In This Issue
Spring Bridge on Concussion: A National Challenge
April 12, 2016 Volume 46 Issue 1

Computational Models of Impact and Blast Force Effects on the Brain Scaling of Animal Injury Models and Prediction of Human TBI

Monday, April 18, 2016

Author: Raúl A. Radovitzky, James Q. Zheng, and Thomas F. Budinger

The goal of our research is to determine the forces transmitted to the brain resulting from the stress field of a military blast, fall, or sport or vehicle collision involving the head. This information is essential to link the external insult to the mechanism of brain tissue injury.

Epidemiological and experimental animal data relate the injury pathology to the physical parameters of the external impact, but there are several limitations. Biofidelic physical models do not include the necessary distribution of tissue stimulants with accurate tissue material properties of the human head and brain, nor the required spatial distribution of pressure sensors, to allow stress field measurements with the needed temporal and spatial fidelity. Human cadaver studies are limited by postmortem changes in material properties as well as the logistics of performing them in realistic blast and projectile impact situations. Sensors in helmets that relate acceleration and deceleration levels to a victim’s behavior reveal no information about the internal mechanisms of tissue injury or even reliable thresholds for concussions.

Computer simulations can provide the high spatial and temporal resolutions needed to evaluate three-dimensional stress fields as a function of time. And the material properties corresponding to tissues of the head, skull, and brain can be assigned to human head models using magnetic resonance imaging (MRI) data and archival tissue property measurements.

Need for Animal Models

Animal models are needed to relate external forces to internal stress fields, but are useful only if the tissue injury probability functions can be translated to expectations for the human head and brain. In this paper we describe a scaling law that allows this vital transformation from animal experiments on many species to the human situation.

Knowledge of the internal force fields generated by the external impact of a broad range of physical threats will enable understanding of the mechanisms of injury. This knowledge is basic to the determination of the tolerances of the brain to external trauma and to the development of improved protection systems. Linkages between the transfer of momentum or rate of change of momentum (force) and types of tissue injury cannot be determined from current experimental data without a validated scaling law from an animal species to the human head and brain. Nor can the distribution of tissue injury and long-term effects be understood without knowledge of the stress fields and accompanying strains for specific physical impacts to the head.

Human Head and Brain Blast Simulation Studies

We performed a critical set of experiments to answer the question, Do the helmet and face mask afford important protection to the soldier exposed to a blast? The two major components of the experiments were a detailed model of the human head and a detailed description of the temporal stress field for blasts up to 800 kilopascals (kPa) arriving at the head.

Simulation of the Human Head and Brain

The design of a simulation model of the human head and brain started with an anatomical description of the position and type of all the tissues. For each tissue we assigned material parameters including elastic properties such as Young’s modulus, the shear modulus, and viscoelastic properties, as well as anisotropy tensors obtained by detailed diffusion tensor imaging using magnetic resonance. We developed this model with the Defense and Veterans Brain Injury Center (Moore et al. 2009); it was the most comprehensive model of human head material properties, with the highest spatial resolution that was practical given the computing power available.

To evaluate the protection afforded by the advanced combat helmet (ACH) we added its material properties (e.g., poroelasticity), including padding and air spacing, as these are critical to the proper description of reflections and transmission at acoustic impedance–mismatched surfaces. We added the ACH face shield position and properties as well.

Simulation of the Blast

The next major component of the experiment was a physically accurate description of the blast itself. This was a three-dimensional hydrodynamic model. The mathematical tools and related algorithms for executing the time-dependent changes of a blast involved extensive calculations that the MIT group has been perfecting for many years.

In our first experiments we showed that a blast wave has a direct transmission of stress waves into the brain through regions of the skull with soft tissue (e.g., ear canals, sinuses, nasal orifices) (Moore et al. 2009). The direction of the blast is an important determinant of the resulting stress fields in the brain, which is also the case for hits to a helmet. Figure 1 shows simulations to quantitatively evaluate the effectiveness of the helmet and shield (Nyein et al. 2010). We found that both the helmet and face shield provided significant protection, an important finding as there had been a major controversy about the possibility that the helmet increased the risk of brain injury. The results made a major impact on military protection science.

Figure 1

Validation of these findings using a skull and tissue stimulants in which pressure sensors were embedded is shown in figure 2. These experiments relate to protective systems for reducing stress fields. The major goal, however, is to determine the relationships between stress fields and resulting brain tissue injuries. For this goal we rely on animal models and some method to translate the data from animal models to the human head and brain.

Figure 2

Animal Models

The experimental arrangements for studying the effects of a range of mechanical events leading to brain tissue injuries differ depending on whether the study is for nonpenetrating low stress-rate collisions (e.g., sports and vehicle collisions), high stress-rate impacts (e.g., projectiles impacting helmets), or much higher stress-rate events from explosive blasts.

Numerous methods are used to create the accelerations and physical stresses involved, but unfortunately many experiments fall short of the actual conditions experienced on the battlefield. Also problematic is the use of fixed animals that do not undergo the rotational accelerations experienced by human subjects because of the differences between a head mass that is loosely tethered to the body and the total head-body mass. Examples of animal experiments used to evaluate battlefield blast effects are reported in this issue (Kovacs, Margulies) and elsewhere (e.g., Bauman et al. 2009). The results from these experiments show either the probability of a specific injury versus the physical insult or the characteristics of tissue injury from different intensity levels or numbers of hits.

Of key importance to understanding the tolerance of the human brain to exposure to blast and other stress fields is a reliable method for relating the results of animal experiments to the human brain. We present such a method below.

