstruction of the connectome.
Xiaoyue Huang
November 23rd
Wei Liu Spectral clustering fMRI data with frequency features
Clustering fMRI data is used for finding the brain's functional network. Not like statistical parametric method, it does not assume model for the data, and only use the similarity matrix as the input. Here to define the similarity. we use coherence between to signal as a correlation on the frequency domain. We found this feature is robust to physiological noise. This is the beginning of the research and we also plan to use the spatial neighborhood as a constraint, and in long run, also want to use DTI information as a prior for a semi-supervised learning.
Gopal Veni
Anuja Sharma
November 30th
Avantika Vardhan
James Fishbaugh
December 7th
Neda Sadeghi
Alton Alexander
