教授 博士生导师 硕士生导师
主要任职: 医学部党委书记兼常务副部长
性别: 男
毕业院校: 复旦大学
学位: 博士
所在单位: 生物医学工程学院
学科: 生物医学工程
电子邮箱: krqin@dlut.edu.cn
开通时间: ..
最后更新时间: ..
点击次数:
论文类型: 会议论文
发表时间: 2021-06-05
页面范围: 1512-1517
摘要: The mechanical rotary blood pumps (RBP) have been designed to provide sufficient physiological perfusion and prevent ventricular collapse caused by suction events for the congestive heart failure (HF) patients. Various control systems could be one of the most crucial objectives for developing RBPs. Some control algorithms need implantable sensors, but they are unreliable due to sensor error, pump thrombus, or short lifespan, etc. Other sensorless control algorithms can eliminate any unreliable sensor but to date those strategies have not been successfully incorporated into the RBPs in clinical. To overcome the above limitations, in this study a new physiological control strategy for RBPs has been proposed. This method depends on a recently developed implantable wireless sensor called CardioMEMS, which can provide hemodynamic information for pulmonary artery pressure (PAP). In this algorithm, a gain-scheduled proportional-integral controller is used to maintain the actual mean PAP measurements close to a user-defined threshold to provide sufficient physiologic perfusion. The algorithm is tested in-silico with (1) reference mean PAP (MPAP) of 17 mmHg and the threshold of MPAP as 15 and 19 mmHg due to +/- 2 mmHg sensor drift during rest and exercise conditions with normal and increased pulmonary vascular resistance (PVR); (2) physiological state quickly changes between rest and exercise with normal PVR for (1); and (3) actual MPAP measurements with 5% and 10% uniformly distributed noise for (1) and (2). Simulation results demonstrated that with the CardioMEMS sensor the proposed control algorithm can achieve adequate physiologic perfusion with satisfied efficacy and robustness.