Abstract

The aims of this investigation were to delineate the internal biomechanics of the spine under vertical impact vector and assess the probability of injury. Male and female whole-body human finite element models were used. The restrained occupants were positioned on the seat, and caudo-cephalad impacts were applied to the base. Different acceleration-time profiles (50–200 ms pulse durations, 11–46 g peak accelerations) were used as inputs in both models. The resulting stress–strain profiles in the cortical and cancellous bones were evaluated at different vertebral levels. Using the peak transmitted forces at the thoracolumbar disc level as the response variable, the probability of injury for the male spine was obtained from experimental risk curves for the various pulses. Results showed that the shorter pulse durations and rise times impart greater loading on the thoracolumbar spine. The analysis of von Mises stress and strain distributions showed that the compression-related fractures are multifaceted with contributions from both the cortical and cancellous bony components of the body. Profiles are provided in the paper. The intervertebral disc may be involved in the fracture mechanism, because it acts as a medium of load transfer between adjacent vertebrae. Injury risks for the shortest pulse was 63%, and for the widest pulse it was close to zero, and injury probabilities for other pulses are given. The present modeling study provides insights into the mechanisms of internal load transfer and describes injury risk levels from caudal to cephalad impacts.

Introduction

Injuries to the thoracolumbar spine occur in automotive and military environments. In the former scenario, they have been recognized to restrained occupants in frontal impacts in vehicles equipped with airbags. For example, in the analysis of Crash Injury and Research Engineering Network database assembled by the U.S. Department of Transportation, spinal injuries have been reported [13]. Axial loading-induced compression and burst fractures were identified in this and other databases. In the latter scenario, they have been recognized in military personnel associated with combat-related events such as underbody blast loading from improvised explosive devices [46]. A majority of injuries to the mounted soldiers, similar to the automotive environment, were compression-related trauma in the form of vertebral body fracture with or without the involvement of the ligament complex [1,6,7].

Injury analysis in both scenarios has attributed the spine trauma due to an external vertical loading vector. Laboratory studies have been conducted to replicate the field injuries by applying the impact load at the sacral end to the isolated postmortem human surrogate spine specimens and pelvis via the seat to the whole-body surrogates, and injury risk curves have been reported in previous studies [8,9]. The role of the shape of the acceleration pulse on pelvis and spine injuries was discussed in the study that used whole body surrogates, and five specimens were repeatedly tested for this purpose [9]. The effects of mass recruitment were discussed based on an analysis of pulse shapes and injuries sustained by the four specimens. Because of the intact nature of the experimental model, it was not possible to record or estimate the forces transmitted to the proximal end of the lumbar spinal column, while injuries were an outcome of this research. More recently, studies have reported injury probability curves for the thoracolumbar spine from the vertical loading vector using isolated subcomponent models. The objectives of this study were to determine the loads sustained at the thoracolumbar disc level with different acceleration pulses using finite element models and delineate the internal mechanics of the spine from caudo-cephalad impacts.

Materials and Methods

Model.

The finite element model developed by Global Human Body Modeling Consortium, GHBMC, was used in this study [10]. No Changes were made to the supplied model. The midsize male and small size female models were used. Briefly, the thoracolumbar spine of the human body model consists of the vertebral bodies and their posterior elements, intervertebral discs, and interconnecting ligaments. Each vertebral body of the lumbar spine consists of the cortical shell, cancellous bone, endplate, and their posterior bony complexes. The anterior and posterior longitudinal ligaments, supra- and interspinous ligaments, capsular and intertransverse ligaments, and ligamentum flavum, superior costotransverse, and the costal and lateral and intracostal transverse ligaments were included. The material properties and element types from the original model were maintained in this study [11,12]. The ligaments were modeled using elastic beam element with material type 074 (*MAT_ELASTIC_SPRING_DISCRETE_BEAM). The nucleus and annulus subcomponents of the thoracolumbar spinal discs were included at each vertebral level. The annulus and nucleus were modeled using three-dimensional solid hexagonal elements. The material type 177 (*MAT_HILL_FOAM) was used for the annulus, and type 001_Fluid (*MAT_ELASTIC_FLUID) was used for the nucleus. One-dimensional elements and material type 074 (*MAT_EALSTIC_SPRING_DISCRETE) were used to model the ligaments. The cancellous bone was modeled using three-dimensional solid elements with material type 024 (*MAT_PIECEWISE_LINEAR_PLASTICITY) and the bone cortical was modeled using two-dimensional quadrilateral element with material type 003 (*MAT_PLASTIC_KINEMATIC). Table 1 shows the material property data used in the model.

