healthdatahub/applications-du-hdh | gitlab.com owners

healthdatahub/applications-du-hdh/cartographie-ecosysteme-snds

The main database of the French National Health Data System (SNDS) contains data from Health Insurance reimbursements, hospital treatment and medical causes of death. In order to characterise its use for health research and innovation, an interactive cartography has been produced to understand the framework of its use and to identify the stakeholders of the SNDS ecosystem. A bibliographic search via PubMed (available here), LiSSa, HAL was conducted to identify scientific articles published starting January 2007 on studies using SNDS data. The list of authors, their affiliations, keywords, the number of citations and much more were collected. A descriptive analysis was carried out in order to assess temporal and geographical trends in the use of SNDS main database. The graphs where generated with networkx, a python package used for the creation manipulation and study of complex networks. To generate the Author/Affiliations graphs we first create the adjacency matrix between the Authors/Affiliations and the article PMIDs. We then use the networkx.Graph class to create the needed undirected graphs, using the adjacency matrices as the data to intialize the graphs.

Last synced at: 5 days ago - Pushed at: 11 months ago - Stars: 4 - Forks: 0

healthdatahub/applications-du-hdh/schema-snds

Formalisation du schéma du SNDS selon la spécification table-schema. Création de produits dérivés.

Last synced at: 5 days ago - Pushed at: 11 months ago - Stars: 9 - Forks: 13

healthdatahub/applications-du-hdh/dico-snds

Ce projet contient l'application Rshiny d'un dictionnaire interactif de la base principale du SNDS. Cet outil est accessible à cette adresse : https://drees.shinyapps.io/dico-snds/

Last synced at: 5 days ago - Pushed at: 11 months ago - Stars: 4 - Forks: 1

healthdatahub/applications-du-hdh/pathology-slides-blur-detection

The functions provided in this repo can be used to label images showing blurred areas. These functions have been designed to highlight blurred areas on digitized melanoma biopsy slides. In that case, the blurring was due to a problem with the scanner's focus at the time of scanning. The possible labels for each image are: “totally blurred”, “partially blurred”, “no blur”. For the “partially blurred” category, an approximation of the blurred area is provided.

Last synced at: 5 days ago - Pushed at: 11 months ago - Stars: 1 - Forks: 0