C. Sweetlin Hemalatha

Agent Based Health Monitoring of Elderly People in Indoor Environments Using Wireless Sensor Networks

Abstract This paper presents a design of a health care monitoring system based on wireless sensor networks (WSN) which is capable of collecting, retrieving, storing and analysing the vital signs of the patient. Vital signs such as body temperature, blood pressure, pulse rate and respiratory rate are monitored. To manage these sensors and for collecting and storing data in a database, a Multi Agent System (MAS) is used. The dynamic nature and mobility of the agents make them suitable for maintaining these sensors in the WSN. The proposed MAS consists of four agents namely Admin agent, Control agent, Query agent and Data agent. Admin agent plays the role of invoking and terminating other agents. Control agent is responsible for storing the data sensed by the sensors into the database. Data agent performs data reduction which is achieved using Epsilon approximation. The use of Data agent in the proposed scheme reduces data traffic and the requirement of secondary storage space. Query agent is responsible for providing a GUI for the doctor to view the patients vital sign details. The proposed agent based system has been implemented using Java language in JADE environment and the results are validated.

Minimal infrequent pattern based approach for mining outliers in data streams

Minimal Infrequent Pattern based Outlier Detection.An algorithm for mining minimal infrequent patterns in data streams.Three simple factors deciding outliers.An algorithm for detecting outliers based on mined minimal infrequent patterns.Experimental results with real time sensor data and publically available UCI data set. Outlier detection is an important task in data mining which aims at detecting patterns that are unusual in a dataset. Though several techniques are proved to be useful in solving some outlier detection problems, there are certain issues yet to be resolved. Most of the existing methods compute distance of points in full dimensional space to detect outliers. But in high dimensional space, the concept of proximity may not be qualitatively meaningful due to the curse of dimensionality and incurs high computational cost. Moreover, the existing methods focus on discovering outliers but do not provide the interpretability of different subspaces that cause the abnormality. Frequent pattern mining based approaches resolve the aforementioned issues. Recently, infrequent pattern mining has attracted the attention of data mining research community which aims at discovering rare associations and researches in this area motivated to propose a new method to detect outliers in data streams. Infrequent patterns are more interesting than frequent patterns in some domains such as fraudulent credit transactions, anomaly detection, etc. In such applications, mining infrequent patterns facilitates detecting outliers. Minimal infrequent patterns are generators of family of infrequent patterns. In this paper, a novel method is presented to detect outliers by mining minimal infrequent patterns from data streams. Three measures namely Transaction Weighting Factor (TWF), Minimal Infrequent Deviation Factor (MIPDF) and Minimal Infrequent Pattern based Outlier Factor (MIFPOF) are defined. An algorithm called Minimal Infrequent Pattern based Outlier Detection (MIFPOD) method is proposed for detecting outliers in data streams based on mined minimal infrequent patterns. The effectiveness of the proposed method is demonstrated on synthetic dataset obtained from vital dataset collected from body sensors and a publicly available real dataset. The experimental results have shown that the proposed method outperforms the existing methods in detecting outliers.

Multi-sensor based in-home health monitoring using Complex Event Processing

Web enablement of sensor data is an important mission in providing anywhere anytime access. Particularly, it is essential in Tele-health monitoring of geriatric patient who are alone at home. For continuous monitoring, it requires the patient to wear wireless body sensors which give information about his vital parameters. Sensor Web Enablement (SWE) provides a platform for making the raw sensor data available on the web so that it becomes accessible to doctors for making clinical diagnosis. The challenges involved are effective collection of sensor data and bringing them to web using Service Oriented Architecture (SOA), complexity in finding relationship between raw events, developing rules for identifying patterns that pose a threat and generating alerts to patient, caregivers and doctors, dealing storage and privacy issues in accessing data on web and developing algorithm for fast and accurate fall detection. This talk addresses those challenges by providing possible solutions.

Multi-Level Search Space Reduction Framework for Face Image Database

In face recognition, searching and retrieval of relevant images from a large database form a major task. Recognition time is greatly related to the dimensionality of the original data and the number of training samples. This demands the selection of discriminant features that produce similar results as the entire set and a reduced search space. To address this issue, a Multi-Level Search Space Reduction framework for large scale face image database is proposed. The proposed approach identifies discriminating features and groups face images sharing similar properties using feature-weighted Fuzzy C-Means approach. A hierarchical tree model is then constructed inside every cluster based on the discriminating features which enables a branch based selection, thereby reducing the search space. The proposed framework is tested on three benchmark and two self-created databases. The experimental results show that the proposed method achieved an average accuracy of 93% and an average search time reduction of 66% compared to existing approaches for search space reduction of face recognition.

Associative Classification based Human Activity Recognition and Fall Detection using Accelerometer

Human fall poses serious health risks especially among aged people. The rate of growth of elderly population to the total population is increasing every year. Besides causing injuries, fall may even lead to death if not attended immediately. This demands continuous monitoring of human movements and classifying normal low-level activities from abnormal event like fall. Most of the existing fall detection methods employ traditional classifiers such as decision trees, Bayesian Networks, Support Vector Machine etc. These classifiers may miss to cover certain hidden and interesting patterns in the data and thus suffer high false positives rates. Hence, there is a need for a classifier that considers the association between patterns while classifying the input instance. This paper presents a pattern mining based classification algorithm called Frequent Bit Pattern based Associative Classification FBPAC that distinguishes low-level human activities from fall. The proposed system utilizes single tri-axial accelerometer for capturing motion data. Empirical studies are conducted by collecting real data from tri-axial accelerometer. Experimental results show that within a time-sensitive sliding window of 10 seconds, the proposed algorithm achieves 99% accuracy for independent activity and 92% overall accuracy for activity sequence. The algorithm gives reasonable accuracy when tested in real time.

