Just as a geeky aside, I enjoyed learning about how to read ECG traces! There are lots of interesting patterns and general principles involved and I guess it could come in useful in a pinch one day. There are quite a few doctors on social media who share and explain traces and it's fun to figure them out first. Or I'm just a morbidly curious nerd.. :-)
I work deep in the weeds on startups using datasets like these and I'm always curious about how people drive value from this raw data. Who is designing things that use this data for anything interesting in the United States? The whole system is designed NOT TO PAY for these advancements. What value if any exists in this data beyond academic research?
A common model is for startups to use publicly available data to build their first model/device. Those devices then collect more application-specific data and use that to iterate design.
Not hard to see how piles of ecg data could be useful.
One example: ICUs typically monitor a dozen or so of these traces per nurses station, so having an automated system for detecting codes would speed up response and avoid missed cardiac events.
The most obvious task is to predict a diagnosis given the ECG history. Are there any studies out there?
Oh I haven't seen bdsp.io, this appears to be a direct clone of the National Sleep Research Resource's hosted software at https://sleepdata.org/
I'm afraid this is gonna be weaponized for the creation of neural-links by big corps.