Our main goal is to use computer modelling methods (e.g. tetrahedra, implicit surfaces, particle system) to obtain realistic models of internal human organs. Amongst other things, this would allow to estimate irradiation errors due to organ movements in cancer therapy with ionising radiation. But our approach may be applied to any domain requiring physically-based modelling. As an application, we gather medical, physical and computing data to develop a personalised simulation of moving and deforming organs, with special interest on breathing simulation. The shape of the lungs and neighbouring environment is based on segmented computer tomography (CT) scan sections performed while the patient's respiration is blocked. After this step, the data are meshed with any model able to handle dynamic alterations. The simulation consists in applying forces due to the organ activity. Most often, the internal (organ deformation) or external (collision between organs) interactions are implemented using a spring or the Lenard-Jones potential. To perform realistic simulations, it is necessary to understand the relation between parameters of the potential and the physical physiological and/or biomechanical properties of the material under studies. This will permit aswell to incorporate heterogeneity and anisotropy in the environment. Keywords: physically based modelling, soft organs, movement and deformation tacking