CPAPtalk.com

Dojers wrote:

GumbyCT wrote: As for Dojers data not showing - how soon did you remove the SD card from the machine? This is not real-time data. The machine needs to first know that you are finished to end the session then do the math. So if you immediately remove the card the data prob won't be there. But then put the card back in and the session and math are complete (by then) then written to the card.

HTH

No, I get up, turn off the machine. Go to the bathroom, get my meds for the day, then when I turn on the computer I get my card and prepare to insert it. Although this morning I DID go quicker which would explain why Sunday didn't show but wouldn't explain why Saturday's still wasn't there. *shrug* It'll get figured out, I got faith.

That is prob not enuff to end the session. While I don't know exactly what ends a session for a PRS1, I do know for the earlier machines it took >59 min. idle, a power interruption, and other things I can't remember.

So try cycling the power or waiting 60 min. Just look at the display to be certain the machine is NOT writing to the card while you are removing the card or cycling the power.

jedimark wrote:

nudave wrote:Not sure if this feature has been requested yet or not...but someway to highlight the waveform and in the bottom status bar give the length of time that's selected.

Hmm.. That be a good idea.. plus it's extra easy to do

The status bar isn't used for much else on the left hand side, so this probably is a very good purpose..

I guess show it in [M]MM:SS:mmmm ?

Question is what happens when Oximeter stuff is selected, which is not synced with the other graphs?

I could just show the time of whatever last area was selected..

Great! Just will save me some calculations in my head..and be more accurate as to the length.

Thanks Jedi!

The Sleepyhead line seems a little thin to me, but maybe that's just my monitor config. Having a 1600x1200 display makes a number of things appear small. Not a big deal either way. It also gets more visible once the line gets longer.

The SPO2 lines probably stand out more because of the light on black display. Maybe it's better to leave it as is.

SPO2 greyed out "bad data" areas on the right.

I don't care about the page turn thing either way. It just looked like a bug to me.

More thoughts:

Start time, end time, duration on window would be nice.

More "regular" scale markings on flow, pressure, etc. would be nice. i.e. 10, 12, 14, 16 instead of 10, 13.3, 16.7 on pressure.

Hmm.. Your screen is huge.. I haven't tried the oximetry stuff yet on my 24" (1920x1200).. I've done all the work on my notebook screen. The version of spo2 I used (well, it was spo2 review) had huge ugly think lines.. That one you just showed me isn't too bad.

I changed the pulse & spo2 to follow the top scale.. Also, I made it autoscale.. even though I said it was difficult. (it wasn't )
Still updates in spurts (I tried full updates, but it was too much realtime data to draw.. slow..).. But I think it's alot nicer now with the scaling.

I can thicken it up.. It might have to be resolution dependant.. Vertical resolution is the main kicker here.. I think if it's over 800pixels high it can thicken some of the lines a bit.

The scale rounding thing is incredibly difficult.. Been through a lot of iterations trying to get that one right.. The thing that helps best is a start and end of the scale being neat, and a evenly divisible number of markers. The autoscaling will help a bit with it anyway..

"Start time, end time, duration on window would be nice. " <-- I can put this in the status bar along with the selection time.. May or may not be pretty. Either that or a dedicated panel down the bottom, but this would rob netbooks of prescious screen space.. (which the status bar does already)

Thanks jedimark. You have done a great work.

I've an idea about the 'no internal clock of CMS50D+'.
You could make it possible to adjust the time scale of the oxymeter data. The user can input the time scale manaully or the program can do the automatic alignment based on the predefined deviation(Ex. I set the deviation is 10 seconds which means after 10s the CPAP start to work, I switched on the oxymeter) by user. So that the data can be fully displayed and viewed.


This is the current situation. Lots of oxymeter data cannot be displayed. The timeline of oxymeter data is not correct.


Again, thanks to your great work. Hope you could solve this problem.

beylow wrote:Thanks jedimark. You have done a great work.

I've an idea about the 'no internal clock of CMS50D+'.
You could make it possible to adjust the time scale of the oxymeter data. The user can input the time scale manaully or the program can do the automatic alignment based on the predefined deviation(Ex. I set the deviation is 10 seconds which means after 10s the CPAP start to work, I switched on the oxymeter) by user. So that the data can be fully displayed and viewed.

This is the current situation. Lots of oxymeter data cannot be displayed. The timeline of oxymeter data is not correct.


Again, thanks to your great work. Hope you could solve this problem.

