Datathon for Intensive Care DAT-ICU event
20-21st of January // PARIS, FRANCE


Participants will have 48 hours to carry out an impactful project using freely available hospital data, for example predicting patient physiology using state of the art algorithms.

MIMIC data as playground

MIMIC is an openly available ICU dataset developed by the MIT Lab for Computational Physiology, comprising deidentified health data associated with ~40,000 critical care patients. It includes demographics, vital signs, laboratory tests, medications, and more.

Who should attend?

We aim to bring together clinicians, data scientists and innovators in healthcare to address current problems in healthcare with data analytics technologies.

8 million patients 40 hospitals

AP-HP (Greater Paris University Hospitals) is a European world-renowned university hospital. Its 39 hospitals treat 10 million people every year: in consultation, emergency, during scheduled or home hospitalizations. AP-HP is the leading employer in the Greater Paris area.


25 projects/teams: the winners will be selected by the jury and the community and will have the chance to experiment their ideas/models on the APHP ICU data (the biggest medical data aggregation).

Big Data

DAT-ICU : Datathon for Intensive Care


Greater Paris University Hospitals, Massachusetts Institute of Technology, the French Intensive Care Society, University Pierre et Marie Curie and University Paris Descartes, with the help of the French Society of Anesthesia & Intensive Care Medicine, have the pleasure to invite you to the event :


DAT-ICU : Datathon for Intensive Care

which will be held on Saturday 20th and Sunday the 21st of 2018

in the Campus Picpus (access through Rothschild Hospital, 5, Rue Santerre, 75012 Paris – metro stations Bel-Air, Picpus and Nation)

DAT-ICU aims at uniting clinicians and data scientists around medical data to answer relevant clinical questions.

The participants will have 48 hours to undertake a clinical project using the MIMIC database in OMOP format. This database, developed by the computational physiology lab of MIT, contains all the information about 50 000 intensive care unit patients (demographics, vital signs, biology, medication, reports, etc.).

You will have the opportunity to create mixed teams. Clinicians will bring the questions which need data mining, along with their expertise of the data. Data scientists judge the technical feasibility and eventually implement the various analysis needed. Interpretation will be everyone’s task.

You will have access to work stations connected to AP-HP calculation clusters provided by IT/IS teams. You will be able to access to the data pre-loaded in Jupyter environments, where will be installed the most popular tools and libraries in R and Python, with Hadoop Spark.

A prize will be awarded by an interdisciplinar jury to congratulate the most innovative approaches. Those projects will have the opportunity to present their results in front of the national congress of the French Intensive Care Society which will be held on Wednesday the 24th of January 2018. They will also be encouraged to write a research protocol to access later on to Greater Paris University Hospitals ICU data, which will be in the same OMOP format. Besides, a special prize could be awarded.

The registration is free on http://blogs.aphp.fr/dat-icu. For logistic reasons, it is restricted to 25 teams. Before the event, you are invited, if possible, to create your team in advance, through the forum on the website, which can you find the relevant expertise you need.

Sincerely yours,
The organizing team

DAT-ICU is endorsed by the French Society of Anesthesia and Intensive Care Medicine (SFAR)