Scholarlink

<p>In cardiac surgeries with extracorporeal circulation, the lungs are temporarily disconnected from the body. To minimize the risk of tissue damage, ischemia or pulmonary cell necrosis, the lungs are subjected to mild hypothermia. Incorporating dynamic heat transfer models offer the potential to enhance temperature regulation through a more advanced approach. Thanks to a thermal two-port network formalism, a better comprehension of the lung global heat impedance is obtained which has enabled a frequency analysis of the lungs. This modeling approach can also be adapted to incorporate the blood perfusion, which serves as a natural temperature regulator in the human body. Such a global characterization is too complex to be implemented in real-time application, therefore a model order reduction is developed through several approximations. In addition to the physiological modeling, when the system parameters are unknown, system identification enables estimating and optimizing these parameters. An efficient online algorithm, called LMRPEM, was developed in Victor et al. (Nonlinear Dyn 110:635-648, 2022) but the estimation is carried out with the full input–output data. In an online context, the data acquisition may be long and the computation time may highly increase, and even go beyond the sampling time. Therefore, one of the main contributions of this paper is to propose a truncated-LMRPEM with an analysis of the window length to be set.</p>

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

Real-time system identification of bio-heat transfers in lungs

  • 期刊论文

  • 发表日期：

    2025年4月17日

查看全文
  • Stéphane Victor,
  • Enso Ndreko,
  • Jean-François Duhé,
  • Pierre Melchior

摘要

In cardiac surgeries with extracorporeal circulation, the lungs are temporarily disconnected from the body. To minimize the risk of tissue damage, ischemia or pulmonary cell necrosis, the lungs are subjected to mild hypothermia. Incorporating dynamic heat transfer models offer the potential to enhance temperature regulation through a more advanced approach. Thanks to a thermal two-port network formalism, a better comprehension of the lung global heat impedance is obtained which has enabled a frequency analysis of the lungs. This modeling approach can also be adapted to incorporate the blood perfusion, which serves as a natural temperature regulator in the human body. Such a global characterization is too complex to be implemented in real-time application, therefore a model order reduction is developed through several approximations. In addition to the physiological modeling, when the system parameters are unknown, system identification enables estimating and optimizing these parameters. An efficient online algorithm, called LMRPEM, was developed in Victor et al. (Nonlinear Dyn 110:635-648, 2022) but the estimation is carried out with the full input–output data. In an online context, the data acquisition may be long and the computation time may highly increase, and even go beyond the sampling time. Therefore, one of the main contributions of this paper is to propose a truncated-LMRPEM with an analysis of the window length to be set.