HPC MSU

Publication Abstract

Creation of a Computational Simulation of Maternal Trauma in Motor Vehicle Accident

Weed, B., & Liao, J. (2013). Creation of a Computational Simulation of Maternal Trauma in Motor Vehicle Accident. Mississippi State: Mississippi State University.

Abstract

Maternal trauma in car crash is a significant source of injury, morbidity, and mortality for both mothers and fetuses. Motor vehicle related mortality in pregnant women is highest in USA and account annually for about 93,000 cases [1]. Additionally, the number of fetal deaths since four decades has increased as much as seven time the mortality of infant deaths [2]. The primary maternal pathology associated with maternal trauma is placental abruption where the placenta detaches, partially or completely, from the uterus. This results in serious complications such as hemorrhaging and termination of pregnancy. The stability of the pregnancy and the successful transfer of nutrients from the mother to the fetus can also be affected, even under minor levels of abruption. Beyond the placental aspects of trauma, the fetus can also be directly injured. These injuries frequently result in death, and have been associated with cognitive and developmental deficits in surviving individuals. Maternal trauma is difficult to study with traditional experiments, because of many logistical and ethical issues. A very small number of studies have succeeded in using post mortem subjects to create some very valuable data sets, but the availability of pregnant human cadaveric subjects or tissues is very low or negligible. There is also not a suitable animal surrogate for use in these studies as only primates are even remotely similar for crash testing, and chimpanzee research in automotive collisions was ceased during the 1970s due to ethical concerns balanced against the relatively poor scientific value caused by anatomical differences. Computational simulations have shown great promise in recreating biological experiments in silico without physical specimens, through the use of finite element simulations. Finite element simulations of trauma has been studied in various tissues and organ systems, including brain, liver, and pregnancy. These simulations are based on constitutive relationships which must capture the mechanical properties of these biological materials. The material properties can display very complex mechanical behavior, include simultaneous occurrence of complex phenomena such as hyperelasticity, viscoelasticity, stress state dependency, damage accumulation, anisotropy, and other ultrastructural effects, all while sustaining extremely large deformations. The creation of material models capable of capturing such trends requires appropriate experimental data and analysis, before mathematical incorporation in the constitutive relationships. In addition to material properties, finite element analysis requires three-dimensional geometric meshes which are deformed during simulations. These geometries can be created using traditional tools for three-dimensional design such as computer aided design tools, which are essentially basic approximations. It has recently become more common to use meshes derived from real anatomy, through the use of meshing tools which convert scans from medical imaging technologies such as MRI or CT scanning into three-dimensional meshes for direct use in finite element software packages. The final major component required for a finite element simulation is the set of loading conditions which will drive the simulation. This study will endeavor to create a computational simulation of maternal trauma based on medical image derived anatomical meshes, and material models based on advanced mechanical testing. The specific goals and aims of this study are outlined below.