Jungyoon Kim

ISSAQ: An Integrated Sensing Systems for Real-Time Indoor Air Quality Monitoring

With growing transportation and population density, increasing global warming and sudden climate change, air quality is one of the critical measures that is needed to be monitored closely on a real-time basis in todays urban ecosystems. This paper examines the issues, infrastructure, information processing, and challenges of designing and implementing an integrated sensing system for real-time indoor air quality monitoring. The system aims to detect the level of seven gases, ozone (O3), particulate matter, carbon monoxide (CO), nitrogen oxides (NO2), sulfur dioxide (SO2), volatile organic compound, and carbon dioxide (CO2), on a real-time basis and provides overall air quality alert timely. Experiments are conducted to validate and support the development of the system for real-time monitoring and alerting.

The joint relationship between organizational design factors and Hr practice factors on direct care workers’ job satisfaction and turnover intent

Background: Human resource (HR) practices, such as training and communication, have been linked to positive employee job commitment and lower turnover intent for direct care workers (DCWs). Not many studies have looked at the combined interaction of HR practices and organizational structure. Purpose: The aim of this study is to examine the relationship between organizational structure (centralization, formalization, and span of control) and HR practices (training, horizontal communication, and vertical communication) on DCW’s job satisfaction and turnover intent. Methodology: Data were collected from 58 long-term care facilities in five states. We used latent class analysis to group facility characteristics into three sets of combinations: “organic,” “mechanistic,” and “minimalist.” We used multivariate regression to test the relationship of each of these groups on DCW’s job satisfaction and turnover intent. Findings: After controlling for state, organizational, and individual covariates, the organic group, which represents decentralized and less formalized structures and high levels of job training and communication, was positively related to job satisfaction and negatively related to intent to leave. On the other hand, the minimalist group, which is characterized by low levels of job-related training and communication, showed no significant differences from the mechanistic group (referent) on job satisfaction and intent to leave. Practice Implications: These findings imply that managers in long-term care facilities may want to consider adopting organic, decentralized structures and HR practices to retain DCWs.

Analysis of energy consumption for wearable ECG devices

Wearable health monitoring systems (WHMS) have received much attention for the last decade due to increasing healthcare costs and the aging world population. Although advances in sensors, embedded systems, wireless communication and intelligent information technologies has brought us tiny wearable ECG devices and efficient algorithms to continuously monitor and analyze ECG signals, energy consumption is a critical concern for practical usage. In this study, we examine three possible models of data transmissions - transmitting all data in real time (type 1), transmitting all data in batch periodically (type 2), or only transmitting the detected events (type 3), in which the third model is our proposed method. A simulated environment and a generic energy consumption model are developed to examine the influence of different strategies and Ventricular Fibrillation (VF) detection methods on energy consumption.

Epidemiologic and Clinical Profiles of Breast Diseases in Niger

This study aimed at characterizing epidemiologic and clinical profiles of breast diseases in Niger during the period of 2010–2013 at the National Hospital of Niamey. Medical records were abstracted for demographic, reproductive, clinical, and treatment information. A process map of patient navigation and barriers to seeking medical care was developed after interviewing 26 local health professionals who encounter and/or manage breast diseases. We identified 245 breast cancers and 122 other breast diseases. Mean age of breast cancer patients was 45.4 (±13.26 years) and that of breast diseases was 31(±12.5 years) with 1/3 of cancers under age 44. Infection-related diseases represented 24% of non-cancers. Male breast diseases represented 4.75% of diseases and 2.05% of cancers. Only 37.1% of cancers had histopathologic confirmation and 90% of cancer patients presented at advanced stages and mastectomy was performed for 66% of breast cancers. Patient and system barriers to care were common in diagnosing and treating breast diseases. Women in Niger have double burden of infectious breast diseases and emerging breast cancer. Younger age and late diagnosis are common features. Reducing barriers to access to care, down-staging of cancer, implementation of clinical guidelines for managing advanced cases are important needs for reducing breast cancer morbidity and mortality in Niger.

Analysis and Modeling of Selected Energy Consumption Factors for Embedded ECG Devices

Heart attack is a life threaten cardiac disease. Although advances in information and communication technologies have brought us tiny embedded electrocardiogram (ECG) devices and efficient algorithms to continuously monitor and analyze ECG signals, energy consumption of the system is a critical concern for practical usage. However, there have been not many studies focusing on energy consumption of mobile ECG monitoring, especially lacking guidelines in selecting proper data communication strategies and computational algorithms. In this paper, we examine three possible modes of data transmissions for energy consumption. A generic energy consumption model and a simulated environment were developed to examine the influence of different communication strategies and ventricular fibrillation detection methods in terms of computational efficiency and power consumption. This paper shows that the total ratio of energy saving between mode transmitting all data and our proposed event transmission is up to 84.67% per 508 s of data segment.

Designing Integrated Sensing Systems for Real-Time Air Quality Monitoring

With increasing global warming and sudden climate change, air quality monitoring is one of the important issues need to be addressed globally in todays ecosystems. This paper examines the issues and challenges of designing and implementing an integrated sensing system for air quality monitoring. The system is aiming to detect the level of six gases, Ozone, Particulate Matter, Carbon Monoxide, Nitrogen Oxides, Sulfur Dioxide, and Carbon Dioxide, as suggested by the US Environmental Protection Agency (EPA), on a real-time basis and provide overall quality alert timely. Experiments are conducted to validate and support the development of an integrated sensing system for real-time air quality monitoring.

A simple and effective algorithm for R-wave detection using smartphones

Developing simple and effective algorithms for detecting cardiovascular diseases (CVDs) using smartphones has been the trend of research in pervasive healthcare. This paper proposes an efficient R-wave detection algorithm and tested it in an Android Smartphone. We use the complete datasets of MIT-BIH arrhythmia database to verify its relative performance with three other algorithms. Our results show that it can obtain an average error rate of 0.22%, a sensitivity of 99.88% and a predictive accuracy of 99.90%; while, only takes 12ms to process 10 seconds data in a smartphone; thus, it is more suitable for use at mobile devices.

BMRC: A Bitmap-Based Maximum Range Counting Approach for Temporal Data in Sensor Monitoring Networks

Due to the rapid development of the Internet of Things (IoT), many feasible deployments of sensor monitoring networks have been made to capture the events in physical world, such as human diseases, weather disasters and traffic accidents, which generate large-scale temporal data. Generally, the certain time interval that results in the highest incidence of a severe event has significance for society. For example, there exists an interval that covers the maximum number of people who have the same unusual symptoms, and knowing this interval can help doctors to locate the reason behind this phenomenon. As far as we know, there is no approach available for solving this problem efficiently. In this paper, we propose the Bitmap-based Maximum Range Counting (BMRC) approach for temporal data generated in sensor monitoring networks. Since sensor nodes can update their temporal data at high frequency, we present a scalable strategy to support the real-time insert and delete operations. The experimental results show that the BMRC outperforms the baseline algorithm in terms of efficiency.