Original Article
Evaluation of vertebral bone strength with a finite element method using low dose computed tomography imaging

https://doi.org/10.1016/j.jos.2021.03.004Get rights and content

Abstract

Background

Focusing on compression fractures of bone by finite elements, we evaluated bone strength based on the computed tomography-based finite element method. However, the exposure dose is an issue. We aimed to investigate the quantity of reduction of the radiation dose with respect to the reference dose by comparing the calculation results of compression fractures of the vertebral body using experimental data obtained from the spine of a pig.

Methods

Computed tomography images of a self-made phantom that enclosed the lower lumbar vertebra of edible wild pigs were obtained under baseline-dose conditions using various lower tube currents. Images obtained under reference-dose conditions were reconstructed using the filtered back-projection method, whereas images obtained under low-dose conditions were reconstructed using both the filtered back-projection method and the iterative reconstruction method. Computer simulations involving the creation of finite element models using all images were implemented for the compression load calculation for vertebral body parts. Based on the calculated results, images of the low-dose and reference-dose conditions were compared.

Results

Using pigs’ lower lumbar vertebrae, finite element model analysis of low-dose X-ray computed tomography images showed that equivalent results can be obtained with a dose of approximately 40% of the standard radiographic reference doses. As for the compression stress intensity, the same results as those under reference-dose conditions were obtained using the iterative reconstruction method in combination with computed tomography-based finite element method.

Conclusions

The combination of the iterative reconstruction method with the computed tomography-based finite element method is an effective image reconstruction method for achieving dose reduction.

Introduction

The contributions of bone mineral density and bone quality to the risk of osteoporotic factors are 70% and 30%, respectively [1]. Dual energy X-ray absorptiometry (DXA) is considered the gold standard for bone mineral density measurement [2]. However, DXA has a limited ability in assessing bone mineral density alone and determining the fracture risk [3]. When calculating bone mineral density using DXA, the effects of calcification and degenerative spondylosis cause errors in the assessment results. Additionally, this procedure does not assess bone quality. Besides DXA, quantitative computed tomography (Q-CT) is also a method used for determining bone mineral density, which is reported to be more accurate than DXA [4].

To solve the aforementioned problems, we will evaluate bone strength based on the computed tomography-based finite element method (CT-FEM). CT-FEM uses digital imaging and medical communication (DICOM) data to create a three-dimensional finite element model to simulate deformation or destruction according to bone structure and external features in relation to external forces. It is possible to specify size and direction as the load conditions on the bone and build a model for analyzing each subject's bones [5]. This procedure allows for the analysis of bone mineral density and quality (bone outline and internal structure).

One disadvantage of CT-FEM is more exposure to radiation than that of DXA. However, exposure can be reduced using the successive approximation image reconstruction method (IR method: iterative reconstruction method). The IR method can capture images using lower doses of radiation than traditional filtered back projection (FBP) methods, without compromising sharpness or contrast [6].

Our purpose was to investigate the quantity of reduction of the radiation dose with respect to the reference dose by comparing the calculation results of compression fractures of the vertebral body using experimental data. The IR method can reduce radiation levels by 90% [7]. We focused on reducing the tube current, which affects the exposure dose, after confirming the accuracy of CT-FEM, where the site destroyed by the load matches the density image. CT-FEM was performed using CT images taken at lower than standard doses. Since image noise can be suppressed by reconstructing the image using the IR method, the usefulness of the IR method was examined.

Section snippets

Research target

The FEM model used in this study was based on the CT data of a self-made phantom created using the lower lumbar spine of an edible wild pig with a human-like structure. Each phantom was filled with three lumbar vertebrae taken from one pig to create a phantom mixed with saline (0.9%) and a small amount of preservative (sorbic acid), for a total of six phantoms (Fig. 1a). The diameter of the container was 20 cm. The arrangement seen from the side of the lumbar spine is shown in Fig. 1b. The CT

Density

Fig. 2 shows the density display image of the reference tube current of 300 mA (IR level 0), and Fig. 3 shows the obtained density display images of 240, 200, 180 A, 120, and 100 mA (IR level 0). The displayed cross section is a median sagittal section image in the calculation model. Additionally, due to the large number of vertebral bodies, only one vertebral body is shown.

In the density images under all conditions, high density was shown in the part corresponding to the dense bone around the

Density imaging comparison

A high-density area was confirmed near the surface corresponding to the cortical bone, and a low-density region was confirmed inside the vertebral body bone corresponding to the cancellous bone. The density display image can be calculated from the CT value using a linear function-type conversion formula. Therefore, a calculation model that accurately reflects the distribution of pixel values on the CT image can be presumed.

Effect of the IR method

Despite the change in the IR, the average CT value remains almost

Limitations

This study has several limitations. First, quadrupedal pigs and bipedal humans differ in the direction of load on the spine. The humans load is from the posterior to anterior. Second, the basic anatomy is similar, but the vertebral body of the pig is axially longer than that of humans. Further, judging from the CT value, the bone density of pigs is estimated to be approximately twice that of humans. If there is only one lumbar spine, it is considered that there is no problem in examining the

Conclusions

CT images taken under low-tube current conditions can be reconstructed using the IR method to obtain results equivalent to the standard-dose conditions. By using successive approximation image reconstruction, the same result as that of the reference dose image was obtained, even under the condition of a tube current of 120 mA.

Compliance with ethical standards

This study was conducted on knees from pigs to be used for consumption, which were to be discarded. There were no ethical considerations applicable because the pigs were not meant to be used for research purposes.

Declaration of competing interest

None.

Acknowledgement

We would like to thank Editage (www.editage.com) for English language editing.

References (21)

  • Y. Kawabata et al.

    The risk assessment of pathological fracture in the proximal femur using a CT-based finite element method

    J Orthop Sci

    (2017 Sep 1)
  • J.H. Keyak et al.

    Male-female differences in the association between incident hip fracture and proximal femoral strength: a finite element analysis study

    Bone

    (2011 Jun 1)
  • Nih Consensus Development Panel on Osteoporosis Prevention

    Diagnosis, and Therapy. Osteoporosis prevention, diagnosis, and therapy

    J Am Med Assoc

    (2001 Feb)
  • W.F. Lems et al.

    Critical issues and current challenges in osteoporosis and fracture prevention. An overview of unmet needs

    Ther Adv Musculoskelet Dis

    (2017 Dec 1)
  • M. Ito et al.

    Multi-detector row CT imaging of vertebral microstructure for evaluation of fracture risk

    J Bone Miner Res

    (2005 Oct)
  • Maximilian T. Löffler et al.

    Automatic opportunistic osteoporosis screening in routine CT: improved prediction of patients with prevalent vertebral fractures compared to DXA

    Eur Radiol

    (2021 Jan 28)
  • K. Watanabe et al.

    Development of a knee joint CT-FEM model in load response of the stance phase during walking using muscle exertion, motion analysis, and ground reaction force data

    Medicina

    (2020 Jan 29)
  • A. Löve et al.

    Six iterative reconstruction algorithms in brain CT: a phantom study on image quality at different radiation dose levels

    Br J Radiol

    (2013 Nov)
  • N.M. Rayudu et al.

    Finite element analysis-based vertebral bone strength prediction using MDCT data: how low can we go?

    Front Endocrinol

    (2020 Jul 28)
  • K. Imai et al.

    In vivo assessment of lumbar vertebral strength in elderly women using computed tomography-based nonlinear finite element model

    Spine

    (2008 Jan 1)
There are more references available in the full text version of this article.
View full text