Frédérique Laforest

Microarray experiments aim at analyzing expression levels of genes using DNA probes. The amount of data managed for each experiment is very large. It is thus essential to provide electronic support for the capture and the management of information describing microarray experiments. We present here the SITRANS Web information system, the aim of which is to help research workers storing, browsing, sharing and publishing data.

This article presents a global view of our proposal for a medical information system of the future. This information system focuses on patient medical records management. As most of the existing systems, it proposes to store information from patient records in a database. But records capture is different: we propose to use weakly-structured documents. Such documents contain paragraphs with some specific constraints represented by XML tags. The end-user writes new information under the form of weakly-structured documents. An internal system translates theses documents into new data for the internal database. Such a document-based user interface provides much more freedom to the end-user, and certainly reduces the distance between the physicians' way of working and the capture system.

This work concerns the management of electronic medical records. Classical systems based on databases do not satisfy end users, as the provided capture forms are never completely adapted to the patient case and to the physician's way of working. Electronic documents proved to be a better user interface paradigm, but still do not provide satisfying querying performance. Our work concerns the use of semi-structured documents as a user interface for a database-based system. The captured documents are analyzed so that information is extracted and used to fill in a database. We illustrate our proposition with the prototype we have developed and a short example.

Most of the information retrieval models for documents are intended for SGML-like structured documents. In the context of medical informatics, the patient record document needs a looser structuration process as an a priori structure can hardly be defined. Thus we first propose an authoring tool that allows us to annotate embedded information, i.e. to give them a context, with qualifiers that are stored in a thesaurus rather than in SGML-like DTD. The retrieval process in the patient records collection takes into account the flexibility of the qualifying process while reformulating the queries.

This article presents a new way to manage computerized medical records, based on a totally-hypermedia system. As a matter of fact, the classical use of a database limits the necessary variability of the medical record, in function of both the patient profile and the care practitioner habits. The system we propose is based on a hospital Intranet, and on the XML language. This language allows the definition of semantic tags in hyperdocuments, and thus information retrieval is ensured through semantic tags indexation.

Various systems are being used for managing medical information. They have several shortcomings. Among these, one can quote: -they do not overcome the complex medical terminology and its diversity; -they do not use a common medical language -there is no easy communication between the softwares used -and so on The proposed system takes into account the multiplicity of information sources within a hospital and the need for individual decision-making from care providers.