Namic activities utah
Download
1 / 11

NAMIC Activities – Utah - PowerPoint PPT Presentation


  • 111 Views
  • Uploaded on

NAMIC Activities – Utah. HD. AF. HNC. TBI. Image/Shape Analysis Volumetric tractography and DTI atlases Robust correspondences for shape Longitudinal shape analysis Segmentation Globally optimal surface estimation Fast, feature/shape-based image lookup Registration

loader
I am the owner, or an agent authorized to act on behalf of the owner, of the copyrighted work described.
capcha
Download Presentation

PowerPoint Slideshow about ' NAMIC Activities – Utah' - hien


An Image/Link below is provided (as is) to download presentation

Download Policy: Content on the Website is provided to you AS IS for your information and personal use and may not be sold / licensed / shared on other websites without getting consent from its author.While downloading, if for some reason you are not able to download a presentation, the publisher may have deleted the file from their server.


- - - - - - - - - - - - - - - - - - - - - - - - - - E N D - - - - - - - - - - - - - - - - - - - - - - - - - -
Presentation Transcript
Namic activities utah
NAMIC Activities – Utah

HD

AF

HNC

TBI

  • Image/Shape Analysis

    • Volumetric tractographyand DTI atlases

    • Robust correspondences for shape

    • Longitudinal shape analysis

  • Segmentation

    • Globally optimal surface estimation

    • Fast, feature/shape-based image lookup

  • Registration

    • Robust metrics for image match

    • Nonsmoothregularizers


Volume tractography
Volume Tractography

  • DTI Atlases to define ROIs

    • Goodletet al.

  • Volumetric tractography for white matter regions

    • Fletcher et al.

  • Group analysis on volumetric regions


Robust shape correspondences
Robust Shape Correspondences

  • Datar et al. 2011

    • Statistics of points and normals

    • Geodesic distances for point-to-point interactions


Longitudinal shape parameterization with t fletcher m datar
Longitudinal Shape Parameterizationwith T. Fletcher, M. Datar

Left atrium trend – before and after ablation

  • Mixed effects model

    • Hierarchical–properly accounts for staggered/missing data in individuals


Segmentation
Segmentation

  • Graph-based image segmentation

  • Represent surfaces a min-cut in properly ordered graph

    • Liu et al. 2009

  • Challenges

    • Objective functions that capture features, smoothness, coupled surfaces

    • Generalizations to 3D shapes



Fast nearest neighbor lookup
Fast Nearest Neighbor Lookup

  • Problem: from a large database of images, most similar images combine to form best segmentation

    • Label voting or nonparametric modeling paradigm

    • Especially important for heterogeneous data

      • Head&neck, cardiac

  • How to find similar shapes?

    • Deformation (slow)

    • Feature-based query (fast)


Feature based lookup
Feature-Based Lookup

  • Strategy

    • Detect features and compare hierarchically

      • Pyramid matching (Grauman 2006)


Robust image registration
Robust Image Registration

  • Applications

    • Correspondence or coordinate system for comparing different individuals or times

      • E.g. longitudinal TBI, Afib before and after ablation

    • Segmentation from atlases or label voting

      • E.g. head and neck, endocardium from DCE

  • Challenges

    • Nonsmooth transformations

      • Singularities, tearing, sliding

    • Outliers/mismatches


Registration formulation
Registration Formulation

  • Strategy

    • Image match: apply robust versions of image metric (e.g. other norms)

    • Regularization: norms and operators that allow for nonsmooth (noninvertable) transformations

  • Issues

    • Well posedness

    • Optimization



ad