Translation of Animal Experiments to Human Injury Predictions

Motivated by crash injury reduction research in the automotive industry as well as efforts to mitigate battlefield injuries using protective gear, animal experiments have been conducted in this area since World War II. They are usually guided by mechanical and physiological principles, but brain size, head tissue architecture, brain to body mass ratio, skull thickness, and other factors vary widely from species to species. Thus some scaling model is needed to relate the animal experimental data to what would be expected in the human head and brain.

Figure 3

Simulations are essential for developing a scaling law. The experiments described here involve simulations with three species—mice, pigs, and humans—that differ greatly in size and head architecture (figure 3). Simulations were performed with these species using three different incident blast pressures. An example is shown in figure 4. The models were based on the anatomy collected by MRI studies and tissue types using parameters similar to those used for human tissue. Once the tissue properties are assigned to volume elements and the physics of a blast is defined in space and time, the simulations give internal pressures relative to exposed stresses.

Figure 4

The human brain receives the highest maximum intracranial pressure from a blast, as observed for three different blast intensities in the transmitted stress range associated with concussion. The pig experiences the lowest intensities, due to the ratio of brain to skull and soft tissues. The comparison of human to swine brain mass is even more extreme (figure 3).

Figure 5

A proper parameter for scaling blast effects is a brain vulnerability parameter, Ƞ (the horizontal coordinate in figure 5). This is a ratio of (acoustic impedance × the mass of the brain) ÷ the sum of (acoustic impedance × the mass of the skull) + (acoustic impedance × the mass of the head tissues external to the skull). This factor results from considering this as a stress wave propagation problem. The eta parameters are 0.02, 0.13, and 0.75 for the mouse, pig, and human, respectively. The result of this work is summarized in figure 5.

The major conclusion is that humans are the most vulnerable among all mammal species, mostly because of their large brain and the relatively sparse tissue and bone protection of the head. This model enables conversion of an injury threshold S-curve for a particular type of brain injury as determined for one species to the S-curve for the human brain. An example of the major difference in predicted human survivability scaling from rabbit blast survival experiments is shown in figure 6.

Figure 6

Scaling based on mass differences among species (Rafaels et al. 2011) leads to the prediction that the human brain can sustain a higher threshold than the species tested. Our new scaling law actually predicts a much lower threshold for brain injury in humans.

This is of major importance as it allows a much broader use of animal experiments wherein survivability would not be the experimental parameter, using instead other measurements that cannot be performed on human subjects (e.g., axonal injury, accumulation of protein aggregates, micropetechiae).

Complementary studies that compare the histopathology, three-dimensional neuronal architecture, and injury endpoints to information that characterizes the mechanics of the impact cannot reveal mechanisms for the observed injuries without knowledge of the force fields in the brain tissue resulting from the impact. This is the principal value and need for computer simulations with high spatial and temporal fidelity.

Proposed Studies

Two activities under way are collaborations with the Boston Marathon blast investigation, where our simulations can provide the exposure levels the victims suffered, and with the Lawrence Berkeley National Laboratory and Boston University on the investigation of stress fields in the basal-midbrain regions (including the hypothalamus, pituitary, and amygdale tissues). The long-term effects of head injuries on pituitary function are not understood, though the prevalence of pituitary dysfunction is known in NFL players, survivors of unconsciousness episodes from collisions, and war veterans.

An illustration of the importance of careful simulations with the material properties of each volume element can be appreciated from considering the complex pressure field when the variation of the pressure wave speed can be a factor of 3. The simulation will require 2 mm voxels to compute the strain map for a range of incident stresses and stress rates.


Computer simulation provides detailed estimates of the forces transmitted to the brain from an impact or blast event that may cause concussion. Animal models are not adequate for determination of the relationship between a collision or blast event and tissue damage in the human subject without a method to translate the risk threshold curves between species as shown in this paper. Simulations can be used to translate the results of animal studies to human injury risk curves and to guide the design of protective gear.


The work described here was supported by the Department of Defense US Army Research Office through the Institute for Soldier Nanotechnologies (contract number W911NF-13-D-0001) and Program Executive Office Soldier Protection and Individual Equipment.


Bauman RA, Ling G, Tong L, Januszkiewicz A, Agoston D, de Lanerolle N, Kim Y, Ritzel D, Bell R, Ecklund J, Armonda R, Bandak F, Parks S. 2009. An introductory characterization of a combat casualty relevant swine model of closed head injury resulting from exposure to explosive blast. Journal of Neurotrauma 26:841–860.

Jean A, Nyein MK, Zheng JQ, Moore DF, Joannopoulos JD, Radovitzky R. 2014. An animal-to-human scaling law for blast-induced traumatic brain injury risk assessment. Proceedings of the National Academy of Sciences 111(43):15310–15315.

Moore DF, Nyein MK, Jerusalem A, Noels L, Radovitzky R. 2009. Computational biology: Modeling of primary blast effects on the central nervous system. Neuroimage 47:T10–T20.

Nyein MK, Jason AM, Yu L, Pita CM, Joannopoulos JD, Moore DF, Radovitzky RA. 2010. In silico investigation of intracranial blast mitigation with relevance to military traumatic brain injury. Proceedings of the National Academy of Sciences 107:20703–20708.

Rafaels K, Bass CR, Salzar RS, Panzer MB, Woods W, Feldman S, Cummings T, Capehart B. 2011. Survival risk assessment for primary blast exposures to the head. Journal of Neurotrauma 28(11):2319–2328.

About the Author:Raúl A. Radovitzky is professor of aeronautics and astronautics at Massachusetts Institute of Technology. James Q. Zheng is director of the Technical Management Directorate and chief scientist for the project manager, Soldier Protection & Individual Equipment, Program Executive Office–Soldier, US Army. Thomas F. Budinger (NAE) is professor emeritus of bioengineering and electrical engineering and computer science, University of California, Berkeley, and radiology, University of California Medical Center, San Francisco, and senior scientist at the Lawrence Berkeley National Laboratory.