Gelatin phantoms are usually used as a surrogate for real tissue samples in our experiments.
Our aim is to derive a viscoelastic material model, capable of handling various gelatin concentrations (tissue compositions), through indentation experiments and a model-based machine learning algorithm, which also generalizes to more complex deformation states and unknown gelatin concentrations.

Therefore, we perform several indentation experiments with varying phantoms to detect the stress-strain curves and generate our own database. We use a robot to move the indentor with different speeds into the phantoms. A force sensor mounted between the indentor and the robot detects the occurring forces over time.