The 2024 PhysioNet Challenge

We are pleased to announce the George B. Moody PhysioNet Challenge 2024, an opportunity for researchers in signal and image processing, and the machine learning communities to advance cardiac care.

This challenge seeks open-source algorithms to digitize and classify electrocardiograms (ECGs) from images or scanned paper, addressing a significant need in areas where digital ECG technology is scarce. This initiative aims to enhance diagnostic capabilities worldwide by training state-of-the-art ECG diagnosis software on historical and globally available hardcopy ECGs, which are at risk of deterioration. This year’s training dataset also incorporates data augmentation with synthetic data and realistic artifacts. MATLAB and Python sample codes are provided, along with tools for synthetic ECG image generation from publicly available ECG time-series.

Whether you have a background in signal and image processing, computer vision, machine learning, or deep learning, the challenge will be intriguing for you! Participation in the unofficial phase is necessary for eligibility in the official phase and prize competition.

 increase engagement from African researchers, the IEEE Signal Processing Society's Challenges and Data Collections Committee is supporting this year’s Challenge with additional prizes for participating teams from Africa and a hackathon at the annual Data Science Africa meeting. For full participation details, evaluation criteria, and prize information, please visit the official challenge webpage.

Please join us in shaping the future of cardiovascular health!

The PhysioNet Challenge team

Email: info@physionetchallenges.org 
Google Group forum: https://groups.google.com/g/physionet-challenges
Web: PhysioNetChallenges.org