Simulation of upper airway flow-limitations in sleep therapy bench testing.

Abstract

Sleep therapy devices, such as Continuous Positive Airway Pressure (CPAP) devices, support and monitor patients with sleep disordered breathing. The assessment of their efficacy typically relies either on time-consuming clinical studies, whose reproducibility in real patients is challenging, or on bench testing designed to reproducibly simulate patient behavior. The purpose of this study is to address the limitations of existing approaches of simulating upper airway behavior for testing sleep therapy devices by introducing a dynamic valve system, governed by a mathematical resistance model. This approach enables real-time adjustment of upper airway obstruction levels using the parameter critical pressure. It allows for a high degree of control over the severity of airflow-limitations, ranging from no flow-limitation to complete obstruction. Additionally, by adjusting the model's parameters it is possible to flexibly simulate various flow-limitation waveforms observed in real patients, reflecting their anatomical variations. Compared to the Starling resistor, which is the current state-of-the-art method for upper airway simulation, the proposed approach improves reproducibility by reducing dependence on material properties and eliminating aging effects associated with the resistor's elastic segment. Bench validation against computer simulations yielded a normalized root mean square error below 4%, with recorded patterns faithfully replicating flow limitation observed in real patients. Its rigorous mathematical framework and modest hardware requirements make the system readily accessible for widespread use. In conclusion, this work provides a more robust, flexible, and comprehensive platform for evaluating sleep therapy devices.