The motivation for that is (actually _was_) as follows:
1. That one night that I spent in my lab was a split night report. The first part was diagnostic. It reported my various apneas, RERAs etc, Heart Rate, Pulse, Arousals, snoring etc.
2. Lots of these parameters are also available to me now with the CPAP + my apple watch (HR+Pulse+body movement)
3. As I wrote, I am interested in building a predictive model that given my CPAP data attempts to tell me how I feel. This part is attemptable (do-able or not, I don't know) by building a dataset of pt 2 over several weeks
4. I am also interested in finding out, which parameters or set of parameters affect my "how I feel" state the most, via the model, assuming step 3 works (if its way off the results, this is pointless)
5. Let's now make an assumption I have those parameters (principal parameters that affect my state). I can then target treatment specifically for those parameters and see if CPAP is the only device that can help, or if there are other treatments to those specific principal components that may _also_ alleviate my condition (either even better, or equal) in which case I don't need to use a CPAP every night and I get better sooner. Obviously, in consultation with my doctors. My doctor, like many other doctors takes a macro level view (oh, the lab report said your AHI is 20.9, lets use CPAP) unless he has specific data to see that may point to other conditions. My goal is to generate that data and involve him and take his professional advice. If he disagrees, fine. I don't plan to make changes myself without professional advice (contrary to what some folks wildly assumed here).
6. My next goal was to see how much effect the CPAP sessions have had on improving those parameters over my period of tests. For being able to measure that, it would have been nice if I could use the diagnostic part of my sleep result as 'this is where I was, sleeping normally before I started CPAP'. This is the reason I wanted a "baseline" which, really, would involve me not sleeping on my back and not in a room that was freezing (the wires and the mask were there, I agree, but I feel the sleeping on the back + very cold room were more influential in the numbers of the report).
Point 6 was before you told me sleep studies are not accurate anyway, in which case it may be moot.
My goal here is to leverage an area I am comfortable in: computer science and not second guess medical science. For that, I plan to engage with my doctor and whoever else in this forum wants to share their perspectives.
Does my perspective make sense?
