12-13-2020, 12:46 AM
Additional analyses of CPAP/BiPap data
Just found out about OSCAR program today. Likely missed it during my last year of searching as OSCAR is also name of a commonly used open source EMR in Canada and one that I use. I've been using SleepyHead for 4 years now and glad to see that OSCAR has taken on further development and maintenance. I was a programmer before I became a doctor and wrote a number of programs to sample oximetry data using a medical pulse oximeter, then made a Zigbee based motion sensor, respiration sensor and oximeter device as part of a design contest which used nightly for several years and acquired a considerable amount of data, but never came up with as nice an overall display of data as Mark Watkins did with SleepyHead. My ancient skills are more in hardware and microprocessor interfacing and feel more comfortable in assembly language and C than C++.
The main thing I'm interested in at this time is to get more details of respiratory events as well as tidal volume and other measures which are clearly in my BiPap's data files but despite a day of poking through code couldn't find out where the raw data decoding was done. What SleepyHead and OSCAR both do well is certainly adequate for getting an idea of how someone is doing on their machine and have recommended SleepyHead to multiple patients so they can get an idea of what's going on when they're sleeping on a night by night basis. Was very impressed that had no trouble installing OSCAR on my Mac and that it had no difficulty in importing my SleepyHead data from my PC that SleepyHead runs on.
My primary interest is looking at higher order data in respiratory rates, apneas and limb movements which can be through either spectral analysis or correlational analysis. Tidal volume can give an idea of stage of sleep and, when wearing my leg accelerometer, noticed that leg movements associated with acceleration of heart rate. Can get an idea of what interesting analyses can be done on physionet.org and CPAP/BiPap data now far more informative than 20 years ago when all that was recorded was CPAP on time for each night. Normally HR and motion not part of CPAP/BiPap data, but can get from taping Android phone to ones leg to get accelerometry and sleeping position data but unfortunately pulse oximeters that are useable all night still fairly pricy (at least in Canada).
Glad to find this site, and again, would like to congratulate OSCAR from taking over from Mark Watkins SleepyHead and the fine job that's been done in porting program to Mac as well as potentially adding in additional features.
RE: Additional analyses of CPAP/BiPap data
Loki, the development team is always in need of individuals with C++ and especially if they can work in the QT environment. If you have an interest in joining the team contact the PM Fred Bonjour by sending a PM to bonjour. I'll be surprised if he doesn't see your post and send an inquiry.
Your 60 series auto BiPAP is supported, and the Philips algorithms have not changed much, although they have made a lot of changes over the years on the code your SD card records. You will be interested to know the respiratory data on Philips is recorded at 5 Hz while the Resmed machines yield a much higher level of detail recording things like respiratory flow and mask pressure at 25 Hz.
RE: Additional analyses of CPAP/BiPap data
I'd like to learn how you use your android phone accelerometer to record leg movement (e.g., what app do you use, how much storage space does a night's recording require) and what you do with that data (e.g., how do you view and analyze it and if you're able to correlate it with oscar data). besides wanting to do this myself, I'd like to add the info to my plm thread.