Table 1

Material parameters used in the definition of lumbar spine components

Spine componentParameterValue used in modelRange and references
TrabecularDensity1.1 × 10−6 kg/mm31.62 × 10−7 to 1.85 × 10−4 kg/mm3 [1315]
Elastic modulus0.44 GPa0.09 to 0.54 GPa [16]
Yield stress2.83 × 10−3 GPa5.6 × 10−4 to 3.71 × 10−3 GPa [16]
CorticalThickness0.3 mm0.22 to 0.39 mm [17]
Elastic modulus18.4 GPa12.4 to 48.26 GPa [18]
Yield stress0.19 GPa
AnnulusDensity1.4 × 10−6 kg/mm3[18]
Bulk modulus1 GPa
NucleusDensity1.4 × 10−6 kg/mm3[19]
Bulk modulus1.72 GPa
LigamentForce deflectionActual curvesCurves [20]
Spine componentParameterValue used in modelRange and references
TrabecularDensity1.1 × 10−6 kg/mm31.62 × 10−7 to 1.85 × 10−4 kg/mm3 [1315]
Elastic modulus0.44 GPa0.09 to 0.54 GPa [16]
Yield stress2.83 × 10−3 GPa5.6 × 10−4 to 3.71 × 10−3 GPa [16]
CorticalThickness0.3 mm0.22 to 0.39 mm [17]
Elastic modulus18.4 GPa12.4 to 48.26 GPa [18]
Yield stress0.19 GPa
AnnulusDensity1.4 × 10−6 kg/mm3[18]
Bulk modulus1 GPa
NucleusDensity1.4 × 10−6 kg/mm3[19]
Bulk modulus1.72 GPa
LigamentForce deflectionActual curvesCurves [20]

Positioning of the Model

The model was seated upright on a seat with the Frankfort plane of the head oriented parallel to the horizontal plane, torso was erect, and feet rested on the footrest. The entire human body model was gravity settled into the seat before applying the vertical acceleration-time pulse to the seat. After settling the model on the seat, the torso was restrained with a five-point seatbelt [21]. The seat and footrest were modeled with two-dimensional elements, and the harness was modeled as a combination of two- and one-dimensional seatbelt elements. The fabric material property was assigned to two-dimensional seatbelt elements. A value of 0.3 for the friction coefficient for the surface-to-surface contact between seatbelt and body, and between the seat and body was used, and this magnitude was used in earlier research [22]. A parametric study reported that “small changes in friction coefficient will have minimal effect on predicted injury outcome.” [23] While a full separate parametric analysis of the role of the friction coefficient was not performed, as the chosen value of the coefficient was in the midrange (0 to 0.6 used in the cited study), the outputs reported in the Results section represent a reasonable estimate. The full sensitivity analysis will be a topic of a future investigation. Figure 1 shows the model setup.

Fig. 1
Model setup showing the male and female models (left and middle). Arrows on the bottom show the direction of the input acceleration-time pulse. The illustration on the right shows the components of the spine. Soft tissues are not shown because of the monocolor illustration.
Fig. 1
Model setup showing the male and female models (left and middle). Arrows on the bottom show the direction of the input acceleration-time pulse. The illustration on the right shows the components of the spine. Soft tissues are not shown because of the monocolor illustration.
Close modal

Loading and Output Data.

The vertical input seat acceleration profile had a pulse duration of 100 ms, and the peak acceleration was 22 g [21]. This was termed as the baseline pulse. For both models, the thoracolumbar disc load-time histories and the peak magnitudes of the forces were obtained for the baseline and parameterized pulses. This is termed as spinal forces. The risk of injuries for the male spine model was obtained using the computed peak spinal forces and human cadaver experimental injury probability curves for the baseline and parameterized pulses [8]. The reasons for not using this type of injury risk analysis for the female model are given in the Discussion section. The von Mises stress distributions and effective plastic strain profiles for the T8, T12, and L3 vertebral bodies and cortical bones of the vertebrae were obtained at its mid-depth. For the cancellous bone, a limiting stress of 4.8 MPa and a limiting strain 9.5%, and for the cortical bone, a limiting stress of 113 MPa and a limiting strain of 1.8% were used in this investigation [24].

Results

The applied baseline acceleration pulse is shown in Fig. 2. The maximum accelerations for the parameterized pulses were 46.0 g, 30.6 g, 22.1 g, 17.2 g, and 11.5 g, for the 50 ms, 75 ms, 100 ms, 150 ms, and 200 ms, durations, respectively. They are depicted in Fig. 3. The shape of the baseline pulse was maintained in all parameterized cases.

Fig. 2
Acceleration pulse for the baseline case
Fig. 2
Acceleration pulse for the baseline case
Close modal
Fig. 3
Acceleration pulses for the baseline and parameterized cases. Note the increasing accelerations with shorter pulse durations.
Fig. 3
Acceleration pulses for the baseline and parameterized cases. Note the increasing accelerations with shorter pulse durations.
Close modal

The time histories for the male and female spines are shown for the baseline and parameterized pulses are shown in Fig. 4 for the male model and in Fig. 5 for the female model. The peak force magnitudes were lower in the female spine than the male spine for all cases. The compressive nature of forces is plotted as negative magnitudes in these time history plots.

