Variabilities in µQCT-based FEA of a tumoral bone mice model

Inter-laboratory reproduction and sensitivity study of a finite element model to quantify human femur failure load: Case of metastases

INTRODUCTION

Metastases increase the risk of fracture when affecting the femur. Consequently, clinicians need to know if the patient's femur can withstand the stress of daily activities. The current tools used in clinics are not sufficiently precise. A new method, the CT-scan-based finite element analysis, gives good predictive results. However, none of the existing models were tested for reproducibility. This is a critical issue to address in order to apply the technique on a large cohort around the world to help evaluate bone metastatic fracture risk in patients. The aim of this study is then to evaluate 1) the reproducibility 2) the transposition of the reproduced model to another dataset and 3) the global sensitivity of one of the most promising models of the literature (original model).

METHODS

The model was reproduced based on the paper describing it and discussion with authors to avoid reproduction errors. The reproducibility was evaluated by comparing the results given in the original model by the original first team (Leuven, Belgium) and the reproduced model made by another team (Lyon, France) on the same dataset of CT-scans of ex vivo femurs. The transposition of the model was evaluated by comparing the results of the reproduced model on two different datasets. The global sensitivity analysis was done by using the Morris method and evaluates the influence of the density calibration coefficient, the segmentation, the orientations and the length of the femur.

RESULTS

The original and reproduced models are highly correlated (r2 = 0.95), even though the reproduced model gives systematically higher failure loads. When using the reproduced model on another dataset, predictions are less accurate (r2 with the experimental failure load decreases, errors increase). The global sensitivity analysis showed high influence of the density calibration coefficient (mean variation of failure load of 84 %) and non-negligible influence of the segmentation, orientation and length of the femur (mean variation of failure load between 7 and 10 %).

CONCLUSION

This study showed that, although being validated, the reproduced model underperformed when using another dataset. The difference in performance depending on the dataset is commonly the cause of overfitting when creating the model. However, the dataset used in the original paper (Sas et al., 2020a) and the Leuven's dataset gave similar performance, which indicates a lesser probability for the overfitting cause. Also, the model is highly sensitive to density parameters and automation of measurement may minimize the uncertainty on failure load. An uncertainty propagation analysis would give the actual precision of such model and improve our understanding of its behavior and is part of future work.

Reproducibility of Densitometric and Biomechanical Assessment of the Mouse Tibia From In Vivo Micro-CT Images

Interventions for bone diseases (e.g. osteoporosis) require testing in animal models before clinical translation and the mouse tibia is among the most common tested anatomical sites. In vivo micro-Computed Tomography (microCT) based measurements of the geometrical and densitometric properties are non-invasive and therefore constitute an important tool in preclinical studies. Moreover, validated micro-Finite Element (microFE) models can be used for predicting the bone mechanical properties non-invasively. However, considering that the image processing pipeline requires operator-dependant steps, the reproducibility of these measurements has to be assessed. The aim of this study was to evaluate the intra- and inter-operator reproducibility of several bone parameters measured from microCT images. Ten in vivo microCT images of the right tibia of five mice (at 18 and 22 weeks of age) were processed. One experienced operator (intra-operator analysis) and three different operators (inter-operator) aligned each image to a reference through a rigid registration and selected a volume of interest below the growth plate. From each image the following parameters were measured: total bone mineral content (BMC) and density (BMD), BMC in 40 subregions (ten longitudinal sections, four quadrants), microFE-based stiffness and failure load. Intra-operator reproducibility was acceptable for all parameters (precision error, PE < 3.71%), with lowest reproducibility for stiffness (3.06% at week 18, 3.71% at week 22). The inter-operator reproducibility was slightly lower (PE < 4.25%), although still acceptable for assessing the properties of most interventions. The lowest reproducibility was found for BMC in the lateral sector at the midshaft (PE = 4.25%). Densitometric parameters were more reproducible than most standard morphometric parameters calculated in the proximal trabecular bone. In conclusion, microCT and microFE models provide reproducible measurements for non-invasive assessment of the mouse tibia properties.

Beam theory for rapid strain estimation in the mouse tibia compression model

The mouse tibia compression model is a leading model for studying bone's mechanoadaptive response to load. In studying this mechanoadaptive response, (FE) modelling is often used to determine the stress/strain within the tibia. The development of such models can be challenging and computationally expensive. An alternate approach is to use continuum mechanics based analytical theories, such as beam theory (BT). However, applying BT to the mouse tibia requires the fibula be neglected, introducing error in the stress/strain distribution. While several studies have applied BT to the mouse tibia, no study has explored the accuracy of this approach. To address these questions, this work investigates the use of BT in determining stress/strain within the mouse tibia. By comparing BT against FE modelling, it was found that BT can accurately predict tibial stress/strain if correction factors are applied to account for the effect of the fibula. The 25, 37, 50 and 75% cross sections are studied. Focusing on the 37% cross section, without correction, BT can have errors of approximately 21.6%. With correction, this is reduced to 6.6%. Such correction factors are presented. The developed BT model is applicable in the diaphysis and distal metaphysis, where the assumptions of BT are valid. This work verifies BT for determining localised strains in a mouse tibia compression model. This is anticipated to provide efficiency dividends, allowing for high throughput modelling of the mouse tibia, advancing study of bone's mechanoadaptive response.