Enhanced Complex Event Processing Framework for Geriatric Remote Healthcare

Enhanced Complex Event Processing Framework for Geriatric Remote Healthcare

V. Vaidehi (VIT University, India), Ravi Pathak (Striim Inc., India), Renta Chintala Bhargavi (VIT University Chennai, India), Kirupa Ganapathy (Saveetha University, India), C. Sweetlin Hemalatha (VIT University, India), A. Annis Fathima (VIT University, India), P. T. V. Bhuvaneswari (Madras Institute of Technology, India & Anna University, India), Sibi Chakkaravarthy S. (Madras Institute of Technology, India & Anna University, India) and Xavier Fernando (Ryerson University, Canada)
DOI: 10.4018/978-1-5225-5396-0.ch016
OnDemand PDF Download:
No Current Special Offers


Advances in information and communication technology (ICT) have paved way for improved healthcare and facilitates remote health monitoring. Geriatric remote health monitoring system (GRHMS) uses WBAN (wireless body area network) which provides flexibility and mobility for the patients. GRHMS uses complex event processing (CEP) to detect the abnormality in patient's health condition, formulate contexts based on spatiotemporal relations between vital parameters, learn rules dynamically, and generate alerts in real time. Even though CEP is powerful in detecting abnormal events, its capability is limited due to uncertain incoming events, static rule base, and scalability problem. To address the above challenges, this chapter proposes an enhanced CEP (eCEP) which encompasses augmented CEP (a-CEP), a statistical event refinement model to minimize the error due to uncertainty, dynamic CEP (DCEP) to add and delete rules dynamically into the rule base and scalable CEP (SCEP) to address scalability problem. Experimental results show that the proposed framework has better accuracy in decision making.
Chapter Preview

1. Introduction

Monitoring the health condition of elderly people and sharing the information with remote health care providers or hospitals is of great demand with increasing population of senior citizens choosing to live independently. A number of physiological sensors can be integrated into a wearable Wireless Body Area Network (WBAN), which can be used for computer-assisted rehabilitation or early detection of medical conditions (Pantelopoulos et al., 2010; Patel et al., 2012). The use of wireless sensors within a body area network makes continuous health monitoring seamless and easy.

Full utilization of WBAN occurs only when the events of interest are detected and respective action is taken in minimum time. Some of the challenges in developing a GRHMS are:

  • 1.

    The number of sensed events is high as every sensor observation generates an event.

  • 2.

    The latency in detecting an abnormal event should be as small as possible to allow appropriate actions to be taken during emergency.

This relies on the feasibility of placing very small biosensors on the human body that are comfortable and do not impair the regular activities of the patient. These sensors worn on the human body collect various physiological changes in order to monitor the health status of a person irrespective of their location. The collected health information is transmitted wirelessly in real time to the doctor. If any emergency is detected, then the physician is immediately informed about the patient’s status by sending appropriate alert message.

Vital sign monitoring plays an important role in detecting abnormalities in physiological parameters such as heart rate, respiration rate, blood pressure etc. Besides vital parameters, the activity of the patient is required to be monitored as it acts as context information to decide the health condition of the patient. Fall of an elderly person is of major concern in healthcare application as most of the unexpected and unpredictable falls leads to death and critical conditions. Hence, early detection of fall activity is required to treat such patients on time. Human activity recognition enables detection of sudden fall of the elderly person.

The vital parameters are highly correlated as any change in the health condition of a patient relates change in other two physiological parameters. Therefore, abnormality detection algorithms that identify the abnormal patterns need to be false proof. CEP based GRHMS proves to be very efficient (Vaidehi et al., 2012; Sanjana et al., 2013; Bhargavi et al., 2013; Ravi Pathak et al., 2015). The earlier work on CEP based IRHMS is extended and integrated in the proposed eCEP.

A data stream is a sequence of data generated continuously by a source (i.e., Sensor(s)). There are several challenges in collecting and processing these data streams as they are continuous and unbounded. Moreover, timely detection of interesting and complex events or patterns and response generation is critical and it depends upon the application. Traditional data processing paradigm does not fully meet the requirements such as response time, memory and throughput.

Complete Chapter List

Search this Book: