Biomedical Engineering Reference
In-Depth Information
Fig. 27.3
Methodology framework for comparative analysis of the
in silico
and
in vivo
studies on
the organ level. Reprinted with permission from Santos et al. (
2010
)
with clinical DXA, and found good agreements in both quantitative and qualitative
results. The latest of the validated studies (Santos et al.,
2010
) was based on the
patient-specific model for the prediction of BMD (Fig.
27.3
). When quantitatively
compared to the DXA derived results for the normal, osteopenic, and osteoporotic
bone, the maximum discrepancy between the
in silico
and
in vivo
measurements
was only 3.92 %. In addition, the authors report, that comparison with the clinical
data has helped them improve the model by selecting the parameters that lead to the
biologically relevant results.
Fracture healing is another area of interest that borders on the cellular and or-
gan levels, in that it focuses on an event associated with the cellular level, such
as sheer forces or angiogenic processes, but the algorithm is still based on a con-
tinuum assumption. One of the first computational models attempting to simulate
tissue differentiation during fracture healing was based on the biphasic poroelastic
FE algorithm that started at granulation and traced the process all the way to bone
resorption (Lacroix and Prendergast,
2002
). The model was validated against histo-
morphometric data from literature, with different fracture gap sizes. The validation
confirmed that the proposed mechanobiological model produced realistic results for
different gap sizes and loading magnitudes on the rate of reduction in interfragmen-
tary strains. Isaksson and colleagues have performed a comparative review of the
existing approaches, and determined that deviatoric strain is the most significant
parameter for the modeling of tissue differentiation (Isaksson et al.,
2006
). Unfor-
tunately, since fracture healing is an inflammatory time-dependent process that is
difficult to monitor, this investigation relied only on previous reports for validation.