Development of numerical models for injury biomechanics research: a review of 50 years of publications in the Stapp Car Crash Conference

Subject-Specific Geometry of FE Lumbar Spine Models for the Replication of Fracture Locations Using Dynamic Drop Tests

For traumatic lumbar spine injuries, the mechanisms and influence of anthropometrical variation are not yet fully understood under dynamic loading. Our objective was to evaluate whether geometrically subject-specific explicit finite element (FE) lumbar spine models based on state-of-the-art clinical CT data combined with general material properties from the literature could replicate the experimental responses and the fracture locations via a dynamic drop tower-test setup. The experimental CT datasets from a dynamic drop tower-test setup were used to create anatomical details of four lumbar spine models (T12 to L5). The soft tissues from THUMS v4.1 were integrated by morphing. Each model was simulated with the corresponding loading and boundary conditions from the dynamic lumbar spine tests that produced differing injuries and injury locations. The simulations resulted in force, moment, and kinematic responses that effectively matched the experimental data. The pressure distribution within the models was used to compare the fracture occurrence and location. The spinal levels that sustained vertebral body fracture in the experiment showed higher simulation pressure values in the anterior elements than those in the levels that did not fracture in the reference experiments. Similarly, the spinal levels that sustained posterior element fracture in the experiments showed higher simulation pressure values in the vertebral posterior structures compared to those in the levels that did not sustain fracture. Our study showed that the incorporation of the spinal geometry and orientation could be used to replicate the fracture type and location under dynamic loading. Our results provided an understanding of the lumbar injury mechanisms and knowledge on the load thresholds that could be used for injury prediction with explicit FE lumbar spine models.

Thoughts and perspectives on biomechanical numerical models under impacts: Are women forgotten from research?

The present paper explores a series of articles in the literature which deal with impact biomechanics of the head and thorax/abdomen segments, investigating the "sex specific properties/data" used in the studies. Statements in these studies are analyzed and point out, the use of male or female subjects for the developments of finite element models and their validation against experimental data. The present analysis raises the question about "androcentrism," and how biomechanical engineering findings and the design of the derived protecting devices are focused on male subjects.

Approximating subject-specific brain injury models via scaling based on head-brain morphological relationships

Most human head/brain models represent a generic adult male head/brain. They may suffer in accuracy when investigating traumatic brain injury (TBI) on a subject-specific basis. Subject-specific models can be developed from neuroimages; however, neuroimages are not typically available in practice. In this study, we establish simple and elegant regression models between brain outer surface morphology and head dimensions measured from neuroimages along with age and sex information (N = 191; 141 males and 50 females with age ranging 14-25 years). The regression models are then used to approximate subject-specific brain models by scaling a generic counterpart, without using neuroimages. Model geometrical accuracy is assessed using adjusted [Formula: see text] and absolute percentage error (e.g., 0.720 and 3.09 ± 2.38%, respectively, for brain volume when incorporating tragion-to-top). For a subset of 11 subjects (from smallest to largest in brain volume), impact-induced brain strains are compared with those from "morphed models" derived from neuroimage-based mesh warping. We find that regional peak strains from the scaled subject-specific models are comparable to those of the morphed counterparts but could be considerably different from those of the generic model (e.g., linear regression slope of 1.01-1.03 for gray and white matter regions versus 1.16-1.19, or up to ~ 20% overestimation for the smallest brain studied). These results highlight the importance of incorporating brain morphological variations in impact simulation and demonstrate the feasibility of approximating subject-specific brain models without neuroimages using age, sex, and easily measurable head dimensions. The scaled models may improve subject specificity for future TBI investigations.

Displacement voxelization to resolve mesh-image mismatch: Application in deriving dense white matter fiber strains

BACKGROUND AND OBJECTIVE

It is common to combine biomechanical modeling and medical images for multimodal analyses. However, mesh-image mismatch may occur that prevents direct information exchange. To eliminate mesh-image mismatch, we develop a simple but elegant displacement voxelization technique based on image voxel corner nodes to achieve voxel-wise strain. We then apply the technique to derive dense white matter fiber strains along whole-brain tractography (∼35 k fiber tracts consisting of ∼3.3 million sampling points) resulting from head impact.

METHODS

Displacements at image voxel corner nodes are first obtained from model simulation via scattered interpolation. Each voxel is then scaled linearly to form a unit hexahedral element. This allows convenient and efficient voxel-wise strain tensor calculation and displacement interpolation at arbitrary fiber sampling points via shape functions. Fiber strains from displacement interpolation are then compared with those from the commonly used strain tensor projection using either voxel- or element-wise strain tensors.

