Nonlinear modelling of renal blood flow of kidneys

The present study proposes a novel method of modelling the response of the renal blood flow (RBF) of rabbit kidneys to sympathetic nerve activity using polynomial nonlinear autoregressive with exogenous input (NARMAX) model. The nonlinear models were mapped into frequency domain using harmonic probing technique to get linear and nonlinear frequency response functions which are invariant descriptor of the system. The nonlinear model has been shown to be effective in accurately representing RBF dynamics compared to linear models. The effectiveness of the proposed modelling methodology is demonstrated by fitting models to different sets of RBF data collected from several rabbits during infusion of vasoconstrictors and vasodilators that decrease and increase RBF respectively. Frequency domain analysis of the fitted models convincingly demonstrate that similarities exist in the renal dynamics amongst different rabbits and that these are not altered by infusion of vasoconstrictors or vasodilators.