Ventilator Control Based on a Fuzzy-Neural Network Approach

Respiratory systems are complex nonlinear systems that exhibit uncertain properties. Since patients' needs are not constant acceptable control of ventilator has to be adaptive to ensure that patients with severe pulmonary disease will obtain the required ventilation support. The AUTOPILOT-BT system was developed to reduce cognitive load of intensivists and at the same time improve mechanical ventilation therapy [1]. The goal of this research was to evaluate a nonlinear adaptive fuzzy-neural network controller, in which a fuzzy controller is used to control the flow and a neural network is used to identify the nonlinear respiratory system. The proposed nonlinear intelligent controller is applied to data recorded during a multicenter study on ARDS patients (adult respiratory distress syndrome). The experimental results demonstrate that this adaptive fuzzy-neural network controller has much better control performance than is obtained with traditional controllers.