Regional infarction identification from cardiac CT images: a computer-aided biomechanical approach

Ken C L Wong, Michael Tee, Marcus Chen, David A. Bluemke, Ronald M. Summers, Jianhua Yao

Research output: Contribution to journalArticle

Abstract

Purpose: Regional infarction identification is important for heart disease diagnosis and management, and myocardial deformation has been shown to be effective for this purpose. Although tagged and strain-encoded MR images can provide such measurements, they are uncommon in clinical routine. On the contrary, cardiac CT images are more available with lower costs, but they only provide motion of cardiac boundaries and additional constraints are required to obtain the myocardial strains. The goal of this study is to verify the potential of contrast-enhanced CT images on computer-aided regional infarction identification. Methods: We propose a biomechanical approach combined with machine learning algorithms. A hyperelastic biomechanical model is used with deformable image registration to estimate 3D myocardial strains from CT images. The regional strains and CT image intensities are input to a classifier for regional infarction identification. Cross-validations on ten canine image sequences with artificially induced infarctions were used to study the performances of using different feature combinations and machine learning algorithms. Results: Radial strain, circumferential strain, first principal strain, and image intensity were shown to be discriminative features. The highest identification accuracy ((Formula presented.) %) was achieved when combining radial strain with image intensity. Random forests gave better results than support vector machines on less discriminative features. Random forests also performed better when all strains were used together. Conclusion: Although CT images cannot directly measure myocardial deformation, with the use of a biomechanical model, the estimated strains can provide promising identification results especially when combined with CT image intensity.

Original languageEnglish (US)
Pages (from-to)1-11
Number of pages11
JournalInternational journal of computer assisted radiology and surgery
DOIs
StateAccepted/In press - Apr 12 2016
Externally publishedYes

Fingerprint

Infarction
Disease Management
Canidae
Heart Diseases
Learning algorithms
Learning systems
Costs and Cost Analysis
Image registration
Support vector machines
Identification (control systems)
Classifiers
Machine Learning
Forests

Keywords

  • Biomechanics
  • Cardiac computed tomography
  • Computer-aided diagnosis
  • Infarction identification
  • Machine learning
  • Myocardial strain estimation

ASJC Scopus subject areas

  • Radiology Nuclear Medicine and imaging
  • Health Informatics
  • Surgery

Cite this

Regional infarction identification from cardiac CT images : a computer-aided biomechanical approach. / Wong, Ken C L; Tee, Michael; Chen, Marcus; Bluemke, David A.; Summers, Ronald M.; Yao, Jianhua.

In: International journal of computer assisted radiology and surgery, 12.04.2016, p. 1-11.

Research output: Contribution to journalArticle

Wong, Ken C L ; Tee, Michael ; Chen, Marcus ; Bluemke, David A. ; Summers, Ronald M. ; Yao, Jianhua. / Regional infarction identification from cardiac CT images : a computer-aided biomechanical approach. In: International journal of computer assisted radiology and surgery. 2016 ; pp. 1-11.
@article{08e3a4cd1a6b4be1ad74733e2881a064,
title = "Regional infarction identification from cardiac CT images: a computer-aided biomechanical approach",
abstract = "Purpose: Regional infarction identification is important for heart disease diagnosis and management, and myocardial deformation has been shown to be effective for this purpose. Although tagged and strain-encoded MR images can provide such measurements, they are uncommon in clinical routine. On the contrary, cardiac CT images are more available with lower costs, but they only provide motion of cardiac boundaries and additional constraints are required to obtain the myocardial strains. The goal of this study is to verify the potential of contrast-enhanced CT images on computer-aided regional infarction identification. Methods: We propose a biomechanical approach combined with machine learning algorithms. A hyperelastic biomechanical model is used with deformable image registration to estimate 3D myocardial strains from CT images. The regional strains and CT image intensities are input to a classifier for regional infarction identification. Cross-validations on ten canine image sequences with artificially induced infarctions were used to study the performances of using different feature combinations and machine learning algorithms. Results: Radial strain, circumferential strain, first principal strain, and image intensity were shown to be discriminative features. The highest identification accuracy ((Formula presented.) {\%}) was achieved when combining radial strain with image intensity. Random forests gave better results than support vector machines on less discriminative features. Random forests also performed better when all strains were used together. Conclusion: Although CT images cannot directly measure myocardial deformation, with the use of a biomechanical model, the estimated strains can provide promising identification results especially when combined with CT image intensity.",
keywords = "Biomechanics, Cardiac computed tomography, Computer-aided diagnosis, Infarction identification, Machine learning, Myocardial strain estimation",
author = "Wong, {Ken C L} and Michael Tee and Marcus Chen and Bluemke, {David A.} and Summers, {Ronald M.} and Jianhua Yao",
year = "2016",
month = "4",
day = "12",
doi = "10.1007/s11548-016-1404-5",
language = "English (US)",
pages = "1--11",
journal = "Computer-Assisted Radiology and Surgery",
issn = "1861-6410",
publisher = "Springer Verlag",

