[IPAG Home Page] [Publications] [Dissertation Archive] [Personnel]

Segmentation of Anatomical Structure

Paper List

See also: The IPAG Dissertation Archive.

o Acrobat .pdf Format o Gzipped Postscript Format.
See note below for access restrictions.

Papers

o Segmentation and Measurement of the Cortex from 3D MR Images Using Coupled Surfaces Propagation

X. Zeng, L.H.Staib, R.T.Schultz and J.S.Duncan.
In IEEE Trans. on Medical Imaging, October, 1999.

Abstract

The cortex is the outermost thin layer of gray matter in the brain; geometric measurement of the cortex helps in understanding brain anatomy and function. In the quantitat ive analysis of the cortex from MR images, extracting the structure and obtaining a representation for various measurements are key steps. While manual segmentation is t edious and labor intensive, automatic, reliable and efficient segmentation and measurement of the cortex remain challenging problems due to its convoluted nature. Here w e present a new approach of coupled surfaces propagation using level set methods to address such problems. Our method is motivated by the nearly constant thickness of t he cortical mantle and takes this tight coupling as an important constraint. By evolving two embedded surfaces simultaneously, each driven by its own image-derived inform ation while maintaining the coupling, a final representation of the cortical bounding surfaces and an automatic segmentation of the cortex are achieved. Characteristics of the cortex such as cortical surface area, surface curvature and cortical thickness are then evaluated. The level set implementation of surface propagation offers the adv antage of easy initialization, computational efficiency and the ability to capture deep sulcal folds. Results and validation from various experiments on both simulated and real 3D MR images are provided.

BibTeX Entry

@Article{ZengIEEE,
author ="X. Zeng and  L. H. Staib and  R. T. Schultz and
	J. S. Duncan",
title  ="Segmentation and Measurement of the Cortex from 3{D} {MR} 
	Images Using Coupled Surfaces Propagation",
journal="IEEE Transactions on Medical Imaging",
year   ="1999",
month  ="October",
pages  ="100-111"}

oA New Approach to 3D Sulcal Ribbon Finding from MR Images

X.Zeng, L.H.Staib, R.T.Schultz, H.Tagare, L.Win and J.S.Duncan.
In Proceedings of the Second International Conference on Medical Image Computing and Computer-Assisted Intervention, September 1999.

Abstract

Sulcal medial surfaces are 3D thin convoluted ribbons embedded in cortical sulci, and they provide distinctive anatomical features of the brain. Here we propose a new approach to automatic intrasulcal ribbon finding, following our work on cortex segmentation with coupled surfaces via level set methods, where the outer cortical surface is embedded as the zero level set of a high-dimensional distance function. Through the utilization of this distance function, we are able to formulate the sulcal ribbon finding problem as one of surface deformation, thus avoiding possible control problems in other work using sliding contour models. Using dynamic programming and deformable surface models, our method requires little manual intervention and results parameterized sulcal ribbon surfaces in nearly real-time. Though a natural follow up to our earlier segmentation work, we describe how it can be applied with general segmentation methods. We also present quantitative results on 15 MR brain images.

BibTeX Entry

@InProceedings{ZengMiccai99,
author   ="X. Zeng and L. H. Staib and R. T. Schultz and
	  H. Tagare and L. Win and J. S. Duncan",
title    ="A new approach to 3{D} sulcal ribbon finding from
	  {MR} images",
pages    ="148-157",
booktitle="Proceedings of Medical Image Computing and Computer-Assisted
	  Intervention ",
year     ="1999",
address  ="Cambridge, UK"}

o Segmentation and Measurementof the Cortex from 3D MR Images

X. Zeng, L.H.Staib, R.T.Schultz and J.S.Duncan.
In Proceedings of the First International Conference on Medical Image Computing and Computer-Assisted Intervention, October, 1998.

Abstract

The cortex is the outermost thin layer of gray matter in the brain; geometric measurement of the cortex helps in understanding brain anatomy and function. In the quantitative analysis of the cortex from MR images, extracting the structure and obtaining a representation for various measurements are key steps. While manual segmentation is tedious and labor intensive, automatic, reliable and efficient segmentation and measurement of the cortex remain challenging problems due to its convoluted nature. A new approach of coupled surfaces propagation using level set methods is presented here for the problem of the segmentation and measurement of the cortex. Our method is motivated by the nearly constant thickness of the cortical mantle and takes this tight coupling as an important constraint. By evolving two embedded surfaces simultaneously, each driven by its own image-derived information while maintaining the coupling, a final representation of the cortical bounding surfaces and an automatic segmentation of the cortex are achieved. Characteristics of the cortex such as cortical surface area, surface curvature and thickness are then evaluated. The level set implementation of surface propagation offers the advantage of easy initialization, computational efficiency and the ability to capture deep folds of the sulci. Results and validation from various experiments on simulated and real 3D MR images are provided.

