Hot to distinguish between Samitrop, ptbxl and code15

[Bjørn-Jostein Singstad]

Dear Organizers

How can we distinguish between Samitrop, ptb-xl and code15 when you are running our code? As far as I understand, we are not able to change any of the scrips that reorganize the data and there are no particular naming convention that allows us to easily distinguish between them.

Best regards

Bjørn-Jostein 

[PhysioNet Challenge]

Dear Bjørn-Jostein,

Good question.

For the training set, the WFDB header files generated by the data preparation scripts contain the source of the records, e.g., with a line of the form # Source: CODE-15% for data from the CODE-15% dataset and similarly for other datasets:

https://physionetchallenges.org/2025/#data-formats

For the validation and test sets, the WFDB header files may or may not contain a source, so the run_model script should not require it to label the ECGs.

Best,
Matt
(On behalf of the Challenge team.)

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

[Alejandro Pascual]

Dear  Organizers,

Taking advantage of this question, I would like to know if in the validation and test sets we can make use of the get_sampling_frecuency function (or others like get_age or get_sex).

We believe that in a potential clinical application, this type of information could be helpful for diagnosis, as it is usually easy to obtain. In particular, the sampling frequency is important for resampling signals to a common standard and for establishing a proper temporal reference for the lead vectors.

Could you please confirm whether the use of these functions is permitted during validation and testing, and whether they will return valid values?

Best,

Alejandro from EPBandoleroLab team.

[PhysioNet Challenge]

Dear Alejandro,

Yes, we will provide the sampling frequency, duration, and other information about a signal as well as the age and sex (when available) of a patient, in the training, validation, and test sets. We include them as comments in the WFDB header files, and you can use extract them from the header file or use the provided functions to extract them.

You can decide whether and how to use these quantities. The sampling frequency helps to interpret a signal, but you may find that using it by itself as a feature is not helpful, and may make your approach less robust to data from different sources, or data from the same source that is collected later.

Best,
Matt
(On behalf of the Challenge team.)

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