An intelligent control strategy for high-flow nasal cannula based on electrical impedance tomography

To improve the precision and adaptability of oxygen delivery in patients with chronic obstructive pulmonary disease, this study proposes an intelligent control method for high-flow nasal cannula (HFNC) based on Electrical Impedance Tomography (EIT). First, an equivalent circuit model of the HFNC-respiratory system was constructed to represent the physiological dynamics of airflow and muscle effort, and its validity was confirmed through physical experiments. Second, a dual closed-loop control architecture was developed, incorporating real-time EIT-derived ventilation information as the outer-loop feedback and flow rate as the inner-loop control target. The system was implemented using both conventional proportional-integral-derivative (PID) and fuzzy PID algorithms. Finally, a simulated lung platform equipped with EIT monitoring was built to experimentally evaluate the flow tracking performance. The results show that the fuzzy PID controller significantly improves control accuracy and stability, reducing the flow error by 46.1% and fluctuation by 69.1% under high-flow conditions compared to conventional PID. The proposed strategy presents a dynamic, individualized approach to respiratory support, demonstrating promise for advancing precision oxygen therapy in clinical settings.