Fig. 4
Spinal forces from the male model for baseline and parameterized pulses
Fig. 4
Spinal forces from the male model for baseline and parameterized pulses
Close modal
Fig. 5
Spinal forces from the female model for baseline and parameterized acceleration pulses
Fig. 5
Spinal forces from the female model for baseline and parameterized acceleration pulses
Close modal

Figure 6 shows the injury risks for the male spine for all cases. The dashed line in the figure shows the 95% confidence intervals corresponding to the force magnitudes for the male spine. The injury risk was the greatest for the pulse with the shortest time, 62%, and the model shows that the risk is null for longest duration pulse. The peak magnitudes of the male spinal forces and associated injury probabilities are shown (Table 2).

Fig. 6
Injury risks for the male spine model for baseline and parameterized acceleration pulses. The estimated mean and ±95% confidence intervals risk curves are also shown. Data are plotted as squares. See Table 2 for details of actual pulses types for discrete data points.
Fig. 6
Injury risks for the male spine model for baseline and parameterized acceleration pulses. The estimated mean and ±95% confidence intervals risk curves are also shown. Data are plotted as squares. See Table 2 for details of actual pulses types for discrete data points.
Close modal
Table 2

Maximum forces for male for the various pulses and the associated injury risks

Pulse typeForce (kN)Injury risk
50 ms7.6462.5%
75 ms6.2023.3%
100 ms5.286.3%
150 ms4.220.4%
200 ms3.300.0%
Pulse typeForce (kN)Injury risk
50 ms7.6462.5%
75 ms6.2023.3%
100 ms5.286.3%
150 ms4.220.4%
200 ms3.300.0%

A comparison of the von Mises stress profiles for the male and female cancellous bones of the T8, T12, and L3 vertebral bodies are shown in Figs. 7 and 8. The regions of the cancellous bone exceeding the 4.8 MPa threshold are the first regions to respond with any injury. In both cases, shorter duration acceleration pulses tended to result in greater magnitudes and concentration of the stress profiles for this vertebral component for at all spinal levels. The stress profiles are shown at the time of peak magnitudes for all the pulses.

Fig. 7
von Mises stress distributions in the male trabecular bone. T8, T12, and L3 refer to the three spinal vertebral bodies. The stresses are expressed as GPa on the right and pulse definitions are shown on the left.
Fig. 7
von Mises stress distributions in the male trabecular bone. T8, T12, and L3 refer to the three spinal vertebral bodies. The stresses are expressed as GPa on the right and pulse definitions are shown on the left.
Close modal
Fig. 8
von Mises stress distributions in the female trabecular bone. T8, T12, and L3 refer to the three spinal vertebral bodies. The stresses are in GPa on the right of Fig. 7 and pulse definitions are shown on the left.
Fig. 8
von Mises stress distributions in the female trabecular bone. T8, T12, and L3 refer to the three spinal vertebral bodies. The stresses are in GPa on the right of Fig. 7 and pulse definitions are shown on the left.
Close modal

A comparison of the effective plastic strain profiles for the male and female spine cancellous bones of the T8, T12, and L3 vertebral bodies are shown in Figs. 9 and 10. The regions of the cancellous bone exceeding the strain threshold of 9.5% are the first regions to respond with any injury. As in the case of the stress distribution profiles, shorter duration acceleration pulses tended to result in greater magnitudes and concentration of the strain profiles. The T12 level had the region with the greatest strain magnitudes. Table 3 shows the peak magnitudes.

Fig. 9
Effective plastic strain distributions in the male trabecular bone. T8, T12, and L3 refer to the three spinal vertebral bodies. The pulse definitions are shown on the left.
Fig. 9
Effective plastic strain distributions in the male trabecular bone. T8, T12, and L3 refer to the three spinal vertebral bodies. The pulse definitions are shown on the left.
Close modal
Fig. 10
Effective plastic strain distributions in the female trabecular bone. T8, T12, and L3 refer to the three spinal vertebral bodies. The pulse definitions are shown on the left.
Fig. 10
Effective plastic strain distributions in the female trabecular bone. T8, T12, and L3 refer to the three spinal vertebral bodies. The pulse definitions are shown on the left.
Close modal
Table 3

Peak magnitudes of stresses in the male and female spines

Pulse (ms)T8T12L1L2L3L4L5
Male spine
508.6211.505.965.805.615.706.20
758.337.725.374.984.804.805.10
1007.527.024.564.304.264.304.60
1504.554.103.903.853.603.703.80
2004.083.603.403.353.203.203.36
Female spine
507.197.656.386.186.025.915.54
758.017.465.185.284.965.184.92
1005.996.614.404.434.304.514.30
1504.405.753.663.593.653.733.58
2003.424.333.183.153.193.213.10
Pulse (ms)T8T12L1L2L3L4L5
Male spine
508.6211.505.965.805.615.706.20
758.337.725.374.984.804.805.10
1007.527.024.564.304.264.304.60
1504.554.103.903.853.603.703.80
2004.083.603.403.353.203.203.36
Female spine
507.197.656.386.186.025.915.54
758.017.465.185.284.965.184.92
1005.996.614.404.434.304.514.30
1504.405.753.663.593.653.733.58
2003.424.333.183.153.193.213.10

A comparison of the von Mises stress profiles for the male and female spine cortical bones of the T8, T12, and L3 vertebral bodies are shown in Figs. 11 and 12, respectively. Figures 13 and 14 show the effective plastic strain distributions for these vertebrae.

