Dynamical system modelling to discriminate tissue types for bipolar electrosurgery

Electrosurgery uses electric current to heat, coagulate and ligate tissue. The current flow induces a voltage due to the bio-impedance of the tissue. The flow of the current should be controlled precisely to avoid tissue damage by overheating. The control requires the proper mathematical formulation of the current and voltage relationship. This mathematical description can be explained as a dynamical system. It is composed of linear time invariant (LTI) system and/or the static nonlinearity. The in-depth understanding of the system behaviour requires parameter estimation of the LTI block and the nonlinearity. It is often done stepwise by identifying the best linear approximation (BLA) to fit the whole combination of LTI block and nonlinearity. In this paper, we study the potential discrimination of a block-oriented model between femoral, mesentery, renal and tendon tissue. The estimated system fits up to 99.84% accurately with the measured signal after optimization for the electrosurgery data of Covidien Inc. The parameters exhibit that a weak nonlinearity exists in the measurement data. Finally, the fractional order LTI systems are explored to cope with the diffusion due to the generated heat.