MAIN RESEARCH PROJECTS |
|
These are the main projects currently :
Collaborations institutional projects) with
|
|
| R2CP : Representation of the Summary of Product Characteristics of drug (SPCs) Considered altogether, the SPCs of drugs can be seen as a large corpus of texts written under the control of the French Drug Agency (AFSSAPS). They describe the properties of drugs such as their indications, contraindications, pharmacokinetics, pharmacodynamics, cautions for use, recommended regimens. They are written using natural language but the vocabulary used is standardized and homogenized. Structuring their representation could lead to new applications. We try to model the content of the various SPC subsections using natural language processing methods and knowledge engineering. The final objective is to develop tools useful to improve medical reasoning and prescribing, to help for the development strategy of new drugs, to improve the quality of drug data bases and facilitate their maintenance. |
|
| ASTI (Decision support system for therapeutical strategy in primary care) The objective is to develop methods and tools to help the physician to select the best medication adapted to the specific medical problems of patients. This project addresses the representation problems of clinical guidelines and the design of a generic engine able of using this knowledge. The first phase of the project has lead to a prototype developed in the field of the hypertension medication. The prototype has the possibility to work either as a critiquing system or to provide the physician with a therapeutical guidance. The extension of ASTI to other diseases such as type 2 diabetes is in progress. The final goal is to extend the approach to various diseases and to promote the development and spread of a therapeutical knowledge base and associated tools to be incorporated to software currently used in primary care.. |
|
| Accessing to the medical knowledge contained in texts remains a difficult task for the physician in front of a patient during a consultation. This projects aims at developing a graphical language which could be used for facilitating the reading of medical documents such as the drug monographies or the clinical guidelines. This language must be easy to learn. It must permit to represent various medical concepts such as diseases, symptoms, drugs, lab tests. A first version of VCM has been developed. Its use has been evaluated for designing a graphical interface aimed at speeding the reading of drug monographies. Other applications of VCM to medical search engines or to the electronic medical records are currently in progress. | |
|
Cadre: Projet CNRS theme: Plateforme logiciel/Sélection d'attributs L’analyse des données issues de puces ADN représente l’un des thèmes de recherche pluridisciplinaires affiché CNRS dans la thématique « le vivant et ses enjeux sociaux". Florence D’alche-Buc, Vincent Corruble, Maria Rifqi et moi même avons eu notification en juin 2002 de l’acceptation d’un projet (MineChips) concernant le développement d’outils d’analyse des données. Début du projet: 2002. Durée prévue: 1 ans |
|
|
Cadre: Projet INSERM-AVENIR Theme: Classification Automat. de Donneées BioPuces basée sur GO) Nous avons développé un ensemble d’outils pour la récolte des données. Il s’agissait en particulier de récupérer automatiquement des informations issues de la base de données en ligne GeneOntology (http://www.geneontology.org) pour compléter la description des gènes. Nous avons d’autre part conçu un programme permettant d’extraire des connaissances sur les fonctions des gènes. Celui-ci est essentiel aux algorithmes de regroupement d’objets que nous utilisons [Bournaud, Courtine & Zucker, 2002] Début du projet: 2001 |
|
| Cadre: Fonds Berkeley-France 2003
Dudoit, Sandrine School of Public Health, UC Berkeley Début du projet Sept/2003. |
|
|
Ce projet est soutenu par l’AFCRST (Association Franco-Chinoise pour Recherche Scientifique et Technique). Il concerne un projet de collaboration avec l’Equipe du Prof. Lingwen Zeng du Centre National pour les BioPuces de Beijing (Capital Biochip Technology). L’organisme de rattachement de l’équipe chinoise est l’institut d’informatique de l’université TsingHua à Beijing Ce projet implique aussi un chercheur de l'équipe d'Apprentissage Automatique de l’académie des sciences de Beijing. Objectif du projet: Mettre en place des puces à ADN dédiées pour les gènes de l'obésité et concevoir des algorithmes dédiés à leur analyse. Début du projet 2003. Fin du projet 2005. Responsable du projet coté Chinois: Lingwen Zeng Respondable du projet coté Français: Jean-Daniel Zucker |
|
| Introduction l Access l Themes l Publications l Projects l Summary |
| Number of hits since 01/01/2003 |