Beginning of the official phase of the PhysioNet Challenge 2022
352 views
Skip to first unread message
PhysioNet Challenge
May 2, 2022, 3:31:47 PM5/2/22
to physionet-challenges
Dear Challengers,
We are pleased to announce the beginning of the official phase of the George B. Moody PhysioNet Challenge 2022: Heart Murmur Detection from Phonocardiogram Recordings!
For the unofficial phase of this year’s Challenge, we asked teams to identify whether a human screener found heart murmurs in a patient’s heart sound recordings. For the official phase of the Challenge, we are also asking teams to identify whether an expert cardiologist referred the patient for further screening or treatment based on a physical examination, heart sound recordings, and echocardiograms – but we are asking teams to do this with only the patient’s heart sound recordings! Therefore, we are challenging teams to use the same phonocardiogram data to perform two related but different classification tasks: murmur identification and clinical outcome identification. For each task, the team with the best score wins.
For these tasks, we have updated the data to include both murmur labels (Present, Unknown, Absent) and clinical outcome labels (Abnormal and Normal) along with minor corrections to some of the demographics information in the data. We have also updated the MATLAB and Python example classifiers for these two tasks. Teams must update their code to perform both classification tasks, but many code bases will only need minimal changes to submit to the official phase. We have also simplified the scoring metric for the murmur identification task and introduced a new cost-based scoring metric for the clinical outcome identification task. You may see a couple of additional minor changes in the coming days as we receive feedback about these changes.
Please see the Challenge webpage for details, including links to the new data and a new submission form for when you’re ready:
https://physionetchallenges.org/2022/
Your Computing in Cardiology abstracts are currently under review, and we look forward to sharing the results with you next month.
Best,
The PhysioNet Challenge Organizers
Quick links for this year's Challenge:
- Registration form
- Submission instructions and form
- Public discussion forum
- Rules and deadlines
We are pleased to announce the beginning of the official phase of the George B. Moody PhysioNet Challenge 2022: Heart Murmur Detection from Phonocardiogram Recordings!
For the unofficial phase of this year’s Challenge, we asked teams to identify whether a human screener found heart murmurs in a patient’s heart sound recordings. For the official phase of the Challenge, we are also asking teams to identify whether an expert cardiologist referred the patient for further screening or treatment based on a physical examination, heart sound recordings, and echocardiograms – but we are asking teams to do this with only the patient’s heart sound recordings! Therefore, we are challenging teams to use the same phonocardiogram data to perform two related but different classification tasks: murmur identification and clinical outcome identification. For each task, the team with the best score wins.
For these tasks, we have updated the data to include both murmur labels (Present, Unknown, Absent) and clinical outcome labels (Abnormal and Normal) along with minor corrections to some of the demographics information in the data. We have also updated the MATLAB and Python example classifiers for these two tasks. Teams must update their code to perform both classification tasks, but many code bases will only need minimal changes to submit to the official phase. We have also simplified the scoring metric for the murmur identification task and introduced a new cost-based scoring metric for the clinical outcome identification task. You may see a couple of additional minor changes in the coming days as we receive feedback about these changes.
Please see the Challenge webpage for details, including links to the new data and a new submission form for when you’re ready:
https://physionetchallenges.org/2022/
Your Computing in Cardiology abstracts are currently under review, and we look forward to sharing the results with you next month.
Best,
The PhysioNet Challenge Organizers
Quick links for this year's Challenge:
- Registration form
- Submission instructions and form
- Public discussion forum
- Rules and deadlines
Reply all
Reply to author
Forward
0 new messages