Intelligent accident mitigation system by mining vital signs using wireless body sensor

Recent surveys show that the number of road accidents has increased predominantly. One of the major causes to which increased accidents are attributed to is physical ailment of drivers. Continuous monitoring of drivers health condition is essential in order to reduce car accidents that occur due to health abnormality. A non-intrusive method is demanded so as to prevent hindering of driving activity. This paper presents a mobile health monitoring system which is an application running in Android based smart phone. The mobile phone acquires vital parameters such as Heart Rate (HR) and Respiration Rate (RR) from a wearable body sensor through Bluetooth. As faster heart beat and shortness of breath are the most common symptoms of heart problem, the proposed system considers heart rate and respiration rate for deciding abnormal health status. A Bayesian Belief network (BBN) is designed to analyze the vital parameters along with the drivers health history and decide whether the health status of the driver is normal or abnormal in real time. Bayesian Network is a powerful representation for uncertain domains like human health status. Also, it improves classification accuracy as it allows seamless integration of additional information such as the drivers health records with sensed vital information for deciding abnormality and thus avoids false alarms. When abnormality is detected, immediately the driver gets a beep alert call in his mobile and call alert is generated to the caregiver in the case of emergency. The proposed system is tested in real time and it gives reasonable accuracy.

Adaptive learning based human activity and fall detection using fuzzy frequent pattern mining

Human activity recognition (HAR) has gained a lot of significance in monitoring the health of people, especially to detect fall among elderly people who live independently. This project proposes a novel method for recognizing activities and detecting fall of a person using body-worn sensors. Traditional algorithms like Naïve Bayes classifier and Support Vector Machine are mainly used for activity classification. However, these systems fail to capture significant association that exists between interesting patterns. Existing accelerometer based wearable systems are not sufficient to determine the fall of a person. Hence, a Fuzzy Associative Classification based Human Activity Recognition (FAC-HAR) system is proposed to overcome the aforementioned drawbacks in detecting abnormal status of a person. The proposed (FAC-HAR) system uses three different sensors namely heartbeat, breathing rate and accelerometer and employs fuzzy clustering and associative classification for abnormality detection. The proposed system introduces a novel learning mechanism is to improve classification accuracy. A classification accuracy of 92% is achieved with the proposed fuzzy frequent pattern mining based human activity recognition.

Mining infrequent patterns in data stream

In recent years researches are focused towards mining infrequent patterns rather than frequent patterns. Mining infrequent pattern plays vital role in detecting any abnormal event. In this paper, an algorithm named Infrequent Pattern Miner for Data Streams (IPM-DS) is proposed for mining nonzero infrequent patterns from data streams. The proposed algorithm adopts the FP-growth based approach for generating all infrequent patterns. The proposed algorithm (IPM-DS) is evaluated using health data set collected from wearable physiological sensors that measure vital parameters such as Heart Rate (HR), Breathing Rate (BR), Oxygen Saturation (SPO2) and Blood pressure (BP) and also with two publically available data sets such as e-coli and Wine from UCI repository. The experimental results show that the proposed algorithm generates all possible infrequent patterns in less time.

ECG anomaly detection using wireless BAN and HEMFCM clustering

In recent days, elderly people living alone at home are steadily increasing throughout the world. This situation drives to develop a health care system for monitoring the health parameters of elderly people and help them to lead ahealthy independent life. This paper presents a system that uses wireless sensors for monitoring the health parameters without disturbing the normal activities of elderly people. The proposed system provides a wearable health care solution using the wireless Shimmer sensor device for collecting ECG data in home PC. ECG data anomaly is detected using rule based classifier. Classification rules are generated based on cluster centroids obtained using a novel scheme named Hybrid Expectation Maximization and Fuzzy C Means (HEMFCM) Clustering. The proposed method is validated using real data collected from different subjects and abnormal data readings from the MIT BIH database. Experimental results show that proposed method achieves 85% classification accuracy which is better than EM and FCM clustering methods.

Constructing finite automata based model for detecting human fall

The fall of a person forms a major cause for serious health decline or injury related death in elderly persons. The existing fall detection algorithms are based on visual sensors. Fall detection using visual sensor has restriction in the coverage region due to privacy. Existing wearable sensor based on accelerometer is not sufficiency for detecting fall as it will not detect fall that occurs due to internal health abnormalities. Both activity and health data differ between persons of different age group. To handle such variation in the data, this paper proposes a novel fall detection scheme based on fuzzy finite automata with probability density function by formulating rule patterns combining the capabilities of automata using both accelerometer and physiological sensors. The proposed scheme provides a systematic approach for continuous monitoring of elderly patients living alone with the ability to detect the fall of the patient, overcoming the demerits of visual sensors.