L. Jatoba

Context-aware mobile health monitoring: Evaluation of different pattern recognition methods for classification of physical activity

There are various applications of physical activity monitoring for medical purposes, such as therapeutic rehabilitation, fitness enhancement or the use of physical activity as context information for evaluation of other vital data. Physical activity can be estimated using acceleration sensor-systems fixed on a persons body. By means of pattern recognition methods, it is possible to identify with certain accuracy which movement is being performed. This work presents a comparison of different methods for recognition of daily-life activities, which will serve as basis for the development of an online activity monitoring system.

Reliable motion artifact detection for ECG monitoring systems with dry electrodes

Reliable signals are the basic prerequisite for most mobile ECG monitoring applications. Especially when signals are analyzed automatically, capable motion artifact detection algorithms are of great importance. This article presents different artifact detection algorithms for ECG systems with dry electrodes. The algorithms are based on the measurement of additional parameters that are correlated with the artifacts. We describe a mobile measurement system and the procedure used for the evaluation of these algorithms. The algorithms are assessed based upon their effect on QRS detection. The best algorithm improved sensitivity (Se) from 98.7% to 99.8% and positive predictive value (+P) from 98.3% to 99.9%, while 15% of the signal was marked as artifact. This corresponds to a decrease in false positive and false negative detected beats by 89.9%. Different metrics to evaluate the performance of an artifact detection algorithm are presented.

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.

Security for mobile low power nodes in a personal area network by means of trusted platform modules

The growing field of ubiquitous applications and the use of resource constrained mobile devices strongly demands for mechanisms to provide the security and privacy of such mobile devices. In this paper we show that especially new teletherapeutic applications are not feasible without strong cryptographic protection of data and platform. Based on the analysis of security requirements, we introduce a mobile low power node that is secured by means of a Trusted Platform Module (TPM). For privacy and security of the communication between the mobile device and a webserver, which is part of the distributed network, we propose a security protocol based on webservice technology that uses the mechanisms of the TPM. Finally, measurements that were done with the secured mobile node are presented. We show that TPMs are well suited for resource constrained mobile devices and are a step towards trusted ubiquitous computing.

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.

System for Body and Mind Monitoring in coaching process

In order to investigate the correlation between stress and cognitive performance, a mobile Body & Mind Monitoring System was implemented. This system has a modular design and contains different modules allowing a long time and noninvasive monitoring of physiological parameters of a test person in his every day life, for instance ECG, GSR, PPG, respiration and physical activity. The functionality of the created set-up was validated and clinical studies can now be conducted.

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.

Web based Teletherapy System for Telemonitoring and Remote Control of Therapeutic Devices

Health Monitoring is a growing field of research and development that gets more and more attention. Monitoring of patients at home by means of telemedical devices is a great chance to reduce costs for the healthcare system and increases the quality of life. Expanding the idea of health monitoring, a system is presented which not only monitors patient’s vital data, but also enables the physician to adapt the therapy from afar by remote controlling the therapeutic device at the patient’s site. Taking therapy of pain with infusion pumps as an example, the shortcomings of current treatment are discussed and optimizations using a teletherapeutic system are shown. The presented system consists of a webserver with webservice interface, which allows bidirectional, secure and fault tolerant communication between the physician and the devices of the patient.

Securely control Infusion Pumps via Internet for efficient Remote Therapy of Pain

Pain therapy is an important part of the current health care system and will become even more significant in future, because of the demographic changes. Though there are shortfalls in medical care of patients with pain. There are not enough specialized clinics and most of the patients do not get adequate therapy. In this paper the authors present a system that can solve these problems in pain therapy and enables specialized clinics to increase their efficiency in order to increase the capacity for pain care. The system combines telemonitoring of patients with the ability to remote control the patient’s infusion pump from afar. The physician is able to adapt the therapy of a patient without the need to visit him personally. Because of the fact, that the infusion pump applies analgesics that are able to cause vital harm, special interest is taken in the security and safety of the teletherapy system.