The development of medical autonomous systems in general and physiological closed-loop systems in particular requires an increased level of safety while operating in a highly uncertain and volatile environment. In order to provide an optimal therapeutic result, an adaption to the individual patient is required. This usually cannot be easily achieved due to the varying clinical environment and the limited information that is acquired during therapy. For example, the fact that not all clinics work with the same sensors leads to a context-specific availability of information.
Fraunhofer IMTE develops algorithms that enable the control of physiological parameters while guaranteeing safe operation for the patient and providing additional information to the clinical personnel. The goal is to derive a control structure that is modular, such that it is device-independent and can be adapted to the clinical context, for example by incorporating novel sensors that provide previously unknown information or by adapting the therapy in accordance with other therapeutic devices in use. This is achieved by using hierarchical control architectures which incorporate models of the patient, the device and the clinical context to optimize the therapy for the individual patient. This setup does not only guarantee the safety of the patient, but it can also provide information about the patients’ current state and predicted future to the clinician.