An Integrated Computational Simulation System for Injury Assessment

Abstract

Injury prediction and prevention are subject areas that will significantly benefit from the use of digital human models (DHMs). The subject of this research is to investigate human simulation to predict injuries. This work over the past few years seeks to integrate high-fidelity computational methods for stress/strain analysis, namely finite element analysis (FEA) with biomechanics predictions through DHMS to yield measures (indices) for the propensity for injury. Indeed, a multi-scale FEA model using continuum-mechanics-based theories was developed in order to obtain highly accurate simulations of the segmental stress fields during a specific task. Previous work by this group is a simulation environment called Santos™ that enables the prediction of human motion including all aspects of its biomechanics. The Santos environment provides a joint- on physics-based, predictive capability including a muscle model. While FEA models are certainly useful as independent tools, their benefits can be fully realized when they are integrated with a complete system-level human model. This integration essentially connects the local model to a virtual environment, whereby the DHM model yields the muscle forces and motion profiles (i.e., the kinematics of the motion across time for each degree-of-freedom for the body). These motion profiles and muscle forces are calculated for each task and are used as input for the multi-scale FEA model. The results of the FEA model executed across a statistically viable set of data is fed into a neural network for learning and evaluating a pre-determined injury index measure that was developed by this group. This paper presents the initial promising results for this integrated multi-scale approach to quantify and predict injury in a particular joint that is undergoing a specific motion. The joint injury index system is developed based on the yield stress of the joint components. The injury index includes both the bone and soft tissue structures of the joint where bone was modeled as elastic material and the soft tissue was modeled as hyper-elastic material with the Noe-Hookean method. This integrated system allows one to study the effects of various motions and task-parameters on knee joints so as to modify tasks, save analysis time, and reduce the likelihood of injury.

Appendices

Appendix 1

For the ankle joint, the compression (A _comp ) and shear (A _shear ) forces are computed as shown in the following equations [8]:

$$ \begin{aligned} A_{comp} & = f_{c - sol} + f_{c - medgas} + f_{c - tibant} + f_{c - tibpost} + GRF_{z} \cos (\theta_{ankle} ) \\ & \quad - GRF_{y} { \sin }\left( {\theta_{ankle} } \right) - mg_{foot} \cos \left( {\theta_{ankle} } \right) \\ A_{shear} & = f_{s - sol} + f_{s - medgas} + f_{s - tibant} + f_{s - tibpost} + GRF_{z} \sin (\theta_{ankle} ) \\ & \quad - GRF_{y} { \cos }\left( {\theta_{ankle} } \right) - mg_{foot} \sin \left( {\theta_{ankle} } \right) \\ \end{aligned} $$

For the knee joint, the muscle compression (M _comp ) and muscle shear (M _shear ) forces are computed as shown in the following equations:

$$ \begin{aligned} M_{comp} & = f_{{c{-}recfem}} + f_{c - vasint} + f_{c - sar} + f_{c - tfl} + f_{c - bifemlh} \\ & \quad + f_{c - bifemsh} + f_{c - grac} - mg_{tibia} \cos \left( \lambda \right) \\ M_{shear} & = f_{s - recfem} + f_{s - vasint} + f_{s - sar} + f_{s - tfl} + f_{s - bifemlh} \\ & \quad + f_{s - bifemsh} + f_{s - grac} - mg_{tibia} \sin \left( \lambda \right) \\ \end{aligned} $$

where $f_{{c{-}sol}} ,f_{{c{-}medgas}} ,f_{{c{-} tibant}} ,f_{{c{-} tibpost}}$ are the compression components of the ankle muscles, $f_{s - sol} ,f_{s - medgas} ,f_{s - tibant } ,f_{s - tibpost}$ are the shear components of the ankle muscles, $f_{c - recfem} ,f_{c - vasint} ,f_{c - sar} ,f_{c - tfl} ,f_{c - bifemlh} ,f_{c - bifemsh} ,f_{c - grac}$ are the compression components of the knee muscles, and $mg_{tibia} , mg_{foot}$ are the shin and foot weight, respectively.

$f_{s - recfem } , f_{s - vasint} , f_{s - sar} , f_{s - tfl } , f_{s - bifemlh} , f_{s - bifemsh} , f_{s - grac}$ are the shear components of knee muscles, $GRF_{z} ,GRF_{y}$ are the z and y components of the ground reaction force, and $\lambda , \theta_{ankle}$ are the knee and ankle angles, respectively (Fig. 8 and Table 1).

Fig. 8

a Muscle force analysis [2, 8], b injury prediction algorithm [2, 8]

Table 1 RMSE and MAE values for right knee walking and stair ascending

Appendix 2

See Tables 2 and 3.

Table 2 Injury index for knee joint components at 17 kg backpack loading

Table 3 Comparison of injury index at different loadings for tibia bone and lateral meniscus in walking