Extending temporal databases to deal with telic/atelic medical data☆

Summary 

Objective

In this paper, we aim at defining a general-purpose data model and query language coping with both “telic” and “atelic” medical data.

Background

In the area of Medical Informatics, there is an increasing realization that temporal information plays a crucial role, so that suitable database models and query languages are needed to store and support it. However, despite the wide range of approaches in the area, in this paper we show that a relevant class of medical data cannot be properly dealt with.

Methodology

We first show that data models based on the “point-based” semantics, which is (implicitly or explicitly) assumed by the totality of temporal database approaches, have several limitations when dealing with “telic” data. We then propose a new model (based on the “interval-based” semantics) to cope with such data, and extend the query language accordingly.

Results

We propose a new three-sorted model and a query language to properly deal with both “telic” and “atelic” medical data (as well as non-temporal data). Our query language is flexible, since it allows one to switch from “atelic” to “telic” data, and vice versa.

Conclusion

In this paper, we demostrate the feasibilty of a database approach copying with both telic and atelic data as needed in several (medical) applications.

Keywords: Temporal databases, Relational data model, TSQL2-like query language, Data semantics, Telic/atelic clinical data