BibTeX Entry

@InProceedings{ZengMiccai98,
author =   "X. Zeng and L. H. Staib and R. T. Schultz and J. S. Duncan",
title =    "Segmentation and measurement of the cortex from {3{D} MR} images",
pages =    "519-530",
booktitle ="Proceedings of Medical Image Computing and Computer-Assisted 
	    Intervention ",
year =     "1998",
address =  "MIT"}

o Boundary finding with correspondence using statistical shape models

Y. Wang and L. H. Staib
In Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition, Stanta Barbara, CA, June 1998.

Abstract

We propose an approach for boundary finding where the correspondence of a subset of boundary points to a model is simultaneously determined. Global shape parameters derived from the statistical variation of object boundary points in a training set are used to model the object. A Bayesian formulation, based on this prior knowledge and the edge information of the input image, is employed to find the object boundary with its subset points in correspondence with boundaries in the training set or the mean boundary.

We compared the use of a generic smoothness prior and a uniform independent prior with the training set prior in order to demonstrate the power of this statistical information.

A number of experiments were performed on both synthetic and real medical images of the brain and heart to evaluate the approach, including the validation of the dependence of the method on image quality, different initialization and prior information.

BibTeX Entry

@InProceedings{WangCvpr98,
author =   "Y. Wang and  L. H. Staib",
title =    "Boundary finding with correspondence using statistical shape models",
pages =    "338-345",
booktitle ="Proceedings of the IEEE Conference on Computer Vision 
            and Pattern Recognition",
year =     "June 1998",
address =  "Santa Barbara, CA"}

o Volumetric Layer Segmentation Using Coupled Surfaces Propagation

X. Zeng, L.H.Staib, R.T.Schultz and J.S.Duncan.
Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition, 1998.

Abstract

The problem of segmenting a volumetric layer of finite thickness is encountered in several important areas within medical image analysis. Key examples include the extrac tion of the cortical gray matter of the brain and the left ventricle myocardium of the heart. The coupling between the two bounding surfaces of such a layer provides impo rtant information that helps to solve the segmentation problem. Here we propose a new approach of coupled surfaces propagation via level set methods, which takes into ac count coupling as an important constraint. By evolving two embedded surfaces simultaneously, each driven by its own image-derived information while maintaining the coupl ing, we capture a representation of the two bounding surfaces and achieve automatic segmentation on the layer. Characteristic gray level values, instead of image gra dient information alone, are incorporated in deriving the useful image information to drive the surface propagation, which enables our approach to capture the homogenei ty inside the layer. The level set implementation offers the advantage of easy initialization, computational efficiency and the ability to capture deep folds of the su lci. As a test example, we apply our approach to unedited 3D Magnetic Resonance(MR) brain images. Our algorithm automatically isolates the brain from non-brain structur es and recovers the cortical gray matter.

BibTeX Entry

@InProceedings{ZengCvpr,
author =   "X. Zeng and  L. H. Staib and  R. T. Schultz and
	    J. S. Duncan",
title =    "Volumetric layer segmentation using coupled 
	    surfaces propagation",
pages =    "708-715",
booktitle ="Proceedings of the IEEE Conference on Computer Vision 
	    and Pattern Recognition",
year =     "1998",
address =  "Santa Barbara, CA"}

o On Multi-Feature Integration for Deformable Boundary Finding

A. Chakraborty, Marcel Worring and J. Duncan.
International Conference on Computer Vision (ICCV) 1995.

Abstract

Precise segmentation of underlying objects in an image is very important especially for biomedical image analysis. Here, we present an integrated approach for boundary finding using region and curvature information along with the gradient. Unlike the previous methods, where smoothing is enforced by penalizing curvature, here the grey level curvature is used as an extra source of information. However, information fusion may not be useful unless used properly. To address that, we present results that highlight the pros and cons of using the various sources of information and indicate when one should get precedence over the others.

BibTeX Entry

@article(chakrabortyiccv95,
author="A.  Chakraborty and M.  Worring and J.  S.  Duncan",
title="On Multi-Feature Integration for Deformable Boundary Finding",
journal=piccv,
pages="846-851",
year="1995")

o Integration of Boundary Finding and Region-Based Segmentation Using Game Theory

A. Chakraborty, and J. S. Duncan
Information Processing in Medical Imaging (IPMI) 1995.

Abstract

Robust segmentation of structures from an image is essential for a var iety of applications in biomedical image analysis. Here we propose a method that integrates region based segmentation and gradient based boundary finding using game theory in an effort to form a unified approach that is robust to noise and poor initialization. The novelty of the method is that this is a bi-directional framework whereby the two seperate modules improve their results through mutual information sharing.

BibTeX Entry

@inproceedings(chakrabortyipmi95,
author="A. Chakraborty and J.S. Duncan",
title ="Integration of Boundary Finding and Region-based
       Segmentation Using Game Theory",
booktitle="XIVth International Conference on Information 
         Processing in Medical Imaging",
publisher="Kluwer press",
pages="189-200",
year="1995")

Note:

  1. The actual papers are only accessible only from computers at Yale University or authorized external users. In that case a valid access username/password is needed. IPAG Members: The accounts for this page are not the same as your usual noodle accounts. Contact webmaster@noodle.med.yale.edu for more information.
  2. You may need to install Adobe Acrobat Reader to view or print these files. (Alternatively on unix machines try xpdf.)
Questions regarding the content of this page should be directed to the group director.

Questions or bug reports regarding this interface should go to webmaster@noodle.med.yale.edu.

Home Page, webmaster@noodle.med.yale.edu