The recording frequency is 1 entry per second.. This has been verified by others around here, so it's not a scaling problem..

The big gap in the middle is because of the little short play session when you first tested it. (it's merging multiple sessions together)
You can delete the xxxxxx.000/.001/.002 files in the SleepApp machine profile folder.. (it will be the shortest set looking at the size of the .001 file)

If you go into the SleepApp\Profiles\Username\XXXXXXX folder that coincides with your oximeter data, you can delete the short entries to fix the day.. At the moment you have to open the Profiles.xml file in an editor and lookup the hex machine code.. Messy, ugly and horrible.. but this is still in under the hood stage..

One day when I'm less fuzzy, I'll sort the machines into one more directory deep.. Naming the categories.. eg SleepApp\Profiles\Username\CMS50\XXXXXXXX
Will make it a little bit easier finding what's what when hacking around with SleepyHeads data.

CMS50 importer takes the second you clicked on the import button, and this becomes the end time of the data.. I _could_ in theory select the last second of the mornings CPAP record and use that if it's available. You'd have to have a habbit of stopping the CMS50 recording very close to stopping the cpap machine. And it would vary every day..

There's still a seemingly random up to 15 second window with the CPAP machine shutting down.. plus the clock drift of the CPAP machine. On top of that there is a (very roughly) 15 second delay between pulse & spo2 data (heart to finger blood pumping). So basically you never will see it sync perfectly unless you've got the start & switch off down to perfection, plus the nifty programming tool that plugs into where the SD card slot is to set the clock accurately.

There is an another way...

I can make the live view save data.. Live view has accurate time stamps taken from the PC's clock. Get a cheap low power netbook, and wire up for the night.. Your tethered anyway.. just don't pull your netbook off the bedside table

You'll get bonus Plethy data this way, and it doesn't matter when you start or stop your oximeter stuff.. You can yank it off your finger and it will stop recording when the device switches off.. (You still have to click the stop button, but it safely waits for you to do so)

I'll make it so when you click stop, it pops up a dialog asking if you want to save the session to the database.

The only issue will be the drift between your PC's clock and your oximeter.. And here I can set the offset field you are talking about. (It's a set once and check every 6 months thing)

I have devised a method to verify the drift by looking at your computers clock and noting the exact time you stop/start one session.. (write it down before going to bed, and again when you switch off in the morning) you can do the maths and calculate the offset. I'll also do another offset on top of that allows the user to set the bodies SPO2 delay.

Mark,, Thanks for all your effort on this. I did get Qt v0.8 alpha to work. You do say that it supports the Contec CMS50. How do I do that? I have a directory in which the SPO2 software that comes with the contec has saved the data from previous nights. I tried the file menu to point to that directory which didn't do anything. Is there some secret thing on the menu that I haven't seen?

Victor

PS. I'm a mathematician, and know another mathematician who works in my field who has the same name as you at Penn State University. Are you any relation?

jedimark wrote:Hi all,

Here's a preview of SleepyHead Qt version 0.8. Not a beta.. If your expecting a beta, please don't download this.. this is waaaay too unfinished to be called that.
(perhaps even more unfinished than what you may have seen previously)

I'll post updates relating to this Qt version in this thread. I will not rename the title of this one to make it easier to find.

Click here for where to get the 10.5MB Behemoth, and to view the README
http://sourceforge.net/projects/sleepyh ... es/Testing

Note the word "Testing" in that url. Did I mention this is not a beta?

IMPORTANT: Previous users must delete or move the SleepApp folder in my documents otherwise SleepyHead WILL crash every time you try to start the app.

Supports PR System One, Contec CMS50, and quite possibly ResMed S9 too

Plenty of things wrong with this already: some of the graphs scales aren't correct yet, overview pages info panel is quite empty, preferences and profiles just aren't written yet, Nothing in the extras tab in daily view actually works or saves.. (I left it in hoping to inspire some ideas on what to really put there)

The notes feature works, but might not be a good idea to rely on it till extra's tab is finished.. It uses the same journal machine object as notes.

Graphs still don't reorder yet. They are resizable/collapsible now (except for Event flags, which is now a central overview/control object). This is rather kludgy, but nonetheless better than not having this feature. You can have full screen with just one graph if your persistent enough with the fidgety handles..

Event flags still currently shows Resmed S9 users a few empty lines related to the PRS1..This is on the todo list. ResMed only has H, OA & CA (At least the data sets I've seen yet)

I really need feedback on what to add at this stage..

