SCIB HeadBrain Work Group University of Pennsylvania

1
SCIB Head/Brain Work GroupUniversity of
Pennsylvania
David Meaney and Susan Margulies
2
Southern Consortium for Brain Injury
BiomechanicsOverall Objective

• The primary objective of the consortium is to
  generate the scientific foundation for a new
  method of predicting the probability of brain
  injury, given an applied input loading condition
  to the head

3
University of PennsylvaniaOverall Objective

• To acquire experimental data for developing and
  validating the finite element model used in the
  NHTSA SIMon program
• Brain Injury Thresholds
• Boundary Conditions
• Material Properties of Brain Tissue
• Cell based models of Brain Tissue

4
Brain Injury Thresholds Progress

• To determine if a specific and serious injury,
  the breakdown of the blood brain barrier, can be
  predicted by the mechanical response of a simple
  finite element model of the rat brain
• DCD model and data obtained - completed
• idealized FEM development - completed
• injury threshold derivation - completed
• development of FEM - completed at Wayne State

5
Brain Injury ThresholdsOutcome

• A simple animal model is used to study the
  threshold for brain injury (contusion) - examine
  the usefulness of different predictors for injury
• The FEM is used to study other experimental TBI
  models (cortical impact) with other SCIB members
• Continued cross-correlation of experimental
  models - one threshold does not seem to
  universally match

6
University of PennsylvaniaOverall Objective

• To acquire experimental data for developing and
  validating the finite element model used in the
  NHTSA SIMon program
• Brain Injury Thresholds
• Boundary Conditions

7
Skull-Brain Boundary Conditions Specific Goals

• To measure the relative motion of the brain
  within the skull to provide more validation data
  for the SIMon model as it begins to transform
  from a research tool into a technology that is
  the basis for a new head impact protection
  standard.

8
Experimental Methods
SPAtial Modulation of Magnetization
High Resolution T1-weighted Image
N14 sets (from 11 subjects)
N47 images from 5 subjects
9
Postures and Motion Sequences
Extension supine (ES)
Neutral supine (NS)
Flexion supine (FS)
Neutral prone (NP)
Flexion prone (FP)
10
Boundary Conditions Progress Outcomes

• Progress
• SPAMM - completed. paper published
• High Resolution T1 - completed. manuscript
  published
• Outcomes
• Cerebellum rotates 1-4 brainstem moves 1mm
  out of the foramen magnum during moderate neck
  flexion
• Foramen magnum should NOT be modeled as a no-slip
  condition

11
University of PennsylvaniaOverall Objective

• To acquire experimental data for developing and
  validating the finite element model used in the
  NHTSA SIMon program
• Brain Injury Thresholds
• Boundary Conditions
• Material Properties of Brain Tissue

12
Brain Material Properties Specific Goals

• To provide biofidelic mechanical responses of the
  living brain that can be incorporated in finite
  element models of the head (like SIMon).
• In vivo, in situ, in vitro indentation tests
• Finite shear properties of Brainstem

13
In Vivo Testing
Mechanical properties are determined by quickly
indenting the exposed brain surface to a depth d
with a rigid indentor while recording force P(t)
and displacement d on computer. Assuming that the
brain is a linear viscoelastic half-space, the
shear modulus G can be calculated from the
relationship
Lee ? Radok (1960) The contact problem for
viscoelastic bodies. J. Appl. Mech., 27 438-444
14
Comparing in vivo, in situ, in vitro
IN VIVO testing
Sodium pentobarbital overdose
IN SITU testing
IN VITRO testing
15
Major Findings

• Preconditioning did not have a significant effect
  on the living brain, but significantly reduced
  the shear moduli in vitro.
• Shear moduli in vitro are consistently lower than
  in situ (likely due to skull confinement).
• The same boundary conditions exist in vivo and in
  situ, and there were no significant differences
  between these two comparable conditions.

