Pressure regulation by control of peristaltic and centrifugal pumps in heart-lung machines

The Heart-Lung Machine (HLM) is a device used during Cardiopulmonary Bypass (CPB) that temporarily takes over the functions of the heart and lung. During CPB, this device removes blood from the body, removes carbon dioxide and oxygenates it, and pumps it back into circulation. While oxygenating and returning the blood to circulation, it also provides continuous monitoring of many physiological parameters such as blood pressure, flow rate, blood temperature and oxygen saturation. The control of arterial blood pressure during CPB is a critical factor that directly affects patient morbidity and mortality. High arterial pressure can significantly increase the risk of cerebral hemorrhage and neurological complications. As part of this thesis, an algorithm has been developed to monitor and regulate the patient's blood pressure during surgery in Heart-Lung Machines. The algorithm aims to reduce the pressure to the target value if the regulation level is exceeded and to restore the pump speed to its pre-regulation level if the pressure falls below the regulation threshold. Accordingly, the speeds of the peristaltic (roller) and centrifugal pumps are controlled. In the initial phase of the study, the operating principles of similar Heart-Lung Machines were examined, and system requirements were identified through discussions with healthcare professionals. Based on this requirements analysis, filtering methods were applied to stabilize the pressure data and ensure the stability of the control algorithm. The filtered pressure data were fed into the control algorithm as input. For the RPM (Revolution Per Minute) values obtained from the algorithm output and the filtered pressure data, a proximity algorithm was developed to prevent instabilities and pressure fluctuations that might occur during the control process. In another phase of the study, experimental analyses of the system were conducted, and based on the obtained data, a fundamental model reflecting the system behavior in pressure regulation was created. Polynomial fitting was applied to the peristaltic pump speed and pressure data obtained at fixed clamp opening ratios applied to the tubing, and a function was developed to calculate the pressure value when using RPM and clamp opening ratios. The model created using this function was tested both in open-loop and closed-loop with a PI controller. The findings of the study showed that the developed control algorithm effectively controlled the pressure for both pump types, and that the centrifugal pumps provided a more stable result in pressure regulation. The effects of the mechanical structure of the peristaltic pumps on the pressure data and the control algorithm were also evaluated.