Publication detail

Registration of FA and T1-weighted MRI Data of Healthy Human Brain based on Template Matching and Normalized Cross-correlation

MALÍNSKÝ, M. PETER, R. HODNELAND, E. LUNDERVOLD, A. LUNDERVOLD, A. JAN, J.

Original Title

Registration of FA and T1-weighted MRI Data of Healthy Human Brain based on Template Matching and Normalized Cross-correlation

Czech Title

Lícování FA a T1-váhovaných MRI dat lidského mozku založena na metodě Template Matching a normalizované křížové korelaci

English Title

Registration of FA and T1-weighted MRI Data of Healthy Human Brain based on Template Matching and Normalized Cross-correlation

Type

journal article

Language

en

Original Abstract

Image registration is an essential process in a large range of medical image applications helping medical experts to increase the structural information and describe mutual relations between images. In a human brain studies it is a crucial step to spatially align diffusion tensor images (DTI) and anatomy data to quantitatively compare neural and structure features obtained from the same subject at different time-points. In this paper we propose a registration workflow for a reliable alignment of diffusion weighted fractional anisotropy (FA) images and T1-weighted high-resolution anatomical images of more than hundred subjects. Use of Template matching algorithm yields the robust inner subject registration and speed up of the registration process which helps in better assess of structural and functional brain connectivity. This method has been compared with well-known B-spline multiresolution registration algorithm.

Czech abstract

Lícování obrazu je stále velmi diskutovanou oblastí v lékařském zobrazování a hraje také pomocnou roli při zlepšování informovanosti o strukturách mozku. Lícování DTI dat a dat anatomických hraje při studiích mozku velmi významnou roli. Tento článek se zabývá fúzí DTI FA a T1 váhovaných MRI dat mozku . Součástí tohoto článku je také návrh hodnocení výsledků sestávající se ze čtyřech hodnocených kritérií. Výsledky této registrace jsou porovnány s B-Spline metodou registrace. Metoda byla otestována na více než stech pacientech a na základě výsledků prohlášena za robustní a spolehlivou.

English abstract

Image registration is an essential process in a large range of medical image applications helping medical experts to increase the structural information and describe mutual relations between images. In a human brain studies it is a crucial step to spatially align diffusion tensor images (DTI) and anatomy data to quantitatively compare neural and structure features obtained from the same subject at different time-points. In this paper we propose a registration workflow for a reliable alignment of diffusion weighted fractional anisotropy (FA) images and T1-weighted high-resolution anatomical images of more than hundred subjects. Use of Template matching algorithm yields the robust inner subject registration and speed up of the registration process which helps in better assess of structural and functional brain connectivity. This method has been compared with well-known B-spline multiresolution registration algorithm.

Keywords

Brain, Neuroscience, Magnetic resonance, Multimodal image registration, Template matching, Diffusion tensor imaging

RIV year

2012

Released

10.12.2012

Publisher

Springer

Location

Německo

Pages from

1

Pages to

12

Pages count

12

BibTex


@article{BUT96701,
  author="Miloš {Malínský} and Roman {Peter} and Erlend {Hodneland} and A.J. {Lundervold} and Arvid {Lundervold} and Jiří {Jan}",
  title="Registration of FA and T1-weighted MRI Data of Healthy Human Brain based on Template Matching and Normalized Cross-correlation",
  annote="Image registration is an essential process in a large range of medical image applications helping medical experts to increase the structural information and describe mutual relations between images. In a human brain studies it is a crucial step to spatially align diffusion tensor images (DTI) and anatomy data to quantitatively compare neural and structure features obtained from the same subject at different time-points. In this paper we propose a registration workflow for a reliable alignment of diffusion weighted fractional anisotropy (FA) images and T1-weighted high-resolution anatomical images of more than hundred subjects. Use of Template matching algorithm yields the robust inner subject registration and speed up of the registration process which helps in better assess of structural and functional brain connectivity. This method has been compared with well-known B-spline multiresolution registration algorithm.",
  address="Springer",
  chapter="96701",
  institution="Springer",
  number="6",
  volume="25",
  year="2012",
  month="december",
  pages="1--12",
  publisher="Springer",
  type="journal article"
}