Recording between the arrest and ROSC

65 views
Skip to first unread message

Liuyi Jin

unread,
Jul 5, 2023, 10:08:16 AMJul 5
to physionet-challenges
Hi,

I'm looking for recordings between the time of cardiac arrest and the start time of ROSC. I see some patients have an estimated ROSC time in hours standing for the estimated start time of ROSC after cardiac arrest (e.g., patient 0634 ROSC is 70, and the first recording is the 4th hour after cardiac arrest). 

However, not all patients have the recordings between the time of cardiac arrest and the start time of ROSC. For example, patient 0286 ROSC is 7, but the first recording of ROSC is the 20th hour, which is later than the starting time of ROSC at 7. From another posted question, I see you answered the ROSC given in the I-Care is an estimated value, and the time length between the time of cardiac arrest and the start time of ROSC for most patients is usually in minutes. This makes me uncomfortable using I-Care for my research or training my model because I cannot know precisely which part of the recording belongs to the part before the ROSC and which part belongs to the part after ROSC.

Another solution to address my concern is whether we have a quantitative metric to determine the point of ROSC, i.e., given an ECG/EEG recording, how do we decide which point is the starting point of ROSC? This for sure needs some medical domain knowledge. If you know, please give me some guidance.

Could anyone help answer my question? Any help is appreciated. 


PhysioNet Challenge

unread,
Jul 5, 2023, 10:09:49 AMJul 5
to physionet-challenges
Dear Liuyi,

The units for the time to return of spontaneous circulation (ROSC) are minutes. They are not hours. If you see part of the website that says that the units for ROSC are hours, then please let us know!

Therefore, for your example, patient 0286 had a ROSC of 7 minutes, and the first recording starts at 20 hours after cardiac arrest. Please see the data section of the webpage for details (and units):
https://moody-challenge.physionet.org/2023/#data

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.
Reply all
Reply to author
Forward
0 new messages