RESULTS

Based on a synthetic displacement field, fiber strains interpolated from voxelized displacement are considerably more accurate than those from strain tensor projection relative to the prescribed ground-truth (determinant of coefficient (R²) of 1.00 and root mean squared error (RMSE) of 0.01 vs. 0.87 and 0.10, respectively). For a set of real-world reconstructed head impacts (N = 53), the strain tensor projection method performs similarly poorly (R² of 0.80-0.90 and RMSE of 0.03-0.07), with overestimation strongly correlated with strain magnitude (Pearson correlation coefficient >0.9). Up to ∼15% of the fiber strains are overestimated by more than the lower bound of a conservative injury threshold of 0.09. The percentage increases to ∼37% when halving the threshold. Voxel interpolation is also significantly more efficient (15 s vs. 40 s for element strain tensor projection, without parallelization).

CONCLUSIONS

Voxelized displacement interpolation is considerably more accurate and efficient in deriving dense white matter fiber strains than strain tensor projection. The latter generally overestimates with overestimation magnitude strongly correlating with fiber strain magnitude. Displacement voxelization is an effective technique to eliminate mesh-image mismatch and generates a convenient image representation of tissue deformation. This technique can be generalized to broadly facilitate a diverse range of image-related biomechanical problems for multimodal analyses. The convenient image format may also promote and facilitate biomechanical data sharing in the future.

Instantaneous Brain Strain Estimation for Automotive Head Impacts via Deep Learning

Efficient brain strain estimation is critical for routine application of a head injury model. Lately, a convolutional neural network (CNN) has been successfully developed to estimate spatially detailed brain strains instantly and accurately in contact sports. Here, we extend its application to automotive head impacts, where impact profiles are typically more complex with longer durations. Head impact kinematics (N=458) from two public databases were used to generate augmented impacts (N=2694). They were simulated using the anisotropic Worcester Head Injury Model (WHIM) V1.0, which provided baseline elementwise peak maximum principal strain (MPS). For each augmented impact, rotational velocity (v_rot) and the corresponding rotational acceleration (a_rot) profiles were concatenated as static images to serve as CNN input. Three training strategies were evaluated: 1) "baseline", using random initial weights; 2) "transfer learning", using weight transfer from a previous CNN model trained on head impacts drawn from contact sports; and 3) "combined training", combining previous training data from contact sports (N=5661) for training. The combined training achieved the best performances. For peak MPS, the CNN achieved a coefficient of determination (R²) of 0.932 and root mean squared error (RMSE) of 0.031 for the real-world testing dataset. It also achieved a success rate of 60.5% and 94.8% for elementwise MPS, where the linear regression slope, k, and correlation coefficient, r, between estimated and simulated MPS did not deviate from 1.0 (when identical) by more than 0.1 and 0.2, respectively. Cumulative strain damage measure (CSDM) from the CNN estimation was also highly accurate compared to those from direct simulation across a range of thresholds (R² of 0.899-0.943 with RMSE of 0.054-0.069). Finally, the CNN achieved an average k and r of 0.98±0.12 and 0.90±0.07, respectively, for six reconstructed car crash impacts drawn from two other sources independent of the training dataset. Importantly, the CNN is able to efficiently estimate elementwise MPS with sufficient accuracy while conventional kinematic injury metrics cannot. Therefore, the CNN has the potential to supersede current kinematic injury metrics that can only approximate a global peak MPS or CSDM. The CNN technique developed here may offer enhanced utility in the design and development of head protective countermeasures, including in the automotive industry. This is the first study aimed at instantly estimating spatially detailed brain strains for automotive head impacts, which employs >8.8 thousand impact simulations generated from ~1.5 years of nonstop computations on a high-performance computing platform.

Analysis of Force Mitigation by Boots in Axial Impacts using a Lower Leg Finite Element Model

Lower extremity injuries caused by floor plate impacts through the axis of the lower leg are a major source of injury and disability for civilian and military vehicle occupants. A collection of PMHS pendulum impacts was revisited to obtain data for paired booted/unbooted test on the same leg. Five sets of paired pendulum impacts (10 experiments in total) were found using four lower legs from two PMHS. The PMHS size and age was representative of an average young adult male. In these tests, a PMHS leg was impacted by a 3.4 or 5.8 kg pendulum with an initial velocity of 5, 7, or 10 m/s (42-288 J). A matching LS-DYNA finite element model was developed to replicate the experiments and provide additional energy, strain, and stress data. Simulation results matched the PMHS data using peak values and CORA curve correlations. Experimental forces ranged between 1.9 and 12.1 kN experimentally and 2.0 and 11.7 kN in simulation. Combat boot usage reduced the peak force by 36% experimentally (32% in simulation) by compressing the sole and insole with similar mitigations for calcaneus strain. The simulated Von Mises stress contours showed the boot both mitigating and shifting stress concentrations from the calcaneus in unbooted impacts to the talus-tibia joint in the booted impacts, which may explain why some previous studies have observed shifts to tibia injuries with boot or padding usage.

