Whereas ultrasonography is totally recognised in the surveillance of pregnancy, the simple observation of too scattered and limited results in the screening of foetal malformations leads us to think that the good practice and the subtleties of the ultrasonographic examination are not yet mastered. The objective of FM-Ultranet was to offer help in real time to practitioners facing a malformation suspicion. FM-Ultranet acts as an electronic assistant for the practitioner while the examination is being performed.
The objective of the FM-Ultranet trial was to experiment an on-the-job training platform for ultrasonographists in the field of pregnancy surveillance to improve early detection of foetal malformation. The training platform aims at helping ultrasonographists in improving their diagnosis skills by comparison with existing clinical cases, exploitation of their own past experience and means to contact external experts. From a technical perspective, the FM-Ultranet trial aimed at integrating state-of-the-art techniques in the fields of case-based reasoning (CBR), medical image analysis, multimedia tools, and computer based on-the-job training. The FM-Ultranet trial addressed the problem of improving foetal malformation detection which is a very difficult task for a non specialist of this domain. The potential impact for the project is wide with the market of several thousand ultrasonographists in Europe, an improved coverage of women pregnancy and a reduction of public health expenses.
We have designed a decision support tool in the field of antenatal diagnosis. The tool allows the sharing of antenatal diagnosis expertise as well as of knowledge. This system relies on a fully documented base containing 130 reference cases of urinary tract malformation with the corresponding iconography. Technically, FM-Ultranet blends Case Based Reasoning (CBR) and the use of modern technologies -in particular internet technologies combined with modern storage technologies. The choice of CBR for getting help in decision making seems to be the most interesting since it simulates the medical way of reasoning. With this tool one can retrieve those cases closest to a case submitted by the physician into a data base. The similarity measure uses a specific medical measure developed by gyneco-obstetricians. Additionally, CBR represents the cases under as a hierarchical base sorted according to the sequence of the the examination stages, this has an important meaning from a clinical point of view. Every case is described by a set of characteristics, sometimes documented sometimes not, grouped into concepts ; this knowledge is governed by a set of rules (medical rules, coherence rules, warnings, ...).
FM-Ultranet was partially funded by the European Commission under contract number IST-1999-20865. The project was carried out between December 2000 to May 2002.
FM-Ultranet consisted of several academic, medical, industrial and institutional partners.
- University Pierre and Marie Curie Paris
- empolis GmbH
- University of Hildesheim (as subcontractor of empolis GmbH, location of this research group from 2001-2004)
- University of Montpellier
- University of Liege
Two members of the Wi2-group have been working within FM-Ultranet.
- Ziad El Balaa, Anne Strauss, Philippe Uziel, Kerstin Maximini, and Ralph Traphöner. FM-Ultranet: a Decision Support System using Case-Based Reasoning, Applied to Ultrasonography. In Lorraine McGinty, editor, Workshop on Case-Based Reasoning in the Health Sciences, Trondheim, Norway, June 2003, NTNU Department of Computer and Information Science.
- Dirk Krechel, Ralph Bergmann, Kerstin Maximini, and Aldo von Wangenheim. A Framework for Radiological Assistant Systems. In Peter Kokol, Bruno Stiglic, Milan Zorman, and Damjan Zazula, editors, Proceedings of the 15th IEEE Symposium on Computer-based Medical Systems (CBMS'02), pages 179-184, Maribor, Slovenia, June 2002, IEEE Computer Society, Los Alamitos, California.