Write a Reply

ODI calculation

Hello, I have a question concerning ODI calculation. In a previous discussion it has been mentioned that the series of ndes*ph variables can include desaturations that occurred in wake time and that it would be possible to calculate ODI if the desaturation events were scrubbed from wake time. Would it be correct to calculate ODI from ndes*ph variables using the ndes*ph/slptime? Thank you!

Thanks for checking out the site. I think you are referencing this old forum thread: https://sleepdata.org/forum/o2-desat-events-for-odi-calculation/

ndes*ph/slptime would be the correct way to calculate ODI, however the numerator may inadvertently include desaturations in wake time that shouldn't be included in the calculation. A better way to calculate may be to use the XML annotation files and then identifying SpO2 Desaturations that occurred strictly in sleep epochs.

Thank you for your reply. Maybe a stupid question: where can I get the XML annotation files? The SHHS files on NSSR website are in .csv format only. Should I try to convert them into XML or is there some other source of XML files? Thank you.

The SHHS polysomnography scoring annotation files are located here: https://sleepdata.org/datasets/shhs/files/polysomnography/annotations-events-profusion

Thanks a lot!