Development and Full Body Validation of a 5th Percentile Female Finite Element Model

To mitigate the societal impact of vehicle crash, researchers are using a variety of tools, including finite element models (FEMs). As part of the Global Human Body Models Consortium (GHBMC) project, comprehensive medical image and anthropometrical data of the 5th percentile female (F05) were acquired for the explicit purpose of FEM development. The F05-O (occupant) FEM model consists of 981 parts, 2.6 million elements, 1.4 million nodes, and has a mass of 51.1 kg. The model was compared to experimental data in 10 validation cases ranging from localized rigid hub impacts to full body sled cases. In order to make direct comparisons to experimental data, which represent the mass of an average male, the model was compared to experimental corridors using two methods: 1) post-hoc scaling the outputs from the baseline F05-O model and 2) geometrically morphing the model to the body habitus of the average male to allow direct comparisons. This second step required running the morphed full body model in all 10 simulations for a total of 20 full body simulations presented. Overall, geometrically morphing the model was found to more closely match the target data with an average ISO score for the rigid impacts of 0.76 compared to 0.67 for the scaled responses. Based on these data, the morphed model was then used for model validation in the vehicle sled cases. Overall, the morphed model attained an average weighted score of 0.69 for the two sled impacts. Hard tissue injuries were also assessed and the baseline F05-O model was found to predict a greater occurrence of pelvic fractures compared to the GHBMC average male model, but predicted fewer rib fractures.

An Objective Evaluation of Mass Scaling Techniques Utilizing Computational Human Body Finite Element Models

Biofidelity response corridors developed from post-mortem human subjects are commonly used in the design and validation of anthropomorphic test devices and computational human body models (HBMs). Typically, corridors are derived from a diverse pool of biomechanical data and later normalized to a target body habitus. The objective of this study was to use morphed computational HBMs to compare the ability of various scaling techniques to scale response data from a reference to a target anthropometry. HBMs are ideally suited for this type of study since they uphold the assumptions of equal density and modulus that are implicit in scaling method development. In total, six scaling procedures were evaluated, four from the literature (equal-stress equal-velocity, ESEV, and three variations of impulse momentum) and two which are introduced in the paper (ESEV using a ratio of effective masses, ESEV-EffMass, and a kinetic energy approach). In total, 24 simulations were performed, representing both pendulum and full body impacts for three representative HBMs. These simulations were quantitatively compared using the International Organization for Standardization (ISO) ISO-TS18571 standard. Based on these results, ESEV-EffMass achieved the highest overall similarity score (indicating that it is most proficient at scaling a reference response to a target). Additionally, ESEV was found to perform poorly for two degree-of-freedom (DOF) systems. However, the results also indicated that no single technique was clearly the most appropriate for all scenarios.

Application of Radial Basis Function Methods in the Development of a 95th Percentile Male Seated FEA Model

Human body finite element models (FEMs) are a valuable tool in the study of injury biomechanics. However, the traditional model development process can be time-consuming. Scaling and morphing an existing FEM is an attractive alternative for generating morphologically distinct models for further study. The objective of this work is to use a radial basis function to morph the Global Human Body Models Consortium (GHBMC) average male model (M50) to the body habitus of a 95th percentile male (M95) and to perform validation tests on the resulting model. The GHBMC M50 model (v. 4.3) was created using anthropometric and imaging data from a living subject representing a 50th percentile male. A similar dataset was collected from a 95th percentile male (22,067 total images) and was used in the morphing process. Homologous landmarks on the reference (M50) and target (M95) geometries, with the existing FE node locations (M50 model), were inputs to the morphing algorithm. The radial basis function was applied to morph the FE model. The model represented a mass of 103.3 kg and contained 2.2 million elements with 1.3 million nodes. Simulations of the M95 in seven loading scenarios were presented ranging from a chest pendulum impact to a lateral sled test. The morphed model matched anthropometric data to within a rootmean square difference of 4.4% while maintaining element quality commensurate to the M50 model and matching other anatomical ranges and targets. The simulation validation data matched experimental data well in most cases.