Ulrich Großmann

An approach to reliable motion artifact detection for mobile long-term ECG monitoring systems using dry electrodes

The authors present a research project aiming to develop a context-aware cardiac permanent monitoring system. A significant challenge addressed by this project is to acquire reliable ECG signals with dry electrodes in order to do an automatic analysis und alerting. A method to detect motion artifacts by the simultaneous measurement of electrode/skin-impedance is proposed. Investigations of the correlation of motion artifacts and impedance signal are presented. Results of applying an impedance signal based artifact detection algorithm on QRS detection is shown.

ECG electrodes for a context-aware cardiac permanent monitoring system

The authors present a research project aiming to develop a context-aware cardiac permanent monitoring system. A significant challenge addressed by this project is to provide ECG electrodes that are suitable for permanent use. Because traditional adhesive gel electrodes seem not to be appropriate, two different approaches for electrodes are investigated: dry electrodes and capacitively coupled electrodes. These types of electrodes have been compared, considering signal quality, necessary analog front-ends, motion artifacts, usability for patients and possible implementations of these electrodes. Methods for artifact detection and artifact supression, which can be applied to these electrode types, are presented.

Obtaining Energy Expenditure and Physical Activity from Acceleration Signals for Context-aware Evaluation of Cardiovascular Parameters

This work presents the design and development of an online daily-life activity measurement system. This system has been conceptualized to be used along with other vital parameter sensor-systems, e.g. blood-pressure and electrocardiogram (ECG), to provide the necessary context information for the evaluation of the health status of cardiovascular risk patients who are not hospitalized, but must be permanently monitored during their daily routines. The activity and energy expenditure are captured and estimated from accelerometers, which are placed on different points of the body. The activity, the ECG and the blood pressure are sent to a base station (smart phone or a PDA) and from there to a data base, to which the physicians have access. Thus it is possible to continuously analyze the vital data of a cardiovascular patient taking into consideration the activity or physical strain.

Physical-Activity as Context-Information for Long-term Monitoring of Cardiovascular Diseases

Cardiovascular diseases are the leading cause of death and disability throughout the world and standard treatment in hospitals or rehabilitation centers remains one of the biggest cost-factors in healthcare. Therefore, the prevention and the secondary-prevention of cardiovascular illnesses are of paramount importance for the entire healthcare system. One possible method of prevention is the continuous telemonitoring of cardiovascular patients. Hence, the authors introduce a research project with the aim of developing a context-aware cardiovascular long-term monitoring system. The system should enable continuous patient-friendly measurements of blood-pressure and electrocardiogram (ECG). Furthermore, the system should allow an objective evaluation of the patient’s physiological parameters taking into consideration context information, such as individual activities and stress conditions. This paper places emphasis on the extraction of context-information, whereas for the evaluation of ECG and blood-pressure, the most significant factor is the physical activity. The patient’s activity or movement can be acquired from the three dimensional acceleration of the body caused by each movement. Thus, an inertial micro-sensor system with a triaxial accelerometer is presented. From the acceleration signals, different methods of movement detection and classification such as Nearest Neighbor rule, K-Nearest Neighbor rule and Neuro-fuzzy classification are investigated. Using the adaptive neuro-fuzzy inference method, an online activity recognition system has been implemented and its accuracy is shown.