A neuro-fuzzy approach for predicting hemodynamic responses during anesthesia

The effect of drugs´ interaction on the hemo-dynamic variables is of great importance when considering patient´s safety and stability. It is also important for control infusion systems during anesthesia. In this article, an adaptive-network fuzzy inference system is used to model the effect of two drugs (propofol and remifentanil) on the mean arterial pressure and heart rate. The clinical data of 45 patients is used to train and test the model. The use of subtractive clustering improved the model performance on the testing data set. The fuzzy model is able to capture the synergistic interaction between the two drugs, but other influences were detected.