Fig. 11
von Mises stress distributions in the male cortical bone. T8, T12, and L3 refer to the three spinal vertebral bodies. The stresses are expressed as GPa on the right and pulse definitions are shown on the left.
Fig. 11
von Mises stress distributions in the male cortical bone. T8, T12, and L3 refer to the three spinal vertebral bodies. The stresses are expressed as GPa on the right and pulse definitions are shown on the left.
Close modal
Fig. 12
von Mises stress distributions in the female cortical bone. T8, T12, and L3 refer to the three spinal vertebral bodies. The stresses are expressed as GPa on the right and pulse definitions are shown on the left.
Fig. 12
von Mises stress distributions in the female cortical bone. T8, T12, and L3 refer to the three spinal vertebral bodies. The stresses are expressed as GPa on the right and pulse definitions are shown on the left.
Close modal
Fig. 13
Effective plastic strain distributions in the male cortical bone. T8, T12, and L3 refer to the three spinal vertebral bodies. The pulse definitions are shown on the left.
Fig. 13
Effective plastic strain distributions in the male cortical bone. T8, T12, and L3 refer to the three spinal vertebral bodies. The pulse definitions are shown on the left.
Close modal
Fig. 14
Effective plastic strain distributions in the female cortical bone. T8, T12, and L3 refer to the three spinal vertebral bodies. The pulse definitions are shown on the left.
Fig. 14
Effective plastic strain distributions in the female cortical bone. T8, T12, and L3 refer to the three spinal vertebral bodies. The pulse definitions are shown on the left.
Close modal

Discussion

The human thoracolumbar spinal column sustains axial loading under physiological and traumatic loading situations. Clinical studies have focused on the former scenario, and the investigation of low back pain issues and spinal stabilization using artificial devices such as arthroplasty are examples. Investigative studies have largely used quasi-static and vibration loading on the spine segment(s) and spinal columns. The traumatic loading scenario is relatively less researched, and it is a dynamic event. Injuries under this scenario occur in sports, automotive, and combat environments. Impact vectors include flexion–extension modes in automotive crash events. Vertical or caudal to cephalad oriented impacts have been identified in both automotive and military scenarios. Frontal impacts to restrained occupants in the automotive and underbody blast impacts from improvised explosive device in combat situations are examples of the vertical loading vector. Although some studies have been conducted using whole body human cadavers and isolated spinal columns, determinations have not been made of the injury risks and stress and strain responses for a variety of accelerative pulses.

As stated in the introduction, the objective of the study was to investigate the human thoracolumbar spine mechanics from caudo-cephalad or vertical impacts using male and female whole-body finite element models. The presently used model has been adopted in other scenarios, including automotive and military environments, as a tool to investigate the biomechanics of different human body regions under different impact vectors [25]. Others such as the Total HUman Model for Safety, THUMS, can be used [26]. Because differences exist between the two models, benchmarking analysis is needed to reinforce the findings. It should be recognized that the results from computational models, unlike dummy tests, may not be identical due to the differences in factors such as selection of the local and regional anatomical components, and element types and constitutive laws used to simulate the material properties of spinal components. For example, disc nucleus and facet joints with fluids do not always have the same formulation in different models. In fact, consensus or standards do not exist in the modeling community. The current outputs are, therefore, considered specific to this model and modeling process. While this is a limitation of this study, it should be noted that the latest model version was used as previous versions did not incorporate details of the lumbar spine. Inclusion of deformable discs and vertebrae and associated joints in the current model render the outputs realistic. It is possible to apply the same input to subsystem models such as the pelvis and spine, although such models may pose their own nuances such as numerical and mass recruitment issues due to segmentation [9].

The proximal disc location was selected to examine the risk of injury in this study for the following reasons. The applied force is transmitted from the seat to the pelvis to the spine and evaluation of the forces at the proximal end of the thoracolumbar spine represents an examination of the transmitted, rather than applied, force to the spinal column. Furthermore, it is possible to measure such force data in a physical model such as an anthropomorphic test device as a load cell can be introduced in this region. Automotive Hybrid III dummies and Test Device for Human Occupant Restraint (THOR) have available load cells to record such data, while the spines are distinct between the two dummy devices. From this perspective, this computational modeling study in conjunction with the physical device tests (not the purpose of this investigation) serves as a tool for the estimation and actual measurement of loads. As stated, this is difficult to accomplish using the human cadaver experimental model.

The data used to determine the injury risks under different pulses were obtained from experiments conducted by the authors of this study [8]. Although brief, those tests were done using unembalmed human cadaver male spinal columns and applying the vertical impact using custom vertical accelerator device [27]. Forty-three spines were subjected to caudo-cephalad impacts, and risk curves were obtained using parametric survival analysis. The specific peak force data from this study that were plotted on the experimental risk curve show the probability of sustaining spinal injuries for different pulses/scenarios. As expected, the greatest probability of injury was associated with the shortest pulse, and the lowest risk was with the longest duration pulse. It should be noted that all pulses corresponded to the same total energy. The change in risk levels represents the influence of pulse duration and rise time on the impact biomechanics of the human thoracolumbar spinal column. With the application of the pulse with the longest dwelling time, the spinal column consisting of the discs and facet joints, and to a certain extent the vertebrae, has adequate time to absorb the impact energy and reduce the likelihood of injury, i.e., the influence of viscoelastic effects. With the shorter duration pulse, the inadequate time appears to be factor for greater force transmission and injures associated with greater risks. Mass recruitment effects are also lower with shorter duration pulse rise times. From this viewpoint, energy-absorbing or padded seat structures appear to be a good countermeasure for mitigating spinal trauma at these input levels.

