Now accepting Challenge entries

[PhysioNet Challenge]

Dear Challengers,

We are excited to accept entries for the official phase of the PhysioNet/Computing in Cardiology Challenge 2020.

Since the beginning of this year’s Challenge, we have made several significant changes:

1. We substantially enlarged the training and test sets, which now include 12-lead ECG recordings from 7 distinct hospital systems, including the largest collection of public recordings and a new, completely hidden test set for evaluating the generalizability of your classifiers.

2. We introduced a new scoring metric that reflects the risk and outcomes of different diagnoses by awarding partial credit to misdiagnoses with low risks or similar outcomes as the true diagnosis or diagnoses.

3. We are now requiring that teams include their pretrained models as well as the code for training their models in their submissions. We have updated our baseline classifiers in MATLAB and Python as examples, and we again encourage you to use these codebases as templates for your entries.

Despite the challenges of this year, participation in this year’s Challenge has exceeded our participation from previous years, and we are eagerly looking forward to your entries.

Best,

Matt

(On behalf of the Challenge team.)

https://physionetchallenges.github.io/
https://physionet.org/

Please 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 it. We will not answer emails about the Challenge to any other address. This email is maintained by a group. Please do not email us individually.