Abstract-After cardiac surgery patients often show a cardiovascular instability, which has to be managed by infusing different vasoactive drugs. This work proposes a closed-loop system to automatically react upon changes in the patients state and trigger infusion of indicated drugs. As input signals mean arterial pressure, cardiac output, central venous pressure and systemic vascular resistance were chosen. The vasoactive drugs used are isosorbide dinitrate, norepinephrine, dopamine and additionally hydroxyethyl starch. The controller itself is designed as a controller system with four internal fuzzy controllers, one for each medication dosage to be calculated. The proposed controller is tested with a simulation environment. The environment is able to simulate a canine cardiovascular system and reactions to different medications. The controller is evaluated for three different diseases, for congestive heart failure, hypotension and hypertension. The fuzzy controller system is able to react adequately in all these test cases in the simulation and regulates the hemodynamic signals back within a range of the desired target values.

ZusammenfassungDie frühe Behandlung mit Extrakorporalen Kreislauf-Unterstützungssystemen bei Patienten mit kardiogenem Schock wirkt sich positiv auf den weiteren Verlauf aus und kann einem Multi-Organversagen entgegenwirken. Um eine frühe Behandlung zu ermöglichen, ist eine kontinuierlicheÜberwachung des Patienten durch geschulte Kardiotechniker notwendig. Unter Notfallbedingungen kann eine ungeteilte Aufmerksamkeit für den Patienten nicht garantiert werden womit es zu Behandlungsfehlern kommen kann.Durch AbstractPatients suffering from cardiogenic shock may benefit with an early application of a portable Extracorporeal Circulatory Support System (ECSS) preventing multi organ failure. This however requires the presence and constant supervision of the patient by trained personal at the emergency site. Under these circumstances full attention to the patient may not be guaranteed and operation errors may occur.With the automation of the portable ECSS optimal perfusion may be achieved with minimal workload for the human operator allowing the safe transportation of the patient to the hospital.The focus of this thesis is the development of an adaptive and robust control system that regulates perfusion based on online data of the patient. While the system needs to be highly dynamic, so that it is able to adapt to different situations, it must ensure maximal patient safety at all times.To develop such control system first an animal model was used to analyze the type of signals acquired during extracorporeal circulation. This information was used as a reference for the creation of a mathematical model. The model includes a cardiovascular system undergoing extracorporeal circulation, a gas exchange model and a medication model. This was integrated into a simulation system that could be used for the creation and evaluation of the designed controller.Fuzzy logic was considered as a control mechanism allowing the easy creation of rules based on the knowledge of trained perfusionists. Since patient pre-conditions and reactions will be different from one case to another an adaptive mechanism is proposed to modify the existing controller and adapt to the specific needs of the patient.A software framework was developed allowing a fast implementation of the control system. This framework was created not only focusing on the automation of the ECSS but also to serve as a basis for the development of control systems for other medical devices with similar requirements.Several simulations are presented showing the performance of the fuzzy controller with the proposed adaptive mechanism. Additional simulations show the response of the designed ECSS controller under different patient scenarios.vii Acknowledgments I want to give thanks to my advisor Prof. Dr. Alois Knoll for giving me the opportunity of working in this thesis, for his support and confidence. I would also like to thank Prof. Dr. Robert Bauernschmitt for taking the time to review my work and acting as a second advisor. From the department of Robotics and Embedded Systems ...

Abstract-For many classification or controlling problems a set of training data is available. To make best use of this training data it would be ideal to feed the data into a learning algorithm, which then outputs a finished, trained fuzzy controller, that is able to classify or control the original system. For the FUZZ-IEEE 2012 a competition was proposed to predict future volumes sold per day in a certain gas station. The training data includes a collection of gas prices at the current and the competitor's gas station and the according volume sold on every consecutive day in a period of about one year. This training data was analyzed and fit to a fuzzy learning algorithm based on the Münsteraner Optimisation System. As a base point a mean value comparison is used and then different features as fuzzy inputs are tested. Also different fuzzy set widths and and sequence of commands are compared. The final controller chosen shows promising results in the test with left out training data sets. Final results still have to be shown with the test data of the competition.

For patients suffering from cardiogenic shock cardiopulmonary resuscitation may not be sufficient to restore normal heart function. However, their chances of survival may be increased with the use of an extracorporeal support system. With this system the patient's organs are perfused while being transported to the nearest hospital for proper treatment. In the automation of an extracorporeal support system the patient's vital signals are constantly monitored and proper adjustments are performed to improve organ perfusion. In this paper, an adaptive fuzzy controller is proposed that uses the knowledge and expertise of a perfusionist as a starting point and reference for regulation. Furthermore it is able to adapt to the patient's specific reactions by manipulating the rule base of the fuzzy controller. The performance of the adaptive fuzzy controller is tested with a simulation model of the cardiovascular system.

Abstract-The automation of tasks in an intensive care unit can be beneficial for the patient's recovery and also can reduce the workload of the physicians. That is why research is looking for different tasks, which can be performed by automated machines. One specific project, our group is working on, is the automation of blood pressure regulation with drugs for patients task of the physicians, knowledge has to be gathered about, how the physicians decide on according treatment actions. As this is an everyday task, the easiest way to analyze and understand the treatment decisions is to look at taken decisions in actual patients. For this purpose a system has been set up, which can be connected to both the patient monitor and the syringe pumps, which infuse the different drugs. There were different requirements from the technical, medical point of view and also from an ethic committee, which had to be taken into account. Given those requirements hardware was chosen and suitable software was implemented. First tests showed, that the system can be operated in the intensive care unit and provide data for further research.