While a similar pattern is expected for the female spine responses, injury risk magnitudes for each pulse were not determined because the experimental data were conducted using male human cadavers. While a general or overall scaling can be done in parallel to the neck injury criteria process used in the U.S. Federal Motor Vehicle Safety Standards, FMVSS 208, because true gender differences exist, there is a need to conduct experiments with female specimens. The authors are pursuing this effort as a future study.

Although the rationale for determining the forces was to compare with experimental data and extract the risks based on experimental injury risk curves, it is a structural parameter. Direct validations of a human body model for spine forces can be done with human cadaver tests; however, recording the spinal load in an intact specimen is challenging/infeasible. Inserting a load cell into the intervertebral disc space will compromise the integrity of the column and acts as a fictitious load path. Because the present model was validated with human cadaver data for parameters such as global and local kinematics, it was used to obtain spinal forces in this study [11]. No other validation studies were done. While this is a limitation, other models should be used for benchmarking, and experiments are needed with targets and accelerometers inserted on the bony surfaces of the spinal vertebrae, and perhaps strain gages can be glued to the cortices. These are future studies.

While this study reported forces at the midbody level, it is possible to determine the forces from the different components, e.g., ligaments that were modeled as one-dimensional elements, by changing the cross section to mid-disc or facet. This separate study will assist in delineating the roles of different components of the spine to resist the caudo-cephalad impacts.

The internal stress and strain profiles were examined for both models as the next investigative step to describe the mechanism of load transfer within the spinal column. This can only be done using computational modeling. As the normalization process (from force to stress) accounts for the geometry, it is possible to determine the potential locations of injury using these intrinsic and local variables. Studies in the spine literature have used such measures as potential indicators of injury. They are not cited due to the limit on the number of references. The chosen values for the cortical and cancellous bones, i.e., the von Mises stress magnitudes of 113 MPa and 4.8 MPa, and effective plastic strain magnitudes of 9.5% and 1.8% were adopted from a previous study [24]. While a consensus does not exist on the exact thresholds for both stress and strain parameters, the type and magnitude used in this study are reasonable.

To determine the patterns of spinal forces, stresses, and strains using a parametric approach, methods such as iso-energy can be used. In this study, the pulses were varied by varying the magnitude of the acceleration to maintain constant energy input. The iso-energy approach was chosen to guide in the future design of experiments for countermeasure development. For example, the role of energy absorbing materials attached to the seat frame to decrease lumbar loads can be investigated with equivalent inputs. A similar approach has been used in automotive studies [28,29].

While this study was based on extending the baseline pulse in terms of the abscissa and ordinate on either ends to conduct the parametric study, the injury probabilities were obtained for the force variable. This is because risk curves were available for the force parameter only. In order to improve the predictability, advanced variables such as pressure-based risk curves as a function of input pulse are needed. This can be achieved with experiments with human cadaver spines and is considered as a topic for future experimental investigators.

The stress and strain profiles had greater regions of involvement (defined as increase in magnitude and area) from the lumbar to the thoracic vertebrae. Although not shown for all vertebrae, this pattern was true for other vertebrae, from the caudal to cephalad direction. This phenomenon represents the vertical load transferring mechanism. The cancellous bone stress profiles were concentrated in the inside regions of the vertebrae, for both spines, at all levels, and for all pulses. The greatest concentration was biased to the thoracolumbar level in both the spines. This pattern was also true for the strain distributions, suggesting that the cancellous bone is more susceptible to fracture at the thoracolumbar spinal level.

An examination of the cortical bone stress–strain profiles revealed a relatively uniform concentration along the periphery at all levels, and this was true for male and female spines and for all pulses. This expected response is due to the cortical anatomy in the human spine. The pattern of increasing stresses along the caudal to cephalad direction is due to the impact vector. The strain distributions reveal that the upper thoracic level is more susceptible to injury (greater magnitudes) than the caudal levels, and this pattern is different from the cancellous bone, where it was concentrated more on the thoracolumbar level. Thus, the injury at the upper thoracic level may stem from the cortical bone, and at the lower level, it may initiate from the trabecular/cancellous bone. In other words, the injury at the upper thoracic level may be due to the cortical bone exceeding the strain threshold while at the caudal levels it may be due to the core of the vertebral body exceeding the trabecular compressive stress magnitude. Other factors might include the size of the vertebra. It is known that intervertebral discs play a role in fractures to the vertebrae [30]. Hydraulic mechanisms applying a bursting-type pressure to the vertebral bodies via their caudal or cephalad endplates have been postulated in earlier studies of human cadaver segmented spine tests. Injury to the trabecular bone in the body may be therefore, due to a combination of these variables. Cortical bone failure may also lead to trabecular damage as the cortical indexes are lower in the human spine than the long bones [24]. In other words, the protective effect is less in the vertebral column. Any curvature changes, especially the accentuation of the kyphosis during the loading process, may predispose the spine to fracture. These findings reveal that the compression-related fractures of spinal vertebrae are multifaceted, and the present computational modeling study has provided some insights into the mechanisms of the internal load transfer and describe the injury and injury risk levels from caudal to cephalad impacts.

