Patient-specific biomechanical modeling of bone strength using statistically-derived fabric tensors

Karim Lekadir, Christopher Noble, Javad Hazrati Marangalou, Corné Hoogendoorn, Bert van Rietbergen, Zeike A. Taylor, Alejandro F. Frangi

Research output: Contribution to journalArticleAcademicpeer-review

8 Citations (Scopus)

Abstract

Low trauma fractures are amongst the most frequently encountered problems in the clinical assessment and treatment of bones, with dramatic health consequences for individuals and high financial costs for health systems. Consequently, significant research efforts have been dedicated to the development of accurate computational models of bone biomechanics and strength. However, the estimation of the fabric tensors, which describe the microarchitecture of the bone, has proven to be challenging using in vivo imaging. On the other hand, existing research has shown that isotropic models do not produce accurate predictions of stress states within the bone, as the material properties of the trabecular bone are anisotropic. In this paper, we present the first biomechanical study that uses statistically-derived fabric tensors for the estimation of bone strength in order to obtain patient-specific results. We integrate a statistical predictive model of trabecular bone microarchitecture previously constructed from a sample of ex vivo micro-CT datasets within a biomechanical simulation workflow. We assess the accuracy and flexibility of the statistical approach by estimating fracture load for two different databases and bone sites, i.e., for the femur and the T12 vertebra. The results obtained demonstrate good agreement between the statistically-driven and micro-CT-based estimates, with concordance coefficients of 98.6 and 95.5% for the femur and vertebra datasets, respectively.
Original languageEnglish
Pages (from-to)234-246
JournalAnnals of biomedical engineering
Volume44
Issue number1
DOIs
Publication statusPublished - 2016

Fingerprint

Tensors
Bone
Health
Biomechanics
Materials properties
Imaging techniques
Costs

Keywords

  • METIS-319018
  • IR-102198

Cite this

Lekadir, Karim ; Noble, Christopher ; Hazrati Marangalou, Javad ; Hoogendoorn, Corné ; van Rietbergen, Bert ; Taylor, Zeike A. ; Frangi, Alejandro F. / Patient-specific biomechanical modeling of bone strength using statistically-derived fabric tensors. In: Annals of biomedical engineering. 2016 ; Vol. 44, No. 1. pp. 234-246.
@article{12ebb746b638492aae2b85bf7e73a144,
title = "Patient-specific biomechanical modeling of bone strength using statistically-derived fabric tensors",
abstract = "Low trauma fractures are amongst the most frequently encountered problems in the clinical assessment and treatment of bones, with dramatic health consequences for individuals and high financial costs for health systems. Consequently, significant research efforts have been dedicated to the development of accurate computational models of bone biomechanics and strength. However, the estimation of the fabric tensors, which describe the microarchitecture of the bone, has proven to be challenging using in vivo imaging. On the other hand, existing research has shown that isotropic models do not produce accurate predictions of stress states within the bone, as the material properties of the trabecular bone are anisotropic. In this paper, we present the first biomechanical study that uses statistically-derived fabric tensors for the estimation of bone strength in order to obtain patient-specific results. We integrate a statistical predictive model of trabecular bone microarchitecture previously constructed from a sample of ex vivo micro-CT datasets within a biomechanical simulation workflow. We assess the accuracy and flexibility of the statistical approach by estimating fracture load for two different databases and bone sites, i.e., for the femur and the T12 vertebra. The results obtained demonstrate good agreement between the statistically-driven and micro-CT-based estimates, with concordance coefficients of 98.6 and 95.5{\%} for the femur and vertebra datasets, respectively.",
keywords = "METIS-319018, IR-102198",
author = "Karim Lekadir and Christopher Noble and {Hazrati Marangalou}, Javad and Corn{\'e} Hoogendoorn and {van Rietbergen}, Bert and Taylor, {Zeike A.} and Frangi, {Alejandro F.}",
year = "2016",
doi = "10.1007/s10439-015-1432-2",
language = "English",
volume = "44",
pages = "234--246",
journal = "Annals of biomedical engineering",
issn = "0090-6964",
publisher = "Springer",
number = "1",

}

Patient-specific biomechanical modeling of bone strength using statistically-derived fabric tensors. / Lekadir, Karim; Noble, Christopher; Hazrati Marangalou, Javad; Hoogendoorn, Corné; van Rietbergen, Bert; Taylor, Zeike A.; Frangi, Alejandro F.

In: Annals of biomedical engineering, Vol. 44, No. 1, 2016, p. 234-246.

Research output: Contribution to journalArticleAcademicpeer-review

TY - JOUR

T1 - Patient-specific biomechanical modeling of bone strength using statistically-derived fabric tensors

AU - Lekadir, Karim

AU - Noble, Christopher

AU - Hazrati Marangalou, Javad

AU - Hoogendoorn, Corné

AU - van Rietbergen, Bert

AU - Taylor, Zeike A.

AU - Frangi, Alejandro F.

PY - 2016

Y1 - 2016

N2 - Low trauma fractures are amongst the most frequently encountered problems in the clinical assessment and treatment of bones, with dramatic health consequences for individuals and high financial costs for health systems. Consequently, significant research efforts have been dedicated to the development of accurate computational models of bone biomechanics and strength. However, the estimation of the fabric tensors, which describe the microarchitecture of the bone, has proven to be challenging using in vivo imaging. On the other hand, existing research has shown that isotropic models do not produce accurate predictions of stress states within the bone, as the material properties of the trabecular bone are anisotropic. In this paper, we present the first biomechanical study that uses statistically-derived fabric tensors for the estimation of bone strength in order to obtain patient-specific results. We integrate a statistical predictive model of trabecular bone microarchitecture previously constructed from a sample of ex vivo micro-CT datasets within a biomechanical simulation workflow. We assess the accuracy and flexibility of the statistical approach by estimating fracture load for two different databases and bone sites, i.e., for the femur and the T12 vertebra. The results obtained demonstrate good agreement between the statistically-driven and micro-CT-based estimates, with concordance coefficients of 98.6 and 95.5% for the femur and vertebra datasets, respectively.

AB - Low trauma fractures are amongst the most frequently encountered problems in the clinical assessment and treatment of bones, with dramatic health consequences for individuals and high financial costs for health systems. Consequently, significant research efforts have been dedicated to the development of accurate computational models of bone biomechanics and strength. However, the estimation of the fabric tensors, which describe the microarchitecture of the bone, has proven to be challenging using in vivo imaging. On the other hand, existing research has shown that isotropic models do not produce accurate predictions of stress states within the bone, as the material properties of the trabecular bone are anisotropic. In this paper, we present the first biomechanical study that uses statistically-derived fabric tensors for the estimation of bone strength in order to obtain patient-specific results. We integrate a statistical predictive model of trabecular bone microarchitecture previously constructed from a sample of ex vivo micro-CT datasets within a biomechanical simulation workflow. We assess the accuracy and flexibility of the statistical approach by estimating fracture load for two different databases and bone sites, i.e., for the femur and the T12 vertebra. The results obtained demonstrate good agreement between the statistically-driven and micro-CT-based estimates, with concordance coefficients of 98.6 and 95.5% for the femur and vertebra datasets, respectively.

KW - METIS-319018

KW - IR-102198

U2 - 10.1007/s10439-015-1432-2

DO - 10.1007/s10439-015-1432-2

M3 - Article

VL - 44

SP - 234

EP - 246

JO - Annals of biomedical engineering

JF - Annals of biomedical engineering

SN - 0090-6964

IS - 1

ER -