Optimization Approaches for the Design of Additively Manufactured Scaffolds - Tissue Engineering: Computer Modeling, Biofabrication and Cell Behavior - page 113

copy of real structures is difficult, time consuming and in most cases not strictly
necessary; therefore, capturing and reproducing the main key properties of natural
structures (such as porosity and pore distribution) in an effective yet simple manner
could represent a valid alternative to obtain the desired results [ 28 , 29 ].
For years, a 'trial-and-error' approach has been adopted to validate scaffold
micro-architectures, with modifications being made to an existing design on the ba-
sis of in vitro or in vivo results [ 30 ]. More recently, Finite Element Analysis (FEA)
contributed to the reduction of experimental tests and to shortening scaffold design
process. FEA has been at first used in TE for a post hoc investigation of the mechan-
ical behavior of scaffolds and to predict their interactions with the surrounding tis-
sues. For example, several finite element studies provided a computational approach
to evaluate the effects of mechanical stimuli induced by fluid flow and mechanical
loads to the cells seeded into a scaffold [ 31 , 32 ]; the computed results were used
to modify geometrical or material parameters and to choose the most suitable ones
for tissue replacement. CAD and µCT-based models of the produced porous scaf-
folds have been normally considered as input geometry. The reasonable accuracy
of simulation results, in comparison with experimental data, has led to the recent
use of FEA as a predictive tool for the apriori design and optimization of scaffold
architectures [ 33 ].
When approaching the fabrication of porous scaffolds, two main issues have to
be simultaneously addressed: one is the optimal porosity, which will depend on
the ability of hosted cells to migrate and proliferate within the scaffold; the other
one regards the capability of the produced scaffold to bear physiological loads once
implanted, without dramatic collapses [ 34 ]. The latter issue is particularly urging in
the case of scaffolds intended for the regeneration of mineralized tissue, where the
scaffold implant should provide a leading framework for the repair of the damage.
The need for an optimization problem therefore emerges: on the one hand, the ideal
scaffold should exhibit a sufficiently large porosity; on the other hand, it should
be stiff and robust enough to withstand physiological loads. Such requirements are
evidently antagonistic, since a large porosity negatively impacts robustness. For this
reason, the development of design strategies capable to optimal trade-off between
these two opposite needs is one of the main challenges of TE.
Topological optimization techniques are often used in combination with FEA in
order to obtain functional scaffold microarchitectures that maximize stiffness while
preserving constraints on the total mass budget. They aim at achieving the best use
of material within a structure subjected to either a single load or a multiple load
distribution and involve the determination of features such as location, number and
shape of “holes”, as well as connectivity of the domain, in order to compute new
microstructures that attain desired properties [ 35 ]. Thus, instead of relying on CAD
tools to design unit cell geometries, topology optimization approaches have been
used to locate void and solid spaces within the initial domain on the basis of the
desired tissue requirements. A pioneering work in this context is represented by the
study of Almeida et al. [ 36 ], who developed a tool combining CAD-based model-
ing and analytic methods to simulate and optimize scaffolds providing an apriori
control of the mechanical properties as a function of porosity and scaffold archi-
tecture. Instead of starting form a predefined unit cell, the tool starts form a dense