Ramos Carlos, Nazeran Homer, Goldman Michael D, Diong Bill
Electrical and Computer Engineering, the University of Texas at El Paso, TX 79968, USA.
Annu Int Conf IEEE Eng Med Biol Soc. 2010;2010:548-52. doi: 10.1109/IEMBS.2010.5626522.
Recent attempts at estimating the parameters for respiratory impedance models from data obtained by Impulse Oscillometry (IOS) have come across difficulties when using the well-established Mead's model of human respiratory impedance. Unconstrained optimization of this model often yields values of chest wall compliance (C(W)) and lung compliance (C(l)) too large to be physiologically feasible. We hypothesize that IOS volume displacements are inconsequential to the lung tissue and chest wall due to the small contributions of these displacements relative to lung capacity. In order to explore the validity of this hypothesis we performed a detailed analysis of Mead's impedance model. The IOS input flow signal was approximated by using a combination of typical waveforms, this signal was then used to excite Mead's electrical circuit model of the respiratory impedance with physiologically realistic parameter values estimated using data obtained from one normal adult, ten adult patients with Cystic Fibrosis, ten patients with Asthma and ten normal children, with focus on normal adult data. Pressure waveforms, energy and integrated pressure values were then obtained and compared at different points of interest in the model. This investigation suggests that the pressures "felt" by the lung tissue and chest wall are too small to have a noticeable effect on them therefore making those particular circuit elements unnecessary when the respiratory system is subject to small displacement volumes such as those used in Impulse Oscillometry. Furthermore, we believe that the very large parameter values often obtained with unconstrained optimization of Mead's model are evidence that C(l) and C(w) could be "shorted-out" when modeling IOS data.
近期，在通过脉冲振荡法（IOS）获取的数据来估计呼吸阻抗模型参数时，若使用成熟的米德人体呼吸阻抗模型，会遇到困难。对该模型进行无约束优化时，常常会得出胸壁顺应性（C(W)）和肺顺应性（C(l)）的值过大，以至于在生理上不可行。我们推测，由于IOS的容积位移相对于肺容量的贡献较小，所以对肺组织和胸壁来说无关紧要。为了探究这一推测的正确性，我们对米德阻抗模型进行了详细分析。通过组合典型波形来近似IOS输入流量信号，然后用该信号激励米德呼吸阻抗的电路模型，模型中的生理现实参数值是根据一名正常成年人、十名成年囊性纤维化患者、十名哮喘患者和十名正常儿童的数据估算得出的，重点是正常成年人的数据。接着，在模型中不同的感兴趣点获取并比较压力波形、能量和积分压力值。这项研究表明，肺组织和胸壁“感受到”的压力过小，对它们不会产生显著影响，因此当呼吸系统受到如脉冲振荡法中使用的小容积位移作用时，那些特定的电路元件就没有必要了。此外，我们认为，对米德模型进行无约束优化时常常得到的非常大的参数值证明，在对IOS数据建模时，C(l)和C(w)可能会被“短路”。
