Issues in data fusion for healthcare monitoring

This paper proposes new architecture for handling data cleaning, data fusion, and context and knowledge generation using multi-tiered communication and a triadic hierarchical class analysis approach.

Abstract

Pervasive healthcare monitoring using body sensors and wireless sensor networks is a rapidly growing area in healthcare monitoring applications. Several issues arise in these systems, such as complex distributed data processing, data fusion, unreliable data communication, and uncertainty of data analysis in order to successfully monitor patients in real time. In this paper, we introduce some of the important issues in healthcare monitoring with focus on software problems such as reliability, network robustness, and context awareness. We describe related works in data filtering, data fusion, and data analysis then we suggest new architecture for handling data cleaning, data fusion, and context and knowledge generation using multi-tiered communication and a triadic hierarchical class analysis approach. The proposed architecture is called "Pervasive Healthcare Architecture" and we discuss how it can be applied to a particular monitoring scenario.