Deformable registration of cortical structures via hybrid volumetric and surface warping

Tianming Liu, Dinggang Shen, Christos Davatzikos

Research output: Contribution to journalArticle

Abstract

Registration of cortical structures across individuals is a very important step for quantitative analysis of the human brain cortex. This paper presents a method for deformable registration of cortical structures across individuals, using hybrid volumetric and surface warping. In the first step, a feature-based volumetric registration algorithm is used to warp a model cortical surface to the individual's space. This step greatly reduces the variation between the model and individual, thus providing a good initialization for the next step of surface warping. In the second step, a surface registration method, based on matching geometric attributes, warps the model surface to the individual. Point correspondences are also established at this step. The attribute vector, as the morphological signature of surface, was designed to be as distinctive as possible, so that each vertex on the model surface can find its correspondence on the individual surface. Experimental results on both synthesized and real brain data demonstrate the performance of the proposed method in the registration of cortical structures across individuals.

Original languageEnglish (US)
Pages (from-to)1790-1801
Number of pages12
JournalNeuroImage
Volume22
Issue number4
DOIs
StatePublished - Aug 2004
Externally publishedYes

Fingerprint

Brain

Keywords

  • Cortical structures
  • Deformable registration
  • Hybrid warping

ASJC Scopus subject areas

  • Cognitive Neuroscience
  • Neurology

Cite this

Deformable registration of cortical structures via hybrid volumetric and surface warping. / Liu, Tianming; Shen, Dinggang; Davatzikos, Christos.

In: NeuroImage, Vol. 22, No. 4, 08.2004, p. 1790-1801.

Research output: Contribution to journalArticle

Liu, Tianming ; Shen, Dinggang ; Davatzikos, Christos. / Deformable registration of cortical structures via hybrid volumetric and surface warping. In: NeuroImage. 2004 ; Vol. 22, No. 4. pp. 1790-1801.
@article{8c5d125d58bd463c90ad45b42aa2cefe,
title = "Deformable registration of cortical structures via hybrid volumetric and surface warping",
abstract = "Registration of cortical structures across individuals is a very important step for quantitative analysis of the human brain cortex. This paper presents a method for deformable registration of cortical structures across individuals, using hybrid volumetric and surface warping. In the first step, a feature-based volumetric registration algorithm is used to warp a model cortical surface to the individual's space. This step greatly reduces the variation between the model and individual, thus providing a good initialization for the next step of surface warping. In the second step, a surface registration method, based on matching geometric attributes, warps the model surface to the individual. Point correspondences are also established at this step. The attribute vector, as the morphological signature of surface, was designed to be as distinctive as possible, so that each vertex on the model surface can find its correspondence on the individual surface. Experimental results on both synthesized and real brain data demonstrate the performance of the proposed method in the registration of cortical structures across individuals.",
keywords = "Cortical structures, Deformable registration, Hybrid warping",
author = "Tianming Liu and Dinggang Shen and Christos Davatzikos",
year = "2004",
month = "8",
doi = "10.1016/j.neuroimage.2004.04.020",
language = "English (US)",
volume = "22",
pages = "1790--1801",
journal = "NeuroImage",
issn = "1053-8119",
publisher = "Academic Press Inc.",
number = "4",

}

TY - JOUR

T1 - Deformable registration of cortical structures via hybrid volumetric and surface warping

AU - Liu, Tianming

AU - Shen, Dinggang

AU - Davatzikos, Christos

PY - 2004/8

Y1 - 2004/8

N2 - Registration of cortical structures across individuals is a very important step for quantitative analysis of the human brain cortex. This paper presents a method for deformable registration of cortical structures across individuals, using hybrid volumetric and surface warping. In the first step, a feature-based volumetric registration algorithm is used to warp a model cortical surface to the individual's space. This step greatly reduces the variation between the model and individual, thus providing a good initialization for the next step of surface warping. In the second step, a surface registration method, based on matching geometric attributes, warps the model surface to the individual. Point correspondences are also established at this step. The attribute vector, as the morphological signature of surface, was designed to be as distinctive as possible, so that each vertex on the model surface can find its correspondence on the individual surface. Experimental results on both synthesized and real brain data demonstrate the performance of the proposed method in the registration of cortical structures across individuals.

AB - Registration of cortical structures across individuals is a very important step for quantitative analysis of the human brain cortex. This paper presents a method for deformable registration of cortical structures across individuals, using hybrid volumetric and surface warping. In the first step, a feature-based volumetric registration algorithm is used to warp a model cortical surface to the individual's space. This step greatly reduces the variation between the model and individual, thus providing a good initialization for the next step of surface warping. In the second step, a surface registration method, based on matching geometric attributes, warps the model surface to the individual. Point correspondences are also established at this step. The attribute vector, as the morphological signature of surface, was designed to be as distinctive as possible, so that each vertex on the model surface can find its correspondence on the individual surface. Experimental results on both synthesized and real brain data demonstrate the performance of the proposed method in the registration of cortical structures across individuals.

KW - Cortical structures

KW - Deformable registration

KW - Hybrid warping

UR - http://www.scopus.com/inward/record.url?scp=3242670879&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=3242670879&partnerID=8YFLogxK

U2 - 10.1016/j.neuroimage.2004.04.020

DO - 10.1016/j.neuroimage.2004.04.020

M3 - Article

VL - 22

SP - 1790

EP - 1801

JO - NeuroImage

JF - NeuroImage

SN - 1053-8119

IS - 4

ER -