For PRS1, point the importer at the folder or SD card containing the P-Series folder
For CMS50, point the importer at the directory containing SpO2.ini, SpO2 Review.ini and Data folder
For Resmed S9, point the importer at the folder or SD card containing the DATALOG & STR.edf file, etc..

Have fun.

/Mark

victormiller wrote:Mark,, Thanks for all your effort on this. I did get Qt v0.8 alpha to work. You do say that it supports the Contec CMS50. How do I do that? I have a directory in which the SPO2 software that comes with the contec has saved the data from previous nights. I tried the file menu to point to that directory which didn't do anything. Is there some secret thing on the menu that I haven't seen?

Victor

PS. I'm a mathematician, and know another mathematician who works in my field who has the same name as you at Penn State University. Are you any relation?

To get at your old CMS50 stuff, you need to point the importer at the directory containing "SpO2 Review.ini", "SpO2.ini" and "Data" files.
I think it's hiding under your "Application Data" folder somewhere.. Where exactly I never looked, because I didn't have the SpO2 software at the time I wrote that code.. If the location is consistent across windows versions (relative to the user folder, not C drive), I may add smarts to help find it later on..

To import new CMS50 stuff, plug your device in, and go to the Oximetry tab, check the serial port is correct, and click on Import button on the bottom right once. Wait for it to finish. (I'm suspecting there is still a progress bar glitch with it) I fixed a bug in the source code affecting CMS50E oximeters today, there is still a bug calculating the length.. These changes aren't in the current windows build yet. If this process doesn't work, you'll have to do it via SpO2 Review software, and import the other way..

I have a fairly common name, and no American relatives that I know of on that side..

I love maths.. I suck at it unless a computer is involved though.

I've saved PRS1 data into a directory on the hard disk in my computer.

When I go into the import menu, most of the directories on my hard disk were missing. I have several disks, and some disks had a few directories, some had none.

I closed the program and restarted and most of the directories are back now. I've seen this more than once.

First, "Could'nt" is spelled wrong. Unless it's spelled differently in Oz.

I had debug info on. I notice these warnings. Since I'm about to mention another data problem, what do these items mean?

Ignoring short session file "G:/PRS1/110719/P-Series/P02480914C29C/p0/0000000539.001"
Warning: PRS1Loader: Could'nt open summary file "G:/PRS1/110719/P-Series/P02480914C29C/p0/0000000539.001"
Warning: PRS1Loader: Could'nt open summary file "G:/PRS1/110719/P-Series/P02480914C29C/p0/0000000565.001"
Warning: PRS1Loader: Could'nt open summary file "G:/PRS1/110719/P-Series/P02480914C29C/p0/0000000603.001"
Warning: PRS1Loader: Could'nt open summary file "G:/PRS1/110719/P-Series/P02480914C29C/p0/0000000679.001"

I left my card in my PC and missed waveforms on a session or two. After I put my card back in, it copied the other data to the card. I mention this, but I'm not sure it's the cause of the problem below.

Look at the following overview screen:



Note the inconsistency between the date labeling on the two lines and what happened to July 16.

There are also some hours missing. Here's the encore info for the same time frame.

archangle wrote:First, "Could'nt" is spelled wrong. Unless it's spelled differently in Oz.

I had debug info on. I notice these warnings. Since I'm about to mention another data problem, what do these items mean?

Ignoring short session file "G:/PRS1/110719/P-Series/P02480914C29C/p0/0000000539.001"
Warning: PRS1Loader: Could'nt open summary file "G:/PRS1/110719/P-Series/P02480914C29C/p0/0000000539.001"
Warning: PRS1Loader: Could'nt open summary file "G:/PRS1/110719/P-Series/P02480914C29C/p0/0000000565.001"
Warning: PRS1Loader: Could'nt open summary file "G:/PRS1/110719/P-Series/P02480914C29C/p0/0000000603.001"
Warning: PRS1Loader: Could'nt open summary file "G:/PRS1/110719/P-Series/P02480914C29C/p0/0000000679.001"

Thanks.. I forgot I've got to spell better in debug code if users can see it

the .001 files only contain a summary, when Sleepyhead can't load this file because it's dodgy (due to what is looking like a firmware bug), SleepyHead recalculates all the summary data.. So it doesn't really need to use it.. (It does also store machine settings, like what the humidifer is set to, etc)

The summary contains things like counts of events like Obstructive, Clear Airway, Hyponeas, etc.. and also some Pressure average calculations.. The averages it contains aren't very accurate, as they only have one decimal place.. I recalculate all those.

