About beta band data loss

[Mohammad Aftab Uddin]

Dear Physionet,

Thank you for the new challenge 2023 and I am excited to participate on this one too. :)

Now, as you have stated, just wanted to confirm the below statement you mentioned at https://moody-challenge.physionet.org/2023/#eeg-data

All EEG data was pre-processed with bandpass filtering (0.5-20Hz) and resampled to 100 Hz

and the sample code you provided has beta band code

beta_psd,  _ = mne.time_frequency.psd_array_welch(signal_data, sfreq=sampling_frequency, fmin=12.0, fmax=30.0, verbose=False)

If you have already used bandpass to filter out signal data over 20Hz, wouldn't that drop a lot of information, already? 

Please ignore if this question sounds novice, I really am one.

Thanks

Aftab

[PhysioNet Challenge]

Dear Aftab,

Thanks for asking this question -- it is a very reasonable question to ask!

We consulted with the team that sourced and shared the data, and they did not actually use a bandpass filter on the data. They may have been describing part of the example code, which considers different frequency bands, but the teams can decide how much (or how little) of the example code to use for their approaches.

We have removed this text from the webpage, and we will replace it with a fuller description of the data pre-processing steps. We try to socialize and "beta test" the Challenge, including the Challenge data, during the unofficial phase, and we very well may update the data in response to participant questions and feedback.

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.