Piotr Augustyniak

Seamless Tracing of Human Behavior Using Complementary Wearable and House-Embedded Sensors

This paper presents a multimodal system for seamless surveillance of elderly people in their living environment. The system uses simultaneously a wearable sensor network for each individual and premise-embedded sensors specific for each environment. The paper demonstrates the benefits of using complementary information from two types of mobility sensors: visual flow-based image analysis and an accelerometer-based wearable network. The paper provides results for indoor recognition of several elementary poses and outdoor recognition of complex movements. Instead of complete system description, particular attention was drawn to a polar histogram-based method of visual pose recognition, complementary use and synchronization of the data from wearable and premise-embedded networks and an automatic danger detection algorithm driven by two premise- and subject-related databases. The novelty of our approach also consists in feeding the databases with real-life recordings from the subject, and in using the dynamic time-warping algorithm for measurements of distance between actions represented as elementary poses in behavioral records. The main results of testing our method include: 95.5% accuracy of elementary pose recognition by the video system, 96.7% accuracy of elementary pose recognition by the accelerometer-based system, 98.9% accuracy of elementary pose recognition by the combined accelerometer and video-based system, and 80% accuracy of complex outdoor activity recognition by the accelerometer-based wearable system.

Content-adaptive signal and data in pervasive cardiac monitoring

Although the pervasive monitoring of patients at cardiac risk is widely recognized as a valuable tool, the content-dependent signal and data adaptability is rarely considered. Consequently the autonomy of wearable recorders is affected by the unnecessary computation and the wireless data carrier is wasted by the unwanted information. The thorough analysis of human ECG interpretation process and detailed studies of diagnostic procedures reveal that the adaptation of the perceptual strategy to the ECG content is an intuitive and common practice of cardiology experts. Considering human interpretation-derived directions we re-arranged a typical machine interpretation software to simulate the human reasoning. Several aspects of data are adjusted accordingly to the automatic rough estimate of the record contents: the interpretation process flow, the result priority, report content and frequency and the local sampling frequency of the ECG included

Time-frequency modelling and discrimination of noise in the electrocardiogram

In widely spread home care applications of ECG recorders, the traditional approach to the problem of noise immunity is no longer sufficient. This paper presents a new ECG-dedicated noise removal technique based on a time-frequency noise model computed in a quasi-continuous way. Our algorithm makes use of the local bandwidth variability of cardiac electrical representation and splits the discrete time sequence into two sub-planes. The background activities of any origin (muscle, power line interference, etc) are measured in the regions of the time-frequency plane, situated above the local bandwidth of the signal. The noise estimate on each particular scale is non-uniformly sampled and needs to be extrapolated to the regions where the components of cardiac representation are normally expected. On the lower scales, the noise contribution is computed with the use of square polynomial extrapolation. The time-frequency representation of noise, partially measured and partially calculated, is arithmetically subtracted from the noisy signal, and the inverse time-frequency transform yields a noise-free cardiac representation. The algorithm was tested with the use of CSE database records with the addition of MIT-BIH database noise patterns. The static and dynamic performance of the algorithm is sufficient to ameliorate the signal-to-noise ratio by more than 11 dB.

Monitoring activities of daily living based on wearable wireless body sensor network.

With recent advances in microprocessor chip technology, wireless communication, and biomedical engineering it is possible to develop miniaturized ubiquitous health monitoring devices that are capable of recording physiological and movement signals during daily life activities. The aim of the research is to implement and test the prototype of health monitoring system. The system consists of the body central unit with Bluetooth module and wearable sensors: the custom-designed ECG sensor, the temperature sensor, the skin humidity sensor and accelerometers placed on the human body or integrated with clothes and a network gateway to forward data to a remote medical server. The system includes custom-designed transmission protocol and remote web-based graphical user interface for remote real time data analysis. Experimental results for a group of humans who performed various activities (eg. working, running, etc.) showed maximum 5% absolute error compared to certified medical devices. The results are promising and indicate that developed wireless wearable monitoring system faces challenges of multi-sensor human health monitoring during performing daily activities and opens new opportunities in developing novel healthcare services.

