On going projects

  • Handling  missing data in mixed data – modelisation of polytrauma patients with Jean-Pierre Nadal and the Traumabase group. The project is in collaboration with APHP (Public Assistance – Hospitals of Paris).  When a patient arrives at the emergency room, there is a succession of important decisions which are taken in particular on the gravity of his state which lead to give him more or less attention. Of course, errors of diagnosis at these stages can be dramatic. Based on an incomplete database (constituted by the gathering of registers of several centers in the Ile-de-France region), the ultimate goal is to develop a decision model to guide and support the emergency workers.
  •  Distributed matrix completion for medical database  with Geneviève Robin and François Husson, Balasubramanian Narasimhan. Personalized medical care relies on comparing new patients profiles to existing medical records, in order to predict patients treatment response or risk of disease based on their individual characteristics, and adapt medical decisions accordingly. The chances of finding profiles similar to new patients, and therefore provide them better treatment,  increase with the number of individuals in the database. For this reason, gathering the information contained in the databases of several hospitals promises better care for every patient. However, there are technical and social barriers to the aggregation of medical data. The size of combined databases often makes computations and storage intractable, while institutions are usually reluctant to share their data due to privacy concerns and proprietary attitudes. Both obstacles can be overcome by turning to distributed computations, which consists in leaving the data on sites and distributing the calculations, so that hospitals only share some intermediate results instead of the raw data. This solves the privacy problem and reduces the cost of calculations by splitting one large problem into several smaller ones. We focus here on handling missing values in this distributed framework.