Nicolas Garcelon

Data science

Nicolas Garcelon
  • Vincent Benoit
  • Hassan Faour

Meilleures publications

Garcelon N. Finding patients using similarity measures in a rare diseases-oriented clinical data warehouse: Dr. Warehouse and the needle in the needle stack. Journal of Biomedical Informatics. 2017 Volume 73, September 2017, Pages 51-61.

Garcelon N. Improving a full-text search engine: the importance of negation detection and family history context to identify cases in a biomedical data warehouse. J Am Med Inform Assoc. 2017 24(3):607-613.

Garcelon N. Epidemiology of PID: Innovative New Way to Identify Patients in the CEREDIH Registry Through a Medical Data Warehouse. J. Clin. Immunol. 2014 34, S361–S362.

Campillo-Gimenez B. Full-text automated detection of surgical site infections secondary to neurosurgery in Rennes, France. Stud Health Technol Inform. 2013 192, 572–575.

Bouchireb K. NPHS2 mutations in steroid-resistant nephrotic syndrome: a mutation update and the associated phenotypic spectrum. Hum. Mutat 2014 35, 178–186.

voir toutes les publications du laboratoire

voir toutes les publications de l'Institut Imagine

The "data science" support team was created at the Imagine Institute in 2012, and has been labeled data science platform in 2017. The platform is made up of three IT engineers. The platform develops methods to accelerate translational research between doctors and researchers, between the hospital and the institute. We create software allowing the storage and analysis of patients’ phenotypic data.

Phenotypic and genetic databases
The aim is to help researchers and clinicians to collect, pool, structure, sustain and secure patient data for local, national or international studies. We developed the "eCohorte" software to quickly set up a database containing all the necessary tools for data entry and retrieval. We now manage 82 databases with a total of 131,000 patients. Much of this work also involves the automated retrieval and load of retrospective data.

The transversal tools
We have developed several tools for researchers, clinicians and managers: BioPancarte for personalized visualization of laboratory results, Auxo for the creation of growth curves, Gecko for the management of clinical studies and inclusions of Patient, Gene2CT for automatic monitoring of clinical trials associated with genetic mutations etc. These software are deposited to the Agency of Protection of the Programs.

Biomedical data warehouse
We developed Dr. Warehouse, a biomedical data warehouse that integrates free text data (hospitalization, consultation, prescription reports, etc.) as well as coded data (biological results).
Dr Warehouse is implemented in Necker-Enfants Malades hospital. We loaded 4 million documents from 20 different sources (EHR, Biomedical databases) from 1996 to 2017 for 480,000 patients. It also contains 36 million biological results.

The main functionalities are to be able to search patients using a 'Google like' search engine, to be able to carry out automatic phenotypic descriptions, and to propose tools to help diagnosis by similarity calculation between patients. Dr. Warehouse's interface provides physicians with ergonomic tools to explore medical records to accelerate the recruitment of patients into clinical studies.
Dr Warehouse is now also released under the GNU GPL (open source License).