Abstract: Creating virtual bodies of real patients and using them for diagnosis and treatment planning offer the potential to further empower clinical decision making by medical experts. Virtual patient modeling allows to examine the mechanical and physiological conditions under which articulations are operating in a variety of activities without putting the patient in hazard. The continuous scientific progress has led to an increased range of musculoskeletal data and knowledge being available, covering multiple scales of the musculoskeletal system. A fuller integration of these modalities can broaden the scientific basis of virtual articulation modeling in patients, but poses challenges for data fusion and coupling of simulations. Here, we present a multimodal strategy to compose virtual models of the knee articulation based on a complementary spectrum of data that enables simulations on different scales.