Searching biosignal databases by content and context: Research Oriented Integration System for ECG Signals (ROISES)

Area of Science:

• Biomedical Engineering
• Data Science
• Medical Informatics

Background:

• Technological advancements have led to massive bio-signal datasets, particularly electrocardiogram (ECG) data.
• Existing ECG data storage formats are diverse and heterogeneous, impeding integrated access and management.
• Extracting new knowledge from bio-signals is crucial for advancing medical procedures.

Purpose of the Study:

• To develop a methodology for unified access to bio-signal databases and metadata.
• To enable transparent content and context-based searching across multiple data resources.
• To address the challenges posed by the diversity and heterogeneity of bio-signal storage formats.

Main Methods:

• Definition of an interactive global ontology to manage similarities and differences between data sources.
• Establishment of similarity mappings and enrichment of the ontology's terminological structure.
• Introduction of the Research Oriented Integration System for ECG Signals (ROISES) for complex queries.

Main Results:

• A novel methodology for unified access to bio-signal databases.
• Decoupling of information retrieval from underlying data source structures.
• Facilitation of transparent, content, and context-based searching across diverse bio-signal data.

Conclusions:

• The proposed methodology and ROISES system effectively address the challenges of accessing and querying heterogeneous ECG data.
• This approach enables more efficient knowledge extraction from large bio-signal datasets.
• It supports the adoption of new medical procedures through improved data accessibility and integration.