Computational Mechanobiology in Cartilage and Bone Tissue Engineering: From Cell Phenotype to Tissue Structure - Computational Modeling in Tissue Engineering - page 346

tissue formation. Due to its importance in tissue regeneration, the lattice model
approach has been extended to include angiogenesis [ 19 ] in an attempt to better
model the process of bone regeneration. Checa and Prendergast [ 20 ] closely
followed the approach of Byrne et al. [ 14 ] to investigate the effect of cell seeding
and mechanical loading on vascularisation and tissue formation inside a scaffold.
High and low loading conditions and different seeding densities were simulated
and homogeneous seeding compared to only peripheral seeding. It was predicted
that reducing the MSC seeding density from 1 to 0.5% greatly improved vascular
infiltration and bone tissue formation, whereas under high seeding density
conditions only peripheral vascularisation and ossification occurred. Likewise,
peripheral seeding was predicted to allow scaffold core vascularisation and sub-
sequent bone formation. While low loading conditions were generally beneficial
for bone formation, increasing the applied stress above a certain threshold value
was predicted to result in decreased vascularisation and a mainly cartilaginous
tissue.
4.2.2 Incorporating Realistic Scaffold Architectures
While the previous two studies modelled a regularly shaped scaffold, an irregularly
shaped scaffold inside a loaded in vivo bone chamber was implemented on the
lattice level by Khayyeri et al. [ 60 ]. Its geometrical representation was based on
lCT scans of a highly porous collagen-glycosaminoglycan (collagen-GAG)
scaffold. On the level of the regularly structured finite element mesh material
properties were homogenised by a rule of mixtures approach using the volume
fractions of tissues and scaffold material present in the element's lattice. This
approach allows for an easy implementation of the scaffold geometry into the
lattice. It is computationally efficient and hence suitable for larger scaffold
structures at the expense of micromechanical accuracy of the solution due to the
application of the rule of mixtures. The effect of a varying scaffold stiffness on
tissue differentiation patterns was investigated. Soft scaffolds were predicted to
result in mainly fibrous tissue formation while increasing stiffness values of up to
1 GPa increased the amount of cartilage and bone present in the bone chamber
significantly.
A micromechanically more accurate assessment of biophysical stimuli in a lCT
based model of an irregular scaffold is necessarily computationally very expensive
since a large number of elements is required to mesh the scaffold structure (and the
pore space). This approach was followed in Sandino et al. [ 99 ] for calcium
phosphate (CaP) and glass scaffolds by performing a solid analysis of applied
compression and a steady-state Newtonian fluid flow analysis to assess stress,
strain, fluid pressure, fluid velocity and flow induced surface shear stress in the
scaffold during early bioreactor culture. The heterogeneity was not only reflected
in the strain pattern, but also in the fluid velocities: Some pores were never
perfused despite sufficient interconnectivity while in others the fluid velocity was