Conclusions

Using a computational whole-body human body model subjected to caudo-cephalad impacts from the pelvis to the spine, this study showed that shorter pulse duration and rise time induce greater loading on the thoracolumbar spine. Shorter pulses were associated with greater acceleration magnitudes, and the loading rate increased from 0.2 g/ms for the widest (200 ms) pulse to the 3 g/ms for the shortest pulse (50 ms). In other words, the increased loading in the thoracolumbar spine was attributed to greater rates associated with shorter pulses. The issue of which of the two parameters, amplitude or time, contribute more to the force increase would be a separate future study. An analysis of the stress–strain distributions showed that compression-related fractures of spinal vertebrae are multifaceted with contributions from both the cortical and cancellous bony components. The intervertebral disc may be involved in the fracture mechanism. The present computational modeling study has provided insights into the mechanisms of the internal load transfer and described the injury and injury risk levels from caudal to cephalad impacts.

Acknowledgment

The research was supported by the National Center for Advancing Translational Sciences, National Institutes of Health, Award No. UL1TR001436, and by the Office of the Assistant Secretary of Defense for Health Affairs, through the Broad Agency Announcement under Award No. W81XWH-16-1-0010. It was also supported by the Department of Veterans Affairs Medical Research. This material is the result of work supported with resources and use of facilities at the Zablocki VA Medical Center, Milwaukee, WI, and the Medical College of Wisconsin. Dr. Yoganandan is an employee of the VA Medical Center. We thank the GHBMC for providing the model for the study. The opinions, interpretations, conclusions, and recommendations are those of the authors and are not necessarily endorsed by the NIH, Department of Defense, or other sponsors.

Funding Data

  • National Institutes of Health (No. UL1TR001436; Funder ID: 10.13039/100000002).

  • U.S. Department of Defense (No. W81XWH-16-1-0010; Funder ID: 10.13039/100000005).

