SCCM
Log In
Forgot username or password?
New User? Sign Up Free.

Data Science Datathon

Pre-Congress Educational Session
Data Science Datathon
February 14-15, 2020
Orange County Convention Center
Orlando, Florida, USA

 

Big medical datasets and advancements in artificial intelligence (AI) and machine learning (ML) are transforming research, education, and the delivery of care for patients in the intensive care unit (ICU). The first Society of Critical Care Medicine (SCCM) datathon will bring together leaders in clinical and data science to tackle complex critical care questions. The event demystifies AI/ML and provides attendees with experience in big data research. Actual de-identified patient care data will be used to investigate the answers to an important clinical question—all in two intense days of hands-on investigation.

Who can participate in the datathon?

Clinical Participants
Clinicians are medical experts in fields such as critical care, anesthesia or emergency medicine who are able to identify and understand clinical questions related to patient care.

The Data Science Datathon is a pre-Congress educational session at the 49th Critical Care Congress. Clinical participants do not need to register for the full Congress to participate in this pre-Congress session. To learn more about registration, visit the Registration Information section below.

Register by January 6, 2020

Register for the Datathon & Congress

Register for the Datathon Only

 

Upon registration, you will receive an opportunity to help shape the datathon agenda by suggesting questions to be researched. More about what makes a good data research question is found in the FAQ section.

Data Scientists
Data Scientists are encouraged to apply for participation in the datathon. This is a unique opportunity to use your data science skills and work collaboratively with medical professionals to explore clinical questions related to critical care. Selected data scientist participants will provide the technical knowledge and expertise in utilizing a variety of software tools to extract information from large cloud-based patient databases in answering clinical research questions.

An immense amount of data originate from critical care; this is your opportunity to learn how to analyze and mine that data for meaningful secondary use. Many research projects initiated at prior datathons have resulted in articles published in peer-reviewed medical and scientific journals.

Data Scientist Applications are due by January 6, 2020
Accepted applicants will be notified in mid-January

Data Scientist Application

During the datathon, medical clinicians and data scientists will be grouped in teams to explore questions submitted by clinicians. They will use a variety of tools and actual patient information in de-identified databases to explore the answers to these questions. The teams will be mentored by experts from the Massachusetts Institute of Technology (MIT) who have participated in previous datathons. Teams will present their findings on Saturday afternoon, and winners for the best research questions and findings will be chosen. Winners will be invited to present their findings Sunday at the 49th Critical Care Congress. All participants will receive details about their teams, questions, and other details in late January 2020. 

During previous datathons, clinical questions resulted in research published in peer-reviewed publications. Examples are listed Clinician Participants: What makes a good research question? section.
 

Check out these videos to learn more about other datathons that have tackled issues related to critical care. 

What happens at an AIMed datathon?

DAT-ICU 2018 : Datathon for Intensive Care

 

NUS-NUHS-MIT Healthcare AI Datathon and EXPO 2018

Clinical participants will have an opportunity to submit research questions to be answered during the datathon. 
Good scientific research calls for repeatable data and research questions that are formulated to accomplish one of these four goals: 
  • Define or measure a specific fact or gather facts about a specific phenomenon 
  • Match facts and theory 
  • Evaluate and compare two theories, models, or hypotheses 
  • Prove that a certain method is more effective than other methods 
Research questions should address what the variables of the experiment are, the relationships among them, and state something about the testing of those relationships. 

You can also check out this book, A Short Guide to Writing About Science by David Porush. 

For some good examples of research questions, here is a list of studies that have been published as a result of datathons. 

These studies have been published in Critical Care Medicine, the official journal of SCCM:  These studies have been published in other journals: 
Participate in pre-Congress educational sessions. Each session is packed with essential clinical information to keep you well informed on various critical care topics. 

Registration Details
Registration for pre-Congress education sessions can be done in conjunction with your Congress registration. Visit the Congress registration page.

If you are already registered for Congress and would like to add a ticketed session, please modify your existing Congress registration. To modify your registration, log in to your account in the SCCM store. Once logged in, select "All Events." Select the "Details" button from the Congress line item. Once on the Congress event page, select "Edit Registration." Select the items you would like to attend and then "Add to Cart." Follow the prompts to complete your transaction.

If you wish to attend a pre-Congress educational session and not attend Congress, please visit the SCCM store and search by the course title to register.

Course prices vary; view the Congress registration form for details.