An important application of cancellous bone modeling is the early detection of osteoporosis by sonography [1]. This disease is characterized by the reduction in volume percentage of cortical bone, weakening the bone and increasing the likelihood of fractures. Many past attempts to model the problem only considered mechanical effects. Recent research laid the foundation to also include electric and magnetic effects [2]. In this talk, we present a fully coupled multiscale approach considering mechanical, electric and magnetic effects [3, 4]. The model consists of a macro- and microscale, which are connected by the finite element square method (FE2). On the microscale, we use a two-phase material model. The first phase is cortical bone, which is described as a piezoelectric and insulating solid. The second phase is bone marrow, which is modeled as a viscoelastic and conducting solid. Electric and magnetic effects are coupled via Maxwell’s equations. To compare different stages of osteoporosis, we created representative volume elements (RVEs), which differ in volume fractions of cortical bone. We present simulation results showing a drastic reduction of the magnetic field strength originating from a small mechanical perturbation for later stages of the disease, which is in conformity with experimental research. Furthermore, we show that the inverse problem – recovering the distribution of cortical bone phase from the magnetic field strength – can be solved with very high accuracies by using artificial neural networks (ANNs) [5]. REFERENCES [1] Kaufman, J. J., Luo, G., Siffert and R. S., Ultrasound simulation in bone, IEEE Trans Ultrason Ferroelectr Freq Control, 55(6), 1205-1218 (2008). [2] Gilbert, R.P., Vasilic, A., Klinge, S., Panchenko, A. and Hackl, K., Applications of Homogenization Theory to the Study of Mineralized Tissue, CRC Press (2020). [3] Blaszczyk, M. and Hackl, K., Multiscale modeling of cancellous bone considering full coupling of mechanical, electric and magnetic effects, Biomech Model Mechanobiol, 21(1), 163-187 (2022). [4] Blaszczyk, M. and Hackl, K., On the effects of a surrounding medium and phase split in coupled bone simulations, in revision (2022). [5] Stieve, V., Blaszczyk, M. and Hackl, K., Inverse modeling of cancellous bone using artificial neural networks, Z Angew Math Mech, 102(6), e202100541 (2022).