2020 News

December 24, 2020: The NIH-funded 2021 Challenge is now open! See below for details. Please read this website for details and share questions and comments on Challenge forum. This year’s Challenge is generously co-sponsored by Google, MathWorks, and the Gordon and Betty Moore Foundation.

We have taken the difficult decision to pause all new applications for credentialed access to PhysioNet until 4th January 2021. We apologize for this inconvenience and we will be working hard to clear the backlog of applications in time for opening again in the New Year. Over the coming months, we will also be implementing changes to PhysioNet that we hope will streamline the process for future applications.

We are delighted to announce that MIMIC-IV has been published on PhysioNet! MIMIC-IV, the latest version of MIMIC, is a database comprising comprehensive clinical information on hospital stays for patients admitted to a tertiary academic medical center in Boston, MA, USA.

Major changes from MIMIC-III include: (1) a modular structure that links core hospital data to multiple data sources, including chest x-ray images; (2) an approach to date shifting that provides approximate year of admission; (3) new sources of data; such as the electronic medicine administration record.

The dataset is available from PhysioNet, and access is managed in the same way as MIMIC-III. If you already have access to MIMIC-III, then you will be granted access after signing the Data Use Agreement in the "Files" section of the project. New users will need to complete the credentialing process first (see: https://mimic-iv.mit.edu/docs/access/ for more details). For detailed guidelines on using MIMIC-IV, see the documentation!

The MIT Laboratory for Computational Physiology is seeking a Post-Doctoral Associate to conduct independent research in health care informatics. The Laboratory is an NIH-supported multi-disciplinary group of clinicians, data scientists and engineers that produced the publicly distributed and growing MIMIC database. It is a rich and open research resource that supports signal processing and machine learning research leading to new knowledge and patient-specific prognostic and therapeutic guidance for critical care. 

The Postdoc will contribute to the design and management of the current and future MIMIC databases, and will conduct multidisciplinary original research together with clinicians. The position provides the opportunity to interact with a world-class laboratory comprised of engineers, mathematicians and clinical staff working at the frontiers of translational medicine and advanced research in the domain of critical care informatics and machine learning. The Research Fellow will both contribute to ongoing research projects and propose innovative new projects suitable for research grant funding.

The ideal candidate has a doctoral degree in science or engineering, or a related discipline to assure high level understanding of the research environment. Experience in relational database development and administration is important, ideally in a medical environment. Competence and experience in a subset of the following is expected: Linux, Python, data management. Knowledge of medical terminology is desirable. Strong interpersonal and communication skills are essential. [Update: this position is no longer available.]

We have invited two “wild card” teams to the 2020 PhysioNet Challenge and increased the resources available for training your models. 

We are now accepting entries for the official phase of the Challenge. For the first time, we are requiring teams to submit their pretrained models and code for training their models. See the full announcement on the Challenge forum and the previous two announcements for details.

We are pleased to announce the release of the HiRID critical care dataset, developed as part of a collaboration between Bern University Hospital and the Swiss Federal Institute of Technology (ETH). HiRID is a freely accessible critical care dataset containing data relating to more than 33 thousand admissions to the Department of Intensive Care Medicine of the Bern University Hospital, Switzerland, an interdisciplinary 60-bed unit admitting >6,500 patients per year. 

We have begun the official phase of the Challenge. Please find a new tranche of data with SNOMED-CT codes as diagnoses. Please note that there are some errors or debatable labels in some of the data. Part of the Challenge will be working out how to deal with these issues. In the next few weeks, we will release more data and reopen the scoring system with a new scoring metric.

We are currently dealing with a high volume of applications for credentialed access to PhysioNet, so please expect significant delays in the review process. We are doing our best to deal with the waitlist quickly, handling applications in the order in which they are received. To help ensure that your application is successful, please remember to:

• Include a copy of your CITI training report (not the certificate).
• Remind your reference to reply promptly when contacted.
• Check your application details are correct before submitting.

We apologize for the inconvenience. Please bear with us during this busy time!

Congratulations to the winners of the 2020 Women in Data Science (WiDS) Challenge. The event attracted 951 teams from over 80 countries who competed on models to predict the outcome of critically ill patients. The dataset used in the study, sourced from the GOSSIS (Global Open Source Severity of Illness Score) Consortium, comprised detailed information on over 130,000 ICU stays. For updates on future challenges by WiDS, subscribe to the WiDS mailing list.

We are delighted to announce the PhysioNet/Computing in Cardiology Challenge 2020 on Classification of 12-lead ECGs. For more information, see the Challenge website: https://physionet.org/content/challenge-2020/

Quick links for this year's Challenge can be found here:

• Registration form: https://bit.ly/37d21iN
• Public discussion forum: https://groups.google.com/forum/#!forum/physionet-challenges
• Rules and deadlines: https://physionetchallenges.github.io/2020/#rules-and-deadlines

More information will be posted on the website linked above (and eventually mirrored on physionet.org/challenge/2020 with a delay as it is available). Please check the Challenge forum for real time updates. Please also post questions and comments in the forum. However, if your question reveals information about your entry, then please email challenge [at] physionet.org. We may post parts of our reply publicly if we feel that all Challengers should benefit from the information contained in our responses. We will not answer emails about the Challenge to any other address.

Many thanks again for your continued support of this event and we hope you enjoy this year's challenge. 

A journal article describing the MIMIC-CXR database was recently published in Scientific Data. The article provides detail regarding the collection, curation, and processing done in order to create the database. The article is open access and available online [1].

The database has also been preprocessed into compressed JPG format images, which have been made available on PhysioNet as the MIMIC-CXR-JPG Database. The database includes labels extracted from the free-text reports using publicly available tools. You can read more about the creation of this resource in our arXiv preprint [2].

Finally, we have created the mimic-cxr GitHub repository for collaborative code development on MIMIC-CXR [3]. The code used to generate MIMIC-CXR-JPG from MIMIC-CXR is available in the repository already. We welcome code contributions from all users, and we encourage discussion of the data via the GitHub issues.

[1] Johnson AE, Pollard TJ, Berkowitz SJ, Greenbaum NR, Lungren MP, Deng CY, Mark RG, Horng S. MIMIC-CXR, a de-identified publicly available database of chest radiographs with free-text reports. Scientific Data. 2019;6.

[2] Johnson AE, Pollard TJ, Greenbaum NR, Lungren MP, Deng C-Y, Peng Y, Lu Z, Mark RG, Berkowitz SJ, Horng S. MIMIC-CXR-JPG, a large publicly available database of labeled chest radiographs. arXiv preprint arXiv:1901.07042. 2019.

[3] https://github.com/MIT-LCP/mimic-cxr

The WiDS Datathon 2020 focuses on patient health through data from MIT’s GOSSIS (Global Open Source Severity of Illness Score) initiative. Brought to you by the Global WiDS team, the West Big Data Innovation Hub, and the WiDS Datathon Committee. Winners will be announced at the WiDS Conference at Stanford University and via livestream, reaching a community of 100,000+ data enthusiasts across more than 50 countries.