Combining ultrasonic and computed tomography scanning to characterize mechanical properties of cancellous bone in necrotic human femoral heads

https://doi.org/10.1016/j.medengphy.2019.02.002Get rights and content

Highlights

  • This experiment is carried out based on patients with femoral head necrosis.

  • The accuracy of the experiment is at millimeter level.

  • This study mainly determined the relationship between density and elastic modulus for the entire femoral head.

  • The material properties of necrotic femoral heads and normal femoral heads are measured in this experiment using ultrasound scanning combined with CT scanning.

Abstract

Image-based finite element modelling has been commonly used to determine the biomechanical behaviours of human femora, particularly for the diagnosis of femoral head necrosis. One of the fundamental aspects of biomechanical modelling is the relationship between bone density, which is obtained from images, and elastic modulus. While there exist some empirical equations relating density with elastic modulus, the characterization of this relationship remains incomplete, especially for necrotic femoral heads. The objective of this study was to determine the relationship between density and elastic modulus by combining ultrasonic scanning and computed tomography (CT). Bone specimens were surgically removed from the femora of eight persons (seven females and one male in the age range of 55–68 years old) and underwent both ultrasonic and CT scanning. The images were processed with MATLAB scripts, and a bilinear interpolation algorithm was used to determine the relationship between the CT-measured densities and ultrasound-measured elastic moduli. The results showed different density–elastic modulus relations between the hardening strap of the necrotic region and non-hardening strap areas of the necrotic region. The uniqueness of this study is the characterization of mechanical properties (in the present study, the density–modulus relationship) from clinical images, which would be valuable in computational biomechanics used for the diagnosis and treatment evaluation of femoral head necrosis.

Introduction

Finite element (FE) analysis has been widely used to study the biomechanics of human femora [1], [2], [3], [4], especially for assessing femoral implantation and treatment of osteonecrosis of the femoral head (ONFH). Typically, FE models require the assignment of bone material properties (e.g., elastic modulus) that derive from the bone mineral density, which can be determined by computed tomography (CT) scanning. One of the most important factors regarding the accuracy of an FE analysis is the utilization of a proper density–elastic modulus formula for the FE model.

The mechanical properties (e.g., elastic modulus) and their relationship with femoral bone density can be obtained through in vitro measurements from cadavers. Traditionally, for instance, compressive tests can be performed on cubic or cylindrical specimens to measure the apparent-level elastic modulus and/or bone strength, which can be related to the measured apparent densities of the specimens [5,6]. Based on those established empirical apparent-level density–modulus relationships, some studies have employed strain gauge measures and mechanical tests on whole bones (such as femora) in combination with CT scanning and FE modelling to validate and determine the most appropriate density–modulus relationships for subject-specific FE models [7], [8], [9], [10], [11], [12]. In addition to mechanical testing and computational modelling, ultrasonic techniques have been used to scan cubic or cylindrical femoral samples to directly measure their elastic moduli [13], [14], [15], [16], [17], [18], [19]. Moreover, ultrasonic measures, being non-invasive, may offer some advantages over mechanical tests [20]. A few studies have combined CT and ultrasonic scanning to determine the density–modulus relationship for cortical and cancellous bone [20,21].

It should be noted that previous studies only measured the average/apparent density and ultrasound-based modulus for an entire specimen, based on which the density–modulus relations were established [21]. However, since the material property assignment in finite element modelling requires a density–modulus relationship for pixel-sized elements, those established relations might have ignored finer details of bone material properties. To the best of the authors’ knowledge, no study has used a combination of ultrasound and CT to measure the material properties of necrotic bones, and no density–modulus relationships have been established for necrotic bone tissues in the literature. This is an important issue because the material property assignment is a fundamental step for finite element modelling of necrotic femora for the diagnosis of femoral head necrosis or for the prediction of femoral head collapse. Since it would be difficult to extract a small necrotic bone sample from the femur and conduct mechanical tests on it, noninvasive ultrasonic scanning of a whole femur would be a better way to measure the material properties for necrotic bone tissues.

Therefore, the current study mainly aimed to determine the relationship between density and elastic modulus for the femoral head from normal subjects and patients with femoral head necrosis by using the ultrasonic scanning technique together with CT imaging. The combination of ultrasonic and CT scanning would enable the determination of the density–elastic modulus relationship at a relatively small scale (e.g., pixel/element size), which could be more feasible for CT-based finite element modelling.

Section snippets

Materials and methods

All bone specimens were derived from eight patients, which were provided by Wangjing Hospital (seven females and one male in the age range of 55–68 years old), including seven specimens with femoral head necrosis (one male and six females) and one normal specimen (female). The sample collection procedure was approved by the Ethics Committee of the China Academy of Chinese Medical Sciences in Wangjing Hospital. The normal specimen was collected from a patient who underwent femoral head

Results

Figs. 3 and 4 show the relationships of lattice information between the density and corresponding elastic modulus of each pixel for both normal and necrotic femoral heads. The relationships between the density and elastic modulus of cancellous bone of the normal and necrotic femoral heads corresponded to different fitted curve equations. In addition, the distribution of mechanical properties (in the present study, the density–elastic modulus relationship) was different for various samples. In

Discussion

The present study investigated the relationship between density and elastic modulus by combining ultrasonic and CT scanning. The formula between density and elastic modulus determined in this study is of great importance in computational modelling to study the biomechanical behaviours of femoral heads. These experimental results indicated that, for patients with osteonecrosis of the femoral head, the relationships between density and elastic modulus were different among three typical regions,

Conclusions

In summary, this paper has determined the density–modulus relationships of femoral cancellous bone from eight human subjects using an ultrasonic scanning technique. The density–elastic modulus relationships for femoral bone tissues of normal patients and patients with osteonecrosis of the femoral head were determined. In addition, experiments have been carried out to categorize the necrotic and normal regions from patients with ONFH to obtain their respective relationship formulas between

Acknowledgments

This study was supported by National Key R&D Program of China (2017YFC0107901), National Natural Science Foundation of China (81770465, 11702008), Support Plan for High-level Faculties in Beijing Municipal Universities (CIT&TCD201804011) and Beijing Excellent Talents Funds (2017000020124G277).

Conflict of interest

There is no conflict of interests in this study.

Ethical approval

WJEC-KT-2016-007-P002

China Academy of Chinese Medical Sciences Wangjing Hospital.

References (34)

1

Both authors contributed equally to this work and should be considered co-first authors.

View full text