Towards Health Data Stream Analytics

Data streams, or data sets which continuously and rapidly grow over time, are a prominent form of clinical data generated during the monitoring and treatment of patients in the health care industry. We propose the name Health Data Stream Analytics (HDSA) to the application of stream data processing...

Full description

Saved in:
Bibliographic Details
Published in2010 IEEE/ICME International Conference on Complex Medical Engineering pp. 282 - 287
Main Authors Zhang, Qing, Pang, Chaoyi, Mcbride, Simon, Hansen, David, Charles Cheung, Steyn, Michael
Format Conference Proceeding
LanguageEnglish
Published IEEE 01.07.2010
Subjects
Biomedical monitoring
Data processing
Data visualization
Electroencephalography
Heart rate
Monitoring
Sensors
Time series analysis
Trend analysis
Online AccessGet full text
ISBN1424468418
9781424468416
DOI10.1109/ICCME.2010.5558827

Cover

More Information
Summary:Data streams, or data sets which continuously and rapidly grow over time, are a prominent form of clinical data generated during the monitoring and treatment of patients in the health care industry. We propose the name Health Data Stream Analytics (HDSA) to the application of stream data processing to clinical data. Our work in this area is demonstrating the useful role Health Data Stream Analytics can play in clinical decision support, patient safety improvement and early detection of adverse patient outcomes. Two major challenges in applying stream data processing to heath care are tailoring query support for the clinical context and dealing with the clinical requirement of online query processing. In this paper, we propose the Anaesthetic Data Analyser (ADA) as a Health Data Stream Analytics System for the anaesthetics specialty and describe how it addresses these challenges. ADA differentiates from current approaches by looking at trends in the data stream rather than a single data value against a preset threshold. The trend analysis supported by ADA is a novel application in this area, and enables support for adverse symptoms monitoring in physiological stream data, alerting clinicians when a pre-defined adverse data pattern is detected in the physiological signals. This paper also describes an online query processing algorithm and the results of experiments on "real world" physiological steam data which indicate the algorithms has sub-second response times for trend queries.
ISBN:1424468418
9781424468416
DOI:10.1109/ICCME.2010.5558827