Multi-scale regularization approaches of non-parametric deformable registrations

Hsiang-Chi Kuo, Keh Shih Chuang, Dennis Mah, Andrew Wu, Linda Hong, Ravindra Yaparpalvi, Shalom Kalnicki

Research output: Contribution to journalArticle

Abstract

Most deformation algorithms use a single-value smoother during optimization. We investigate multi-scale regularizations (smoothers) during the multi-resolution iteration of two non-parametric deformable registrations (demons and diffeomorphic algorithms) and compare them to a conventional single-value smoother. Our results show that as smoothers increase, their convergence rate decreases; however, smaller smoothers also have a large negative value of the Jacobian determinant suggesting that the one-to-one mapping has been lost; i.e., image morphology is not preserved. A better one-to-one mapping of the multiscale scheme has also been established by the residual vector field measures. In the demons method, the multiscale smoother calculates faster than the large single-value smoother (Gaussian kernel width larger than 0.5) and is equivalent to the smallest single-value smoother (Gaussian kernel width equals to 0.5 in this study). For the diffeomorphic algorithm, since our multi-scale smoothers were implemented at the deformation field and the update field, calculation times are longer. For the deformed images in this study, the similarity measured by mean square error, normal correlation, and visual comparisons show that the multi-scale implementation has better results than large single-value smoothers, and better or equivalent for smallest single-value smoother. Between the two deformable registrations, diffeormophic method constructs better coherence space of the deformation field while the deformation is large between images.

Original languageEnglish (US)
Pages (from-to)586-597
Number of pages12
JournalJournal of Digital Imaging
Volume24
Issue number4
DOIs
StatePublished - Aug 2011

Fingerprint

Mean square error

Keywords

  • Deformation registration
  • Demon algorithm
  • Diffeomorphic algorithm
  • Multi-scale regularization

ASJC Scopus subject areas

  • Radiology Nuclear Medicine and imaging
  • Radiological and Ultrasound Technology
  • Computer Science Applications
  • Medicine(all)

Cite this

Multi-scale regularization approaches of non-parametric deformable registrations. / Kuo, Hsiang-Chi; Chuang, Keh Shih; Mah, Dennis; Wu, Andrew; Hong, Linda; Yaparpalvi, Ravindra; Kalnicki, Shalom.

In: Journal of Digital Imaging, Vol. 24, No. 4, 08.2011, p. 586-597.

Research output: Contribution to journalArticle

@article{bb629a262332462db9018fd3701f46d5,
title = "Multi-scale regularization approaches of non-parametric deformable registrations",
abstract = "Most deformation algorithms use a single-value smoother during optimization. We investigate multi-scale regularizations (smoothers) during the multi-resolution iteration of two non-parametric deformable registrations (demons and diffeomorphic algorithms) and compare them to a conventional single-value smoother. Our results show that as smoothers increase, their convergence rate decreases; however, smaller smoothers also have a large negative value of the Jacobian determinant suggesting that the one-to-one mapping has been lost; i.e., image morphology is not preserved. A better one-to-one mapping of the multiscale scheme has also been established by the residual vector field measures. In the demons method, the multiscale smoother calculates faster than the large single-value smoother (Gaussian kernel width larger than 0.5) and is equivalent to the smallest single-value smoother (Gaussian kernel width equals to 0.5 in this study). For the diffeomorphic algorithm, since our multi-scale smoothers were implemented at the deformation field and the update field, calculation times are longer. For the deformed images in this study, the similarity measured by mean square error, normal correlation, and visual comparisons show that the multi-scale implementation has better results than large single-value smoothers, and better or equivalent for smallest single-value smoother. Between the two deformable registrations, diffeormophic method constructs better coherence space of the deformation field while the deformation is large between images.",
keywords = "Deformation registration, Demon algorithm, Diffeomorphic algorithm, Multi-scale regularization",
author = "Hsiang-Chi Kuo and Chuang, {Keh Shih} and Dennis Mah and Andrew Wu and Linda Hong and Ravindra Yaparpalvi and Shalom Kalnicki",
year = "2011",
month = "8",
doi = "10.1007/s10278-010-9313-6",
language = "English (US)",
volume = "24",
pages = "586--597",
journal = "Journal of Digital Imaging",
issn = "0897-1889",
publisher = "Springer New York",
number = "4",

}

TY - JOUR

T1 - Multi-scale regularization approaches of non-parametric deformable registrations

AU - Kuo, Hsiang-Chi

AU - Chuang, Keh Shih

AU - Mah, Dennis

AU - Wu, Andrew

AU - Hong, Linda

AU - Yaparpalvi, Ravindra

AU - Kalnicki, Shalom

PY - 2011/8

Y1 - 2011/8

N2 - Most deformation algorithms use a single-value smoother during optimization. We investigate multi-scale regularizations (smoothers) during the multi-resolution iteration of two non-parametric deformable registrations (demons and diffeomorphic algorithms) and compare them to a conventional single-value smoother. Our results show that as smoothers increase, their convergence rate decreases; however, smaller smoothers also have a large negative value of the Jacobian determinant suggesting that the one-to-one mapping has been lost; i.e., image morphology is not preserved. A better one-to-one mapping of the multiscale scheme has also been established by the residual vector field measures. In the demons method, the multiscale smoother calculates faster than the large single-value smoother (Gaussian kernel width larger than 0.5) and is equivalent to the smallest single-value smoother (Gaussian kernel width equals to 0.5 in this study). For the diffeomorphic algorithm, since our multi-scale smoothers were implemented at the deformation field and the update field, calculation times are longer. For the deformed images in this study, the similarity measured by mean square error, normal correlation, and visual comparisons show that the multi-scale implementation has better results than large single-value smoothers, and better or equivalent for smallest single-value smoother. Between the two deformable registrations, diffeormophic method constructs better coherence space of the deformation field while the deformation is large between images.

AB - Most deformation algorithms use a single-value smoother during optimization. We investigate multi-scale regularizations (smoothers) during the multi-resolution iteration of two non-parametric deformable registrations (demons and diffeomorphic algorithms) and compare them to a conventional single-value smoother. Our results show that as smoothers increase, their convergence rate decreases; however, smaller smoothers also have a large negative value of the Jacobian determinant suggesting that the one-to-one mapping has been lost; i.e., image morphology is not preserved. A better one-to-one mapping of the multiscale scheme has also been established by the residual vector field measures. In the demons method, the multiscale smoother calculates faster than the large single-value smoother (Gaussian kernel width larger than 0.5) and is equivalent to the smallest single-value smoother (Gaussian kernel width equals to 0.5 in this study). For the diffeomorphic algorithm, since our multi-scale smoothers were implemented at the deformation field and the update field, calculation times are longer. For the deformed images in this study, the similarity measured by mean square error, normal correlation, and visual comparisons show that the multi-scale implementation has better results than large single-value smoothers, and better or equivalent for smallest single-value smoother. Between the two deformable registrations, diffeormophic method constructs better coherence space of the deformation field while the deformation is large between images.

KW - Deformation registration

KW - Demon algorithm

KW - Diffeomorphic algorithm

KW - Multi-scale regularization

UR - http://www.scopus.com/inward/record.url?scp=80053902482&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=80053902482&partnerID=8YFLogxK

U2 - 10.1007/s10278-010-9313-6

DO - 10.1007/s10278-010-9313-6

M3 - Article

C2 - 20574767

AN - SCOPUS:80053902482

VL - 24

SP - 586

EP - 597

JO - Journal of Digital Imaging

JF - Journal of Digital Imaging

SN - 0897-1889

IS - 4

ER -