Multi-layer Data Model: A Contribution to Semantic Interoperability

The aim of the paper is to describe a data model representing fusion of clinical events (e.g. diagnosis, medications, etc.) with parameters and signals in semantic formalized form and in relation with the data evaluation.

Nowadays there are many approaches to the tasks of representation nomenclature, data structure, linking data and information sources and data decomposition. There are methods for the fusion of different types of data and information from different sources, and there are methods for formalized semantic descriptions of some types of biological signals (typically ECG). However, it is not sufficient to store measured signals in structured files but this data must be stored in such structured way as other patient data enabling easier access and satisfying requirements of semantic interoperability. In addition to data, it is very important to complete the structured information with evaluation of measured and derived data.

All of these areas are very important for the most comprehensive record of the identified states and processes carried out by the patient (in the clinic) as well as an experimental subject (experimental environment). The aim is to arrive at a description, which always contains sufficient context and allows the interpretation of such a record in the most relevant context.

The result of our work is a proposal of Multi-layer data model that is based on three main elements: representation of event; representation of signal and representation of rules (knowledge). They are represented as individual layers, according to level of abstraction and position in the knowledge. These layers are then connected and form a description of the observed and measured conditions and processes allowing description of their interconnections and relationships.