Application of Neural Network and Finite Element Method for Multiscale Prediction of Bone Fatigue Crack Growth in Cancellous Bone - Multiscale Computer Modeling in Biomechanics and Biomedical Engineering - page 24

Fig. 14 Averaged contour of Cr.Dn after 1.E5 loading cycles obtained using the proposed
multiscale FENN approach and corresponding Cr.Dn distribution on specific trabecular bone
specimens obtained by FE
expected to lead to more accurate prediction of damage accumulation, as suggested
by Cowin [ 8 ].
Figure 14 shows the contour of Cr.Dn after 1.E5 loading cycles obtained using
the proposed multiscale FENN approach. The trabecular bone specimen results are
averaged and passed back to the entire femur in the form of averaged Cr.Dn and
Cr.Le using the validated NN.
Observations of the Cr.Dn contour of the proximal femur reveal a very similar
distribution of BV = TV (Fig. 8 ). Keyak et al. [ 31 ] compared experimental fracture