References

1.
Pintar
,
F. A.
,
Yoganandan
,
N.
,
Maiman
,
D. J.
,
Scarboro
,
M.
, and
Rudd
,
R. W.
,
2012
, “
Thoracolumbar Spine Fractures in Frontal Impact Crashes
,”
Ann. Adv. Automot. Med.
,
56
, pp.
277
283
.https://www.ncbi.nlm.nih.gov/pubmed/23169137
2.
Kaufman
,
R. P.
,
Ching
,
R. P.
,
Willis
,
M. M.
,
Mack
,
C. D.
,
Gross
,
J. A.
, and
Bulger
,
E. M.
,
2013
, “
Burst Fractures of the Lumbar Spine in Frontal Crashes
,”
Accid. Anal. Prev.
,
59
, pp.
153
163
.10.1016/j.aap.2013.05.023
3.
Jakobsson
,
L.
,
Björklund
,
M.
, and
Westerlund
,
A.
,
2016
, “
Thoracolumbar Spine Injuries in Car Crashes
,”
IRCOBI
, Malaga, Spain, Sept. 14–16, Paper No. IRC16-22, pp.
104
112
.http://www.ircobi.org/wordpress/downloads/irc16/pdf-files/22.pdf
4.
Danelson
,
K.
,
Watkins
,
L.
,
Hendricks
,
J.
,
Frounfelker
,
P.
,
Pizzolato-Heine
,
K.
,
Valentine
,
R.
, and
Loftis
,
K.
,
2018
, “
Analysis of the Frequency and Mechanism of Injury to Warfighters in the Under-Body Blast Environment
,”
Stapp Car Crash J.
,
62
, pp.
489
513
.https://pubmed.ncbi.nlm.nih.gov/30609005/
5.
Owens
,
B. D.
,
Kragh
,
J. F.
, Jr.
,
Wenke
,
J. C.
,
Macaitis
,
J.
,
Wade
,
C. E.
, and
Holcomb
,
J. B.
,
2008
, “
Combat Wounds in Operation Iraqi Freedom and Operation Enduring Freedom
,”
J. Trauma
,
64
(
2
), pp.
295
299
.10.1097/TA.0b013e318163b875
6.
Schoenfeld
,
A. J.
,
Goodman
,
G. P.
, and
Belmont
,
P. J.
, Jr.
,
2012
, “
Characterization of Combat-Related Spinal Injuries Sustained by a U.S. Army Brigade Combat Team During Operation Iraqi Freedom
,”
Spine J.
,
12
(
9
), pp.
771
776
.10.1016/j.spinee.2010.05.004
7.
Helgeson
,
M. D.
,
Bevevino
,
A. J.
, and
Hilibrand
,
A. S.
,
2013
, “
Update on the Evidence for Adjacent Segment Degeneration and Disease
,”
Spine J.
,
13
(
3
), pp.
342
351
.10.1016/j.spinee.2012.12.009
8.
Yoganandan
,
N.
,
DeVogel
,
N.
,
Moore
,
J.
,
Pintar
,
F.
,
Banerjee
,
A.
, and
Zhang
,
J.
,
2020
, “
Human Lumbar Spine Responses From Vertical Loading: Ranking of Forces Via Brier Score Metrics and Injury Risk Curves
,”
Ann. Biomed. Eng.
,
48
(
1
), pp.
79
91
.10.1007/s10439-019-02363-5
9.
Yoganandan
,
N.
,
Moore
,
J.
,
Arun
,
M. W.
, and
Pintar
,
F. A.
,
2014
, “
Dynamic Responses of Intact Post Mortem Human Surrogates From Inferior-to-Superior Loading at the Pelvis
,”
Stapp Car Crash J.
,
58
, pp.
123
143
.https://pubmed.ncbi.nlm.nih.gov/26192952/
10.
Elemance LLC
,
2019
, “
Global Human Body Models Consortium User Manual: M50 Detailed Occupant Version 5.1.1, Version 3.1 F05 for LS-DYNA®
,” Elemance LLC, Society of Automotive Engineers, Warrendale, PA.
11.
Aira
,
J.
,
Guleyupoglu
,
B.
,
Jones
,
D.
,
Koya
,
B.
,
Davis
,
M.
, and
Gayzik
,
F. S.
,
2019
, “
Validated Thoracic Vertebrae and Costovertebral Joints Increase Biofidelity of a Human Body Model in Hub Impacts
,”
Traffic Inj. Prev.
,
20
(
Suppl. 2
), pp.
S1
S6
.10.1080/15389588.2019.1638511
12.
Poulard
,
D.
,
Kent
,
R. W.
,
Kindig
,
M.
,
Li
,
Z.
, and
Subit
,
D.
,
2015
, “
Thoracic Response Targets for a Computational Model: A Hierarchical Approach to Assess the Biofidelity of a 50th-Percentile Occupant Male Finite Element Model
,”
J. Mech. Behav. Biomed. Mater.
,
45
, pp.
45
64
.10.1016/j.jmbbm.2015.01.017
13.
Richards
,
A. M.
,
Coleman
,
N. W.
,
Knight
,
T. A.
,
Belkoff
,
S. M.
, and
Mears
,
S. C.
,
2010
, “
Bone Density and Cortical Thickness in Normal, Osteopenic, and Osteoporotic Sacra
,”
J. Osteoporosis
,
2010
, pp.
1
5
.10.4061/2010/504078
14.
Yoganandan
,
N.
,
Pintar
,
F. A.
,
Stemper
,
B. D.
,
Baisden
,
J. L.
,
Aktay
,
R.
,
Shender
,
B. S.
, and
Paskoff
,
G.
,
2006
, “
Bone Mineral Density of Human Female Cervical and Lumbar Spines From Quantitative Computed Tomography
,”
Spine
,
31
(
1
), pp.
73
76
.10.1097/01.brs.0000192684.12046.93
15.
Yoganandan
,
N.
,
Pintar
,
F. A.
,
Stemper
,
B. D.
,
Baisden
,
J. L.
,
Aktay
,
R.
,
Shender
,
B. S.
,
Paskoff
,
G.
, and
Laud
,
P.
,
2006
, “
Trabecular Bone Density of Male Human Cervical and Lumbar Vertebrae
,”
Bone
,
39
(
2
), pp.
336
344
.10.1016/j.bone.2006.01.160
16.
Kopperdahl
,
D. L.
, and
Keaveny
,
T. M.
,
1998
, “
Yield Strain Behavior of Trabecular Bone
