SleepyHead Test Build for Mac OS X

Progress report on BUGs:

I just installed SH 0.8.2 on my Mac that's running LEOPARD (not Snow Leopard).

Install went cleaning, but I had to reload the data: It tried to load the old data, but stuff was still not working well. (If I'd known where SH 0.8.1 had stored the data I would have just deleted it.

But---restoring the data WORKED. For the most part.

I have now successfully:

1. imported data from a folder on my Mac's hard drive!
2. imported more than one set of data during the same session!
3. clicked on different days and the detailed data changes and shows the correct day's graphs!
4. clicked on a day without data and then clicked on a day with data and the new day's data shows up!
5. successfully gone back and forth between the welcome screen and the daily data!

But there are still issues:
First, the skew in the tick marks is still there. Here's an example. First image is of a pair of OA's as shown in the OLDER version of SH 0.8.2 on my Mac:


And here's what they look like in NEW SH 0.8.2 on my mac:


And as that last image shows, there's now an issue with some of the graphs not showing up at all or showing up looking funky. Indeed, I've gone back through 216 days of data with wave forms, and I've noticed the pressure, leak, and snore curves no longer consistently show up in the graphical data---and this is a new problem as near as I can tell. Some days all three of these graphs show up; some days two show up; some days one shows up. When a graph shows up, it sometimes looks normal, and other times it looks funky: There's no axis in the y-direction, the data is "all or nothing" or the graph is simply a box with a scale in the x-direction, no scale in the y-direction, and no data plotted at all. Sometimes there's not even an unlabeled box around where the graph is supposed to go. And there's no obvious patterns that I can see on when the graphs look bad vs. when they look good. In the 216 days of data that I just went through, the breakdown
looks like this: So the only "obvious" patterns are that the Snore graphs are always there, but they're seriously messed up about half the time; the leak graphs are either there in correct form or they are totally missing---not even a box around where the data is supposed to be; and the Pressure graphs are usually badly messed up (an empty box most of the time) or completely missing (not even an empty box)

1. Downloaded to a MacBook Pro with Snow Leopard, and installed easily.

2. I was a little puzzled about the time indications in the summary. These are labeled "sleep" and "wake", and a total is given.
Actually, I think they represent the "start" and "end" of the data recording; there can be gaps in this (where there is no valid data -- mask off), and there are times when the mask is on but the person is awake. Thus, the labels "sleep" and "wake" do not seem useful nor accurate.

3. The individual index measurements seem to match my report for Encore. But Encore adds CA, OA, and H to get the total AHI; SleepHead seems to only include OA and H in its AHI. Seems like a lack of consistency, that will confuse people.

Again, many thanks for helping to make this tool available on a Mac!

foo wrote:3. The individual index measurements seem to match my report for Encore. But Encore adds CA, OA, and H to get the total AHI; SleepHead seems to only include OA and H in its AHI. Seems like a lack of consistency, that will confuse people.

SleepyHead includes CA in my reported AHI...from night before last report.

AHI 1.28

CA 0.16
H 0.64
OA 0.48

Totaling 1.28

If yours doesn't then something is wrong. I am not using the Mac build though. I use Windows.

OK, I compared the output from Encore Viewer and SleepyHead more carefully, for the same day.

Encore:
CA 2.2
OA .2
H 1.8
AHI 4.2

SleepyHead:
AHI 3.91
Hypopnea 1.65
Obstructive .21
Clear Airway 2.06

So, the numbers in each set add up correctly, but the underlying numbers are different.

One thing to note: my use on this night had about a 30 min gap where I took off the mask, and then put it back on.
Could that contribute to this difference in reporting?

Just wanted to post that I installed version 0.8.2-2 this morning on the latest version of Lion (the install and first launch went well) - got the pop up message about it wanting to clean up after itself and let it - then reimported my data. One thing I noticed - just before I did this install I had fired up my previous version (sorry, can't remember the version number - I know I installed it back in July) and looked at last nights sleep data and I know there was some "leaks" data in that graph - however, now on the newest 0.8.2-2 version todays leaks graph is blank. Don't know if that is a bug for you or just something gone bad during the data import. I can try to delete my data folder and reimport off the sim if you want.

I just read robysues message a few above mine - sounds like my missing graph is likely the same issue as already reported.

i just can say that this is new for me, but thanks...
cheers,
sally

foo wrote:OK, I compared the output from Encore Viewer and SleepyHead more carefully, for the same day.

Encore:
CA 2.2
OA .2
H 1.8
AHI 4.2

SleepyHead:
AHI 3.91
Hypopnea 1.65
Obstructive .21
Clear Airway 2.06

So, the numbers in each set add up correctly, but the underlying numbers are different.

In comparing my own SH reports to Encore reports, the underlying numbers are often different, but not THIS different: In my data, the underlying numbers in SH are simply rounded at two decimal places instead of the one decimal place in Encore. So if I round the SH numbers for OAI, CAI, and HI, I typically get the Encore numbers. This CAN and DOES lead to the AHI being different occasionally due to roundoff error in Encore.

But that's not what's going on here.

Here's a question that may be useful to jedimark and team for debugging:

If you actually COUNT the number of hypopneas and apneas of each type that show up as flagged in the data and divide by the therapy time in SH and Encore, do you get the indices listed? And do SH and Encore agree on how much therapy time there was for this set of data?

One thing to note: my use on this night had about a 30 min gap where I took off the mask, and then put it back on.
Could that contribute to this difference in reporting?

Don't know. I wouldn't think so. Did you leave your machine RUNNING for that 30 minute gap? Or was the machine turned OFF? Does SH's therapy time and Encore's therapy time differ by more than a minute?

In my data, SH and Encore always agree on therapy time. But then, I only have 3 to 5 minute gaps where I turn the machine off before going to the bathroom.

An alpha build of SleepyHead 0.8.4, packaged as a standard DMG file, is now available at http://sourceforge.net/projects/sleepyh ... %20Builds/ . Please delete your old alpha version of SleepyHeadQt before you install this new alpha version of SleepyHead. If you have any problems with the new DMG file, please report them here.

Remember, this is still alpha code, so lots of things are missing, and it has lots of bugs. This is not ready for daily use, but it would be greatly appreciated if Mac users would give it a workout and report all bugs to Mark.

Mark has completely rewritten the graphing code and made lots of other changes, so please test all parts of the program.

You may (or may not) need to delete your SleepApp folder and reimport your data.

Sleepyhead downloaded to my snow leopard machine. Every thing is clear and visible. Each night the pressure in the field on the lower left shows 3 different numbers. The unit is on Cpap set to 14.It does show a red line for 14,but each day the pressure is different and not as high as 14. I am going to have the machine pressure checked this week.
When you click on a date with no data you see the data from the last day larger in size and no scroll bar.

Bobby, thanks for the feedback. Can you post some screenshots?

I tried this with skitch hope that it's big enough.

Here's one where the lower pressure is very low. There are some old data from last month that do read 14 high pressure. but the low reading is 11. There is no other data that reads this low.

Bobby, yah, that doesn't look right.

Do you have EPR set to 3?

Are you using the ramp? Starting at what pressure?

I'm using EPR 3 but no ramp. My therapy goes downhill if I switch to EPR 2.