Real-time Health Surveillance Research Articles

3D-printed epifluidic electronic skin for machine learning-powered multimodal health surveillance.

The amalgamation of wearable technologies with physiochemical sensing capabilities promises to create powerful interpretive and predictive platforms for real-time health surveillance. However, the construction of such multimodal devices is difficult to be implemented wholly by traditional manufacturing techniques for at-home personalized applications. Here, we present a universal semisolid extrusion-based three-dimensional printing technology to fabricate an epifluidic elastic electronic skin (e3-skin) with high-performance multimodal physiochemical sensing capabilities. We demonstrate that the e3-skin can serve as a sustainable surveillance platform to capture the real-time physiological state of individuals during regular daily activities. We also show that by coupling the information collected from the e3-skin with machine learning, we were able to predict an individual's degree of behavior impairments (i.e., reaction time and inhibitory control) after alcohol consumption. The e3-skin paves the path for future autonomous manufacturing of customizable wearable systems that will enable widespread utility for regular health monitoring and clinical applications.

A vision-based fall detection framework for the elderly in a room environment using motion features and DAG-SVM

Falls are one of the major health risks for the elderly in public health care issues, thus fall detection is particularly necessary for the seniors. In this paper, we propose a novel computer vision-based fall activity detection framework which is based on a new feature extraction method of human motion. We detect falls by analyzing the motion process of human body rather than conventional human shape analysis. Approximate ellipse is first adopted to fit human silhouette, and the areas outside the ellipse are removed to confirm human body silhouette (HBS). Then HBS is divided into three regions, and the motion features are extracted from the centroid coordinate changes of each region. A directed acyclic graph support vector machine (DAG-SVM) classifier is finally applied to detect falls from normal daily activities. The advantage of the proposed method is that it can not only effectively distinguish falls from fall-like activities but also detect the falls parallel to camera optical axis. Experimental results show that our method achieves a state-of-the-art accuracy in fall detection. This method is lightweight and thus can be integrated into embedded devices for real-time health surveillance.

Evidence-based model for real-time surveillance of ARDS

Real-time health surveillance becomes important and necessary with the increase of the elderly population to preserve their quality of life. Real-time models aim to provide alerts before the severe illness occurs. Acute respiratory distress syndrome is a crucial disease of the respiratory system that threats the health of the elderly. This paper proposes a real-time model for the surveillance of ARDS based on belief functions theory. Non-invasive physiological signals are considered such as heart rate, respiratory rate, oxygen saturation and mean airway blood pressure. Different linear and nonlinear parameters are extracted from these signals; then a parameters selection procedure is performed to reduce their dimensionality. Afterwards, classifiers are constructed using parameters distributions defined in the evidence framework. Real-time prediction is then performed by combining all classifiers decisions. As results, high performances are obtained over the testing sets with performances of 77% and 71% for sensitivity and specificity, respectively.

Real-time Secure Health Surveillance for Smarter Health Communities

Pervasive healthcare services with smart decision making capability and ubiquitous communication technologies can forge future smart communities. Real-time health surveillance for early detection of life-threatening diseases through advanced sensing and communication technology can provide better treatment, reduce medical expenses and save lives of community residents (i.e., patients). However, the assurance of data privacy is the prime concern for such smart health technologies. This research aims to describe a privacy-preserving cloud-based system for real-time health surveillance through change detection of multiple vital health signs of smart community members. Vital signs data generated from IoT-enabled wearable devices are processed in real-time in a cloud environment. This article focuses on the development of a predictive model for the smart community considering the sensitivity of data processing in a third-party environment (e.g., cloud computing). We developed a vital sign change detection system using Holt’s linear trend method (to enable prediction of data with trends) where fully homomorphic encryption is adapted to perform computations on an encrypted domain that can ensure data privacy. Moreover, to reduce the overhead of the fully homomorphic encryption method over large medical data we introduced a parallel approach for encrypted computations using a MapReduce algorithm of Apache Hadoop. We demonstrated the proposed model by evaluating some case studies for different vital signs of patients. The accuracy and efficiency of the implementation demonstrate the effectiveness of the proposed model for building a smart community.

Spatial and Spatio-Temporal Log-Gaussian Cox Processes: Extending the Geostatistical Paradigm

In this paper we first describe the class of log-Gaussian Cox processes (LGCPs) as models for spatial and spatio-temporal point process data. We discuss inference, with a particular focus on the computational challenges of likelihood-based inference. We then demonstrate the usefulness of the LGCP by describing four applications: estimating the intensity surface of a spatial point process; investigating spatial segregation in a multi-type process; constructing spatially continuous maps of disease risk from spatially discrete data; and real-time health surveillance. We argue that problems of this kind fit naturally into the realm of geostatistics, which traditionally is defined as the study of spatially continuous processes using spatially discrete observations at a finite number of locations. We suggest that a more useful definition of geostatistics is by the class of scientific problems that it addresses, rather than by particular models or data formats.