,”
J. Biomech.
,
31
(
7
), pp.
601
608
.10.1016/S0021-9290(98)00057-8
17.
Ritzel
,
H.
,
Amling
,
M.
,
Posl
,
M.
,
Hahn
,
M.
, and
Delling
,
G.
,
1997
, “
The Thickness of Human Vertebral Cortical Bone and Its Changes in Aging and Osteoporosis: A Histomorphometric Analysis of the Complete Spinal Column From Thirty-Seven Autopsy Specimens
,”
J. Bone Miner. Res.
,
12
(
1
), pp.
89
95
.10.1359/jbmr.1997.12.1.89
18.
McElhaney
,
J. H.
,
1966
, “
Dynamic Response of Bone and Muscle Tissue
,”
J. Appl. Physiol.
,
21
(
4
), pp.
1231
1236
.10.1152/jappl.1966.21.4.1231
19.
Yang
,
K. H.
, and
Kish
,
V. L.
,
1988
, “
Compressibility Measurement of Human Intervertebral Nucleus Pulposus
,”
J. Biomech.
,
21
(
10
), p.
865
.10.1016/0021-9290(88)90059-0
20.
Yoganandan
,
N.
,
Kumaresan
,
S.
, and
Pintar
,
F. A.
,
2000
, “
Geometric and Mechanical Properties of Human Cervical Spine Ligaments
,”
ASME J. Biomech. Eng.
,
122
(
6
), pp.
623
629
.10.1115/1.1322034
21.
McEntire
,
J.
,
Logsdon
,
K.
,
Vasquez
,
K.
,
Daniel
,
R.
,
Brozoski
,
F. T.
,
Lidl
,
G. M.
, and
Mazuchowski
,
E. L.
,
2018
, “
Thoracolumbar Injuries Resulting From Whole-Body Exposure to Vertical Acceleration
,”
Military Health System Research Symposium
, Kissimmee, FL, Aug. 20–23, p. 1.
22.
Arun
,
M. W.
,
Umale
,
S.
,
Humm
,
J. R.
,
Yoganandan
,
N.
,
Hadagali
,
P.
, and
Pintar
,
F. A.
,
2016
, “
Evaluation of Kinematics and Injuries to Restrained Occupants in Far-Side Crashes Using Full-Scale Vehicle and Human Body Models
,”
Traffic Inj. Prev.
,
17
(
Suppl. 1
), pp.
116
123
.10.1080/15389588.2016.1197394
23.
Xiao
,
S.
,
Yang
,
J.
,
Forman
,
J.
,
Panzer
,
M.
,
Xiao
,
Z.
, and
Crandall
,
J.
,
2016,
Effect of Contact Friction Between Seatbelt and Human Body Model on Simulation of Rib Fracture in Frontal Impact
,”
Eighth International Conference on Measuring Technology and Mechatronics Automation
(
ICMTMA
), Macau, China, Mar. 11–12, pp.
255
257
.10.1109/ICMTMA.2016.69
24.
Zhao
,
J. Z.
, and
Narwani
,
G.
,
2007
, “
Biomechanical Analysis of Hard Tissue Responses and Injuries With Finite Element Full Human Body Model
,”
20th International Technical Conference on the Enhanced Safety of Vehicles
(
ESV
),
NHTSA
,
Lyon, France
, 18–21, pp.
1
15
.https://www.semanticscholar.org/paper/BIOMECHANICAL-ANALYSIS-OF-HARD-TISSUE-RESPONSES-AND-Zhao-Narwani/cc7099750ea9acb27c2a63a6b5ef606900378e37
25.
Umale
,
S.
,
Yoganandan
,
N.
,
Pintar
,
F. A.
, and
Arun
,
M. W. J.
,
2018
, “
Factors Influencing the Effectiveness of Occupant Retention Under Far-Side Impacts: A Parametric Study
,”
J. Mech. Behav. Biomed. Mater.
,
84
, pp.
235
248
.10.1016/j.jmbbm.2018.05.021
26.
Iwamoto
,
M.
,
Nakahira
,
Y.
, and
Kimpara
,
H.
,
2015
, “
Development and Validation of the Total HUman Model for Safety (THUMS) Toward Further Understanding of Occupant Injury Mechanisms in Precrash and During Crash
,”
Traffic Inj. Prev.
,
16
(
Suppl. 1
), pp.
S36
48
.10.1080/15389588.2015.1015000
27.
Yoganandan
,
N.
,
Pintar
,
F. A.
,
Schlick
,
M.
,
Humm
,
J. R.
,
Voo
,
L.
,
Merkle
,
A.
, and
Kleinberger
,
M.
,
2015
, “
Vertical Accelerator Device to Apply Loads Simulating Blast Environments in the Military to Human Surrogates
,”
J. Biomech.
,
48
(
12
), pp.
3534
3538
.10.1016/j.jbiomech.2015.06.008
28.
Kuppa
,
S.
,
Eppinger
,
R. H.
,
McKoy
,
F.
,
Nguyen
,
T.
,
Pintar
,
F. A.
, and
Yoganandan
,
N.
,
2003
, “
Development of Side Impact Thoracic Injury Criteria and Their Application to the Modified ES-2 Dummy With Rib Extensions (ES-2RE)
,”
Stapp Car Crash J.
,
47
, pp.
189
210
.https://pubmed.ncbi.nlm.nih.gov/17096250/
29.
Maltese
,
M. R.
,
Eppinger
,
R. H.
,
Rhule
,
H. H.
,
Donnelly
,
B. R.
,
Pintar
,
F. A.
, and
Yoganandan
,
N.
,
2002
, “
Response Corridors of Human Surrogates in Lateral Impacts
,”
Stapp Car Crash J.
,
46
, pp.
321
351
.https://pubmed.ncbi.nlm.nih.gov/17096232/
30.
Nowicki
,
B. H.
,
Yu
,
S.
,
Reinartz
,
J.
,
Pintar
,
F.
,
Yoganandan
,
N.
, and
Haughton
,
V. M.
,
1990
, “
Effect of Axial Loading on Neural Foramina and Nerve Roots in the Lumbar Spine
,”
Radiology
,
176
(
2
), pp.
433
437
.10.1148/radiology.176.2.2367657