Although noninvasive ventilation has been increasingly used in clinics and homes to treat respiratory diseases, the problem of air leaks should not be neglected because they may affect the performance of the ventilation and even pose a threat to life. The detection and estimation of the leakage are required to implement auto-compensation, which is important in the development of intelligent ventilation. In this study, the methods of detection and estimation of the leakage were established and validated. Ventilation experiments were performed based on the established experimental platform. The air flow and pressure were detected at different locations of the airway to determine the relationship between the leakage and the other variables. The leakage was estimated using linear predictor models. The curves describing the relationships among pressure, flow and volume changed regularly with the leakage. For pressure-controlled ventilation, the leakage could be estimated by the detected peak flow and by the ventilation volume of one breathing cycle. The methods for the leakage estimation were validated. Volume-controlled ventilation was also studied. Although the leakage could be estimated using the detected peak pressure, the limitation of volume-controlled ventilation was obvious for noninvasive ventilation (NIV). Leaks could be detected and estimated using a linear predictor model via the flow/pressure curve. The use of this model is a potential method for the auto-compensation of noninvasive ventilation.

中文翻译：

---

无创通风中空气泄漏的检测和估计：平台研究

尽管无创通气已越来越多地用于诊所和家庭治疗呼吸系统疾病，但漏气问题不容忽视，因为它们可能影响通气性能，甚至对生命构成威胁。泄漏的检测和估计需要实现自动补偿，这对于智能通风的发展非常重要。在这项研究中，建立并验证了泄漏的检测和估计方法。在已建立的实验平台上进行通风实验。在气道的不同位置检测气流和压力，以确定泄漏与其他变量之间的关系。使用线性预测模型估计泄漏。描述压力之间关系的曲线，流量和体积随泄漏规律变化。对于压力控制通气，可以通过检测到的峰值流量和一个呼吸周期的通气量来估计泄漏。验证了泄漏估计的方法。还研究了容量控制通气。尽管可以使用检测到的峰值压力来估计泄漏，但容量控制通气的局限性对于无创通气 (NIV) 是显而易见的。泄漏可以通过流量/压力曲线使用线性预测模型来检测和估计。该模型的使用是无创通气自动补偿的一种潜在方法。泄漏可以通过检测到的峰值流量和一个呼吸周期的通气量来估计。验证了泄漏估计的方法。还研究了容量控制通气。尽管可以使用检测到的峰值压力来估计泄漏，但容量控制通气的局限性对于无创通气 (NIV) 是显而易见的。泄漏可以通过流量/压力曲线使用线性预测模型来检测和估计。该模型的使用是无创通气自动补偿的一种潜在方法。泄漏可以通过检测到的峰值流量和一个呼吸周期的通气量来估计。验证了泄漏估计的方法。还研究了容量控制通气。尽管可以使用检测到的峰值压力来估计泄漏，但容量控制通气的局限性对于无创通气 (NIV) 是显而易见的。泄漏可以通过流量/压力曲线使用线性预测模型来检测和估计。该模型的使用是无创通气自动补偿的一种潜在方法。泄漏可以通过流量/压力曲线使用线性预测模型来检测和估计。该模型的使用是无创通气自动补偿的一种潜在方法。泄漏可以通过流量/压力曲线使用线性预测模型来检测和估计。该模型的使用是无创通气自动补偿的一种潜在方法。