}

TY - JOUR

T1 - Regional infarction identification from cardiac CT images

T2 - a computer-aided biomechanical approach

AU - Wong, Ken C L

AU - Tee, Michael

AU - Chen, Marcus

AU - Bluemke, David A.

AU - Summers, Ronald M.

AU - Yao, Jianhua

PY - 2016/4/12

Y1 - 2016/4/12

N2 - Purpose: Regional infarction identification is important for heart disease diagnosis and management, and myocardial deformation has been shown to be effective for this purpose. Although tagged and strain-encoded MR images can provide such measurements, they are uncommon in clinical routine. On the contrary, cardiac CT images are more available with lower costs, but they only provide motion of cardiac boundaries and additional constraints are required to obtain the myocardial strains. The goal of this study is to verify the potential of contrast-enhanced CT images on computer-aided regional infarction identification. Methods: We propose a biomechanical approach combined with machine learning algorithms. A hyperelastic biomechanical model is used with deformable image registration to estimate 3D myocardial strains from CT images. The regional strains and CT image intensities are input to a classifier for regional infarction identification. Cross-validations on ten canine image sequences with artificially induced infarctions were used to study the performances of using different feature combinations and machine learning algorithms. Results: Radial strain, circumferential strain, first principal strain, and image intensity were shown to be discriminative features. The highest identification accuracy ((Formula presented.) %) was achieved when combining radial strain with image intensity. Random forests gave better results than support vector machines on less discriminative features. Random forests also performed better when all strains were used together. Conclusion: Although CT images cannot directly measure myocardial deformation, with the use of a biomechanical model, the estimated strains can provide promising identification results especially when combined with CT image intensity.

AB - Purpose: Regional infarction identification is important for heart disease diagnosis and management, and myocardial deformation has been shown to be effective for this purpose. Although tagged and strain-encoded MR images can provide such measurements, they are uncommon in clinical routine. On the contrary, cardiac CT images are more available with lower costs, but they only provide motion of cardiac boundaries and additional constraints are required to obtain the myocardial strains. The goal of this study is to verify the potential of contrast-enhanced CT images on computer-aided regional infarction identification. Methods: We propose a biomechanical approach combined with machine learning algorithms. A hyperelastic biomechanical model is used with deformable image registration to estimate 3D myocardial strains from CT images. The regional strains and CT image intensities are input to a classifier for regional infarction identification. Cross-validations on ten canine image sequences with artificially induced infarctions were used to study the performances of using different feature combinations and machine learning algorithms. Results: Radial strain, circumferential strain, first principal strain, and image intensity were shown to be discriminative features. The highest identification accuracy ((Formula presented.) %) was achieved when combining radial strain with image intensity. Random forests gave better results than support vector machines on less discriminative features. Random forests also performed better when all strains were used together. Conclusion: Although CT images cannot directly measure myocardial deformation, with the use of a biomechanical model, the estimated strains can provide promising identification results especially when combined with CT image intensity.

KW - Biomechanics

KW - Cardiac computed tomography

KW - Computer-aided diagnosis

KW - Infarction identification

KW - Machine learning

KW - Myocardial strain estimation

UR - http://www.scopus.com/inward/record.url?scp=84963653991&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=84963653991&partnerID=8YFLogxK

U2 - 10.1007/s11548-016-1404-5

DO - 10.1007/s11548-016-1404-5

M3 - Article

C2 - 27072840

AN - SCOPUS:84963653991

SP - 1

EP - 11

JO - Computer-Assisted Radiology and Surgery

JF - Computer-Assisted Radiology and Surgery

SN - 1861-6410

ER -