Book chapter
Diffusion MRI Registration Using Orientation Distribution Functions
Information Processing in Medical Imaging, pp.626-637
Lecture Notes in Computer Science, Springer Berlin Heidelberg
2009
DOI: 10.1007/978-3-642-02498-6_52
PMCID: PMC3860746
PMID: 19694299
Abstract
We propose a linear-elastic registration method to register diffusion-weighted MRI (DW-MRI) data sets by mapping their diffusion orientation distribution functions (ODFs). The ODFs were reconstructed using a q-ball imaging (QBI) technique to resolve intravoxel fiber crossing. The registration method is based on mapping the ODF maps represented by spherical harmonics which yield analytic solutions and reduce the computational complexity. ODF reorientation is required to maintain the consistency with transformed local fiber directions. The reorientation matrices are extracted from the local Jacobian and directly applied to the coefficients of spherical harmonics. The similarity cost of the registration is defined by the ODF shape distance calculated from the spherical harmonic coefficients. The transformation fields are regularized by linear elastic constraints. The proposed method was validated using both synthetic and real data sets. Experimental results show that the elastic registration improved the affine alignment by further reducing the ODF shape difference; reorientation during the registration produced registered ODF maps with more consistent principle directions compared to registrations without reorientation or simultaneous reorientation.
Details
- Title: Subtitle
- Diffusion MRI Registration Using Orientation Distribution Functions
- Creators
- Xiujuan Geng - National Institute on Drug Abuse, NIH, TaiwanThomas J Ross - National Institute on Drug Abuse, NIH, TaiwanWang Zhan - Department of Radiology, University of California, San Francisco, USAHong Gu - National Institute on Drug Abuse, NIH, TaiwanYi-Ping Chao - Department of Electrical Engineering, National Taiwan University, TaiwanChing-Po Lin - Institute of Neuroscience, National Yang-Ming University, TaiwanGary E Christensen - Department of Electrical and Computer Engineering, University of Iowa, USANorbert Schuff - Department of Radiology, University of California, San Francisco, USAYihong Yang - National Institute on Drug Abuse, NIH, Taiwan
- Resource Type
- Book chapter
- Publication Details
- Information Processing in Medical Imaging, pp.626-637
- Publisher
- Springer Berlin Heidelberg; Berlin, Heidelberg
- Series
- Lecture Notes in Computer Science
- DOI
- 10.1007/978-3-642-02498-6_52
- PMID
- 19694299
- PMCID
- PMC3860746
- eISSN
- 1611-3349
- ISSN
- 0302-9743
- Language
- English
- Date published
- 2009
- Academic Unit
- Electrical and Computer Engineering; Radiation Oncology; Radiation Research Laboratory
- Record Identifier
- 9984047683402771
Metrics
22 Record Views