Wearable Patient Home Monitoring Based on ECG and ACC Sensors

Advanced developments in existing information technologies are extending the range of treatment and health services available at patients home. This paper proposes a mobile monitoring system that integrates wearable ECG and ACC mobile sensors. We present an algorithm determining the correlation between the heart rate and movements based on automatic analysis of ACC and ECG signals. The system was tested on seven healthy adults who were asked to perform normal daily activity. As a result, we examined general physi- cal state during normal daily activities of the subject, based on the analysis of sensor signals.

Ubiquitous Cardiology: Emerging Wireless Telemedical Applications

The Internet and other technological developments are now playing increasing roles in consumer health and the delivery of health services. Ubiquitous Cardiology: Emerging Wireless Telemedical Applications provides developmental solutions and explanations for cardiovascular diagnostics. Useful to field researchers, academicians, and healthcare practitioners, this Premier Reference Source presents a collection of studies on medical data redundancy, priority, and validity.

Wearable wireless heart rate monitor for continuous long-term variability studies

Prototyping of a home care system for activity surveillance and sleep assessment targeted to elderly people involves the compromise of wearing comfort and measurement performance. We propose a wearable heart rate variability monitor connected via wireless digital link to a home-embedded infrastructure of multimodal health surveillance system. The coin-size wearable recorder acquires and processes the electrocardiogram and sends annotated tachogram data accordingly to the subjects status and programed schedule. Thanks to remote programmability, in case of predefined thresholds excess, the recorder response is immediate, whereas the regular reports are organized in packets and delivered in bulk in short transmission sessions. This approach significantly reduces the data rate and the energy required to supply the communication module. The prototype weighting 11.2 g is based on the ARM7 (Atmel Corporate Headquarters 2325 Orchard Parkway San Jose, CA, USA) processor running at 18 MHz and with a 300-mA h rechargeable battery allows for up to 10 days of seamless tachogram monitoring.

Assessment of electrocardiogram visual interpretation strategy based on scanpath analysis

Automated ECG interpretation systems are supposed to follow human expert reasoning. Despite well-established standards, the visual interpretation strategy of the human is still undisclosed today. This paper presents a new approach to the interpretation process research based on eyetrack features captured from a human expert during biosignal visual inspection. This approach required a set of visual tasks consisting in ECG interpretation by volunteers of different degrees of expertise. The recorded eyeglobe trajectories were analysed in the context of medical data represented in the displayed ECG traces and revealed interesting information on diagnostic data distribution and principles of interpretation strategies. The scanpath-derived data make benefit of oculomotoric habits gathering in everyday practice, unconscious mutual perception-recognition interactions and are not affected by human memory or verbalization limits. For these reasons, they provide more objective assessment than any other method willingly controlled by the human. Besides new information about the ECG contents and quantitative descriptions of medical data distribution, our experiment reveals some eyetrack parameters as distinctive for interpretation skills estimation.

A Graph Representation of Subject's Time-State Space

Surveillance systems are currently the most developed branch of assisted living applications providing the disabled or elderly people with unprecedented security in their independent life. This paper presents a design of a telemedical surveillance system, where graph theory is used to describe subjects’ states. Patient’s states expressed by sets of medically-derived parameters and his or her daily activity (a behavioral pattern) are represented by attributed probabilistic graphs with indexed and labeled nodes. This representation provides high flexibility in a state and transient description as well as a reliable measure of behavior divergence, which is a basis for automatic alerting. The system is designed for the subject’s apartment and supports a localization-dependent definition of his or her usual and unusual behavior. The apartment topology is also represented in the form of a graph determining subject’s pathways and states. This approach has been found very flexible in all aspects of personalization, appropriate to work with the behavioral presumption set or with the auto-adaptive artificial intelligence recognition engine. Also the patient’s state, thanks to the semantic description may be easily extended or refined if necessary by adding new, complementary data capture methods.

Data Integration in Multimodal Home Care Surveillance and Communication System

This paper presents the data capture methodology and design of a home care system for medical-based surveillance and man-machine communication. The proposed system consists of the video-based subject positioning, monitoring of the heart and brain electrical activity and eye tracking. The multimodal data are automatically interpreted and translated to tokens representing subject’s status or command. The circadian repetitive status time series (behavioral patterns) are a background for learning of the subject’s habits and for automatic detection of unusual behavior or emergency. Due to mutual compatibility of methods and data redundancy, the use of unified status description vouches for high reliability of the recognition despite the use of simplified measurements methods. This surveillance system is designed for everyday use in home care, by disabled or elderly people.