Overall, we conclude that the mechanical behavior
of a living brain is similar to that of a dead
brain
16
Brain Material Properties Specific Goals

• To provide biofidelic mechanical responses of the
  living brain that can be incorporated in finite
  element models of the head (like SIMon).
• In vivo, in situ, in vitro indentation tests
• Brainstem undergoing large strains

17
Brainstem properties (Finite shear)

• 4-week porcine brainstems (N15)
• 3 specimens per subject

parallel or perpendicular
cross-sectional
18
Methods Brainstem Testing in Simple Shear

• Simple Stress Relaxation
• Shear strains 50, 40, 30, 20, 10, 5, 2.5, then
  50
• Two preconditioning runs

Arbogast et al., J Biomech. 1997
19
Results
Anisotropic strain energy functions for
instantaneous response
WWmatrix Wfiber
G0 initial shear modulus of matrix-controlled
material g1, g2 relative shear relaxation moduli
t1, t2 characteristic times q stiffness
of axonal fibers
20
Brainstem Finite Shear Properties

• Parameters fit simultaneously to 50 test data in
  all 3 directions predicted data from all other
  strains well (average R2 85)
• Fiber stiffness at finite deformation is nearly
  10x stiffer than brainstem matrix - previously
  we reported fiber 3x stiffer than matrix at 2.5
  strain Arbogast and Margulies, 1999
• Brainstem matrix component is the most compliant
  brain tissue region with a instantaneous shear
  modulus of 12.7 Pa. Previously we reported
  average cerebrum tissue modulus of 526.9 Pa
  Prange and Margulies, 2002

21
Brain Material Properties Progress

• To provide biofidelic mechanical responses of the
  living brain that can be incorporated in finite
  element models of the head (like SIMon).
• In vivo, in situ, in vitro tests - completed,
  paper published
• Brainstem - completed studies, manuscript
  submitted

22
Cell-based Models of Brain TissueSpecific Goals

• To provide approximations of the variation in
  cellular strains that occur within the brain
  during impact, estimates that can be incorporated
  in finite element models of the head (like SIMon).

23
Studying cellular kinematics in vitroOrganotypic
Cultures

• Use P4-P6 rat pups
• Transverse sections 350 mm thick
• Cultured on 0.005 think laminin treated silastic
  membranes
• Fed 3X per week with Neurobasal-A supplemented
  with B27, glucose, and L-glutamine
• Incubated for 10 days on a rocker (2 rocks per
  minute)

24
Tracking bead/nuclei movement in gels and cultures
Silicone gel or Organotypic tissue
Labeled beads or nuclei
Deformed Substrate
Microscope Objective
25
Approximating the constructs and tissue
mm
mc
RVE
26
Assumptions

• Cells follow hookean-type elastic behavior
• Mechanical properties of the CNS matrix are
  similar to other soft tissues

27
Nodal equation and solution
28
The fraction of moving nuclei in organotypic
tissue increases with applied strain
29
Predicted coefficient of variance indicates cells
are much stiffer than extracellular matrix
Non-linear matrix KcellKmatrix 16 sKcell.2
Applied strain
30
Softening of brain tissue at finite strainsAre
the cells softening?

• Measured under simple shear conditions
• Repeatable behavior
• Prevents use of linear formulations
• Approximate 35 decrease in tangential stiffness
• Mechanism cytoskeletal re-alignment/failure?

Prange and Margulies, 2002
31
Brain Injury ThresholdsMajor Outcome

• Thresholds for brain injury can be predicted with
  finite element approaches, but the
  cross-correlation of predictions among animal
  models continues
• Brainstem motions indicate a mobile/free boundary
  condition at the foramen magnum
• Brain properties measured from in vitro testing
  are reasonable approximations of the in vivo
  properties
• Composite models suggest a very soft
  extracellular component integrated with stiffer
  cellular components

32
Brain Injury ThresholdsFuture Directions

• Continue the development of animal models - in
  vivo imaging to track cell motion during impact
• Material property testing - influence of age
• Using material properties and finite element
  models - new physcial model validation studies to
  confirm model predictions
• Developing better transformations between
  individual cell types and tissue - are individual
  cell populations at risk

33
Acknowledgements

• Southern Consortium for Injury Biomechanics
• Ashton Foundation