Computational skeletal muscle models for signal identification and sensor development.
Investigators: Prof. Röhrle, Prof. Haasdonk, A/Prof. Sagar
The main aim of this project will be to generate a realistic in silico testbed for sensors detecting electrical signals or physiological data such as surface EMG or ion concentrations within a muscle. The testbed shall be a part of a musculoskeletal system. Initially, chemo-electro-mechanical skeletal muscle models are used to determine virtual EMG data based on a realistic part of the musculoskeletal system and a realistic (motor unit) recruitment model. Further, based on the chemo-electro-mechanical models, one can simulate the flux of ions crossing the membrane. Based on this flux, on can compute the induced a magnetic field by means of Maxwell’s equation. The in silico data can then subsequently be used as a gold standard to develop new inverse methods leading to improved signal identification, to investigate sensor placement, or to test new sensor concepts. Particularly the ability to determine the respective signals or data during contraction is unique and of extreme value. The model-inherent drawbacks of considerable computing times shall be tackled within this project by closely collaborating with USTUTT-2, which is focusing on novel model-order-reduction techniques.