Four new tranches of data with SNOMED-CT code

[PhysioNet Challenge]

Dear Challengers,

We are releasing four new tranches of 12-lead ECG data with SNOMED-CT coded labels to complement the two previously released databases. Altogether, six databases are now available:

A. 6,877 recordings from China Physiological Signal Challenge in 2018 (CPSC2018):  https://storage.cloud.google.com/physionet-challenge-2020-12-lead-ecg-public/PhysioNetChallenge2020_Training_CPSC.tar.gz 
B. 3,453 recordings from China 12-Lead ECG Challenge Database: https://storage.cloud.google.com/physionet-challenge-2020-12-lead-ecg-public/PhysioNetChallenge2020_Training_2.tar.gz
C. 74 recordings from the St Petersburg INCART 12-lead Arrhythmia Database: https://storage.cloud.google.com/physionet-challenge-2020-12-lead-ecg-public/PhysioNetChallenge2020_Training_StPetersburg.tar.gz 
D. 516 recordings from the PTB Diagnostic ECG Database: https://storage.cloud.google.com/physionet-challenge-2020-12-lead-ecg-public/PhysioNetChallenge2020_Training_PTB.tar.gz
E. 21,837 recordings from the PTB-XL electrocardiography Database: https://storage.cloud.google.com/physionet-challenge-2020-12-lead-ecg-public/PhysioNetChallenge2020_PTB-XL.tar.gz
F. 10,344 recordings from a Georgia 12-Lead ECG Challenge Database: https://storage.cloud.google.com/physionet-challenge-2020-12-lead-ecg-public/PhysioNetChallenge2020_Training_E.tar.gz

The first two of these databases have been updated with SNOMED-CT codes. Information about these databases can be found here: CPSC2018. The second database is unused data from CPSC2018 and NOT the CPSC2018 test data. The next three of these databases are previously posted public datasets. Information on their composition can be found here: St Petersburg INCART Database, PTB Diagnostic ECG Database, PTB-XL Database. The recent appearance of the PTB-XL database has granted us the opportunity to vastly increase the scale of the Challenge, and we are enormously grateful to all of the contributors to each database. The sixth database is entirely new, posted for this Challenge, and represents a unique demographic of the Southeastern United States. 

We have sequestered a seventh and final private database (not listed above) along with representative samples from each training database.

In total, 43,101 labeled recordings of 12-lead ECGs from four countries (China, Germany, Russia, and the USA) across 3 continents have been posted publicly for this Challenge, with approximately the same number hidden for testing, representing the largest public collection of 12-lead ECGs.

Please note that there are bound to be some errors or debatable labels in each database. Although we have updated some of the data and labels from the unofficial period of the Challenge, many errors will persist. Part of the Challenge is to work out how to deal with these issues. Some databases have human overread machine labels, and some have single or multiple human labels, so the quality will vary, as well as the demographics and diagnoses. 

Please note, too, that the latest data includes many new classes.  We will only be scoring your algorithms on classes that are captured in the scoring function, which we are in the process of revamping.

In the next few days, we will be reopening the scoring system, and posting an updated scoring metric. Thank you for your patience as we have assembled this database. 

More soon! 

The Challenge team.

https://physionetchallenges.github.io/

Please post questions and comments in the forum. However, if your question reveals information about your entry, then please email chal...@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. 

---

https://physionetchallenges.github.io/2020/

www.PhysioNet.org

[Joachim Behar]

Dear Challenge organizers,

Thank you for the updated databases and instruction.

It is important to clarify the following:

1. whether the test set databases will exclusively include short 12-lead recordings (as mostly represented in the training set) or whether it will include long continuous recordings for minutes to hours (e.g. INCART). The ML strategy might differ significantly depending on that.

2. what classes will be considered in the prospective score. 

Kind regards,
Joachim

Please post questions and comments in the forum. However, if your question reveals information about your entry, then please email c...@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. 

---

https://physionetchallenges.github.io/2020/

www.PhysioNet.org

[PhysioNet Challenge]

Hi Joachim,

Thanks for the opportunity to clarify :)

As noted in the previous announcement, all test data will be standard 12 lead ECGs collected during clinical practice.  

To clarify the particulars, these are usually between 30 and 60 seconds, although you should write your code to adapt to variable lengths. It's unlikely there are extremely long ECGs (say 30 minutes) from which someone makes a diagnosis in the last few minutes.  Due to the large volume of data, we cannot be entirely certain.

All the best,

Gari

---

https://physionetchallenges.github.io/2020/

www.PhysioNet.org

--
You received this message because you are subscribed to the Google Groups "physionet-challenges" group.
To unsubscribe from this group and stop receiving emails from it, send an email to physionet-challe...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/physionet-challenges/80210442-34a0-43ef-86a2-4e0bd4e45f28o%40googlegroups.com.