Estimation of force and torque development based on dynamic muscle properties
Figure 1: Development of a testing environment for muscle assessment bases on FEM simulation.
Figure 2: Combination of theoretical knowledge, modelling and measurements within a test-bench to establish an approach for force / torque estimation of muscles.
Contact persons
Project Description
Especially elderly people are suffering from a lack of mobility, typically resulting from ongoing age-related muscular atrophy. To relive this process, early application of countermeasures such as muscle development by training have proven as a valuable tool in clinical practice. Nevertheless, currently there is no feasible tool to assess the muscle's force / torque development, which is solely based on physiologic and dielectric tissue properties. However, such an approach has the potential to provide guidance for determination of muscle health, the best level of support provided by an external thera-peutic device like an exoskeleton or similar.
One way to establish such a force / torque estimator is the algorithmic fusion of different measuring modalities, assessing indicators of physiological, morphological and metabolic aspects during muscle activity. At the present stage of research, it seems realistic that the algorithmic fusion of surface Electromyogram (sEMG) and Electrical Impedance Myography (EIM) may solve this task.
Project Goals
The starting point of our investigation is a comprehensive analysis of the physiological, morphological and metabolic processes, involved within the process of muscle activity and their impact onto the measured EIM and EMG signals. Here, application of feasible source separation strategies shall provide deeper insights into the superimposed processes shaping these signals, later condensed within feasible models approximating the dynamic process of muscle contraction and relaxation.
To settle this goal, anatomically correct Finite Element Models (FEM) representing the human extremities are designed and consequently validated by measurements within a mechanical test bench. Hence, establishing a test environment, quantifying the impact of electrode placement along the extremities and variation of dielectric and geometric muscle properties during differing movement patterns. For illustration, see figure 1.
The applied patterns are constructed in such a way that mostly stimulate one specific muscle group, later combined to multiple simulations. Based on these measurements, features related to force development will be defined for both modalities. The most sensitive of these later forms the inputs of an algorithmic fusion approach, establishing force / torque estimation during muscle activity, see figure 2.
Considering the morphological complexity of human tissues, our research is currently restricted to the large muscle groups of the lower and upper extremities. For instance, the quadriceps of the upper leg and the biceps of the upper arm. Despite, it is planned to transfer the concepts to other muscle groups.