Predictive computer models for medical classification problems based on patient data and expert knowlegde

From 01-02-2000 to 31-01-2003

Description

The central theme of this interdisciplinary research is the development as well as the clinical evaluation of a new class grey-box computer models for robust medical classification problems which combine in an optimal way the knowledge extraction from patient data (black box) wtih medical expert knowledge (white box). The knowledge to be extracted can be very diverse, such as the preoperative prediction of malignancy of tumors, characterization of the pathology, preoperative staging, etc. In order to develop and evaluate this general methodology, the research is mainly focused on the preoperative discrimination of ovarian tumors. During the project we will investigate how the obtained results can be used by other research groups : not only by other centers of gynaecologic oncology but also by other divisions in U.Z. Leuven (e.g. for the treatment of breast and prostate tumors and in studies of the Sudden Infant Death Syndrome).

Team

• Sabine Van Huffel, Promoter

Financing

Funding: KU Leuven - Internal Funding KU Leuven

Program/Grant Type: BOF-Research - BOF Research Programmes