BioSignalML: An abstract model for physiological time-series data

This thesis proposes a new standard framework, BioSignalML, for describing physiological time-series data, in order to address some of the challenges created by the diverse range of formats used for biosignal storage and exchange. I argue that the main cause of difficulties for researchers, wanting to share and exchange biosignal data, is a lack of standardisation as to how metadata is represented and assigned meaning, even for common attributes, and not due to differing data formats. The BioSignalML model seeks to address the lack of metadata standardisation and considers data and metadata to be of equal importance. Instead of specifying another new storage format, BioSignalML uses open data and metadata standards, in particular those from the Semantic Web, to describe signals in their existing formats. An abstract model and a new ontology, the BioSignalML Ontology, provide a shared framework for biosignals, allowing their information to be extended, unified, and interlinked with other resources on the Web. This approach, of applying Semantic Web technology to biosignals in a general way, is believed to be novel. The BioSignalML so ware library allows applications to create and access signal data and metadata in a format independent way. This library forms the basis of a web-accessible biosignal repository that allows standard web applications, including Semantic Web tools, to be used to query, browse, annotate and process data in signal collections. All so ware is freely available as an open-source resource. The utility of BioSignalML is demonstrated by its application in three separate areas — to enhance a public repository of physiological signal recordings; to facilitate the integration of biosignal data with physiological modelling applications; and to manage polysomnography and respiratory flow data used in product research and development. In all these applications, users benefit from improved query, annotation, and data services. BioSignalML is designed to facilitate data integration and provide metadata consistency, both within and between research groups, and across a wide range of research domains, in a way that allows for future extension.

Name thesis:degree_name

PhD

Level thesis:degree_level

Doctoral

Grantor dc:publisher

ResearchSpace@Auckland

Year dc:date.issued

2013

Author dc:creator

• Brooks, David

Advisors dc:contributor.advisor

• Titchener, MR
• Smaill, BH

Statement dc:rights

• Items in ResearchSpace are protected by copyright, with all rights reserved, unless otherwise indicated. Previously published items are made available in accordance with the copyright policy of the publisher.

Licence dc:rights.uri

• https://researchspace.auckland.ac.nz/docs/uoa-docs/rights.htm

Handle dc:identifier.uri

https://hdl.handle.net/2292/22026

OAI identifier oai:identifier

oai:researchspace.auckland.ac.nz:2292/22026