--

I just discovered the cause of a waveform sync error that was making some events not match up.. This has been fixed in the code.. (It ran deeper than I first thought.. the sync problems can happen anywhere.. I have now fixed it "properly")

Now SleepyHead events will match even more accurately..

archangle wrote:I left my card in my PC and missed waveforms on a session or two. After I put my card back in, it copied the other data to the card. I mention this, but I'm not sure it's the cause of the problem below.

Look at the following overview screen:



Note the inconsistency between the date labeling on the two lines and what happened to July 16.

There are also some hours missing. Here's the encore info for the same time frame.

Yes.. Noted..

Part of the problem comes from the ordinary BarChart skipping empty days.. This skip feature may need to just die... the code is ugly and crufty. Which means in the interest of consistency, plain barcharts may end up having gaps on days where there is no data to make scrolling/panning work neater. This looks extra dodgy when it can't fit a date for every bar, it switches to the standard date ticker line charts use.

The other issue, which is the main reason it's messed up here is the Session Times display is nowhere near finished yet..
For some dumb reason it won't show the far right bit of data..

EDIT: Also, some short sessions were turned off because they looked ugly.. (Do you really want to see <5 minute sessions here?)

I still don't like the entire overview page.. I need to rethink it.. I'm not entirely sure whether it should be interactive at all. (If it stays interactive, it's way too complicated slow and ugly to update the summary data when the graphs are panned)

I've been tinkering elsewhere, as there's lots of nastier parts that needs work.

jedimark wrote:
Part of the problem comes from the ordinary BarChart skipping empty days.. This skip feature may need to just die... the code is ugly and crufty. Which means in the interest of consistency, plain barcharts may end up having gaps on days where there is no data to make scrolling/panning work neater. This looks extra dodgy when it can't fit a date for every bar, it switches to the standard date ticker line charts use.

The other issue, which is the main reason it's messed up here is the Session Times display is nowhere near finished yet..
For some dumb reason it won't show the far right bit of data..

EDIT: Also, some short sessions were turned off because they looked ugly.. (Do you really want to see <5 minute sessions here?)

I still don't like the entire overview page.. I need to rethink it.. I'm not entirely sure whether it should be interactive at all. (If it stays interactive, it's way too complicated slow and ugly to update the summary data when the graphs are panned)

I've been tinkering elsewhere, as there's lots of nastier parts that needs work.

I agree that you should show "no data" days.

I'd prefer to see <5 minute sessions. Sometimes I'll turn the machine on and do cpap when I'm wanting to look at the data to try something out. Or to test that data collection is working. It might confuse the newbies if the test session they did doesn't show up.

I do love being able to scroll and zoom the overview data. It's a big improvement over encore. We probably need some indication as to what time period the statistics cover. I could live with the statistics disappearing when you scroll/zoom and having to hit a "calculate statistics" button.

This is already a great program. It already beats the heck out of Encore and EncorePro. (Is that "damning with faint praise?")

archangle wrote:
I agree that you should show "no data" days.

I'd prefer to see <5 minute sessions. Sometimes I'll turn the machine on and do cpap when I'm wanting to look at the data to try something out. Or to test that data collection is working. It might confuse the newbies if the test session they did doesn't show up.

I do love being able to scroll and zoom the overview data. It's a big improvement over encore. We probably need some indication as to what time period the statistics cover. I could live with the statistics disappearing when you scroll/zoom and having to hit a "calculate statistics" button.

This is already a great program. It already beats the heck out of Encore and EncorePro. (Is that "damning with faint praise?")

An awful lot of short sessions get ignored because they generally break things. There is a threshold that can be lowered though, it's currently set to 5 minutes.. I may add an option later on to select what you view as a useless session. Codewise, allowing less than 30 seconds is a bad idea. I forgot why.. but it was a good reason. :}

I need a bit more convincing on the overview stuff.. Perhaps I'll feel differently once the day skipping code is trashed, and the silly edge case bugs are fixed..

--

I just reworked the core database load/store routines.. I removed my cruddy binary file packer and used Qt's built in one.. This means the next build will definitely require a reimport of all data.. I did fix the Purge feature, so it will give a warning message and offer to do it automatically, so shouldn't have to folder hunt the SleepApp folder.

It would be a good idea to use SpO2 Review to import your CMS50 data first before playing with it in SleepyHead until the next version (otherwise you will lose oximeter data)