OXIMETERS

I taped up the cable without testing it and it worked first time - and I didn't even have access to the cable end to see what the exact colors were - it was cast in plastic so I just cut into it and then figured out which wire I had - and I was lucky - I did not cut into a wire I did not need to cut. It turns out I've done this kind of crap a lot - way too much. What I could not find was an exact list of what is what crossovers are in a 9 pin null modem (and I was too lazy to ohm mine out) so I went by the diagram in the N-200 book, sort of, so that I could more or less do what I finally decided that they probably wanted me to do.

What I really did not understand was that generally, printers are DTE so that they can be attached to a modem (traditionally, few ever were) and you used a special cable - there are some modes on the N-200 where you are supposed to be talking to a printer - and I thought that it would swap the connector into DCE mode and then I could use the straight male-female.

If you find that there is anything to any of this conjecture, please let me know. Right now, I believe that is all conjecture and they handled all conversion with the "right" cable.

Ruby seems a good choice for a monitoring program - you can encapsulate the read object into something that fixes the timestamps, for example, and if it turns out to not be complex enough you can encase that code. The question I have is mysql or simple serial files. I tend to like the latter - because I think they are more reliable and flexible - unless you want access to the data while you are posting it. The biggest issue I see with ruby is that it is a GC language and an unplanned GC might take more than 10 character times (or however large your input buffer is). You know how to deal with that - you plan GC and run it when you can predict that you have time to deal with the GC time. A good article about Ruby GC is here.

Of course, there is one more possibility - the overnight data capture unit - The laptop sits by the bed and makes a secure connection on the Internet - raw data is sent across an encrypted connection identified only by a secret session number which is elsewhere associated back to the patient.

The central collector would certainly want a journaled database to record additions in case of failure.

At this central point, data can be checked by technicians who can phone someone up who is having trouble getting their clip attached. But I'm sure someone is doing that already,

As usual, I was wrong, Holding the low sat and high rate buttons first gives you the opportunity to set the two digit year (0-99) and then the month and day, then the hour and minute. While holding, use the dial to adjust. Release and push again to set the next "thing". I am on the correct minute now. You don't need a gadget - and I am not sure where I picked up that misinformation.

I've been able to get the trend data output easily enough (just type a "W" to hyperterminal) and the event output an (X) to hyperterminal and now I'm looking at other formats. Clearing and clock setting, so far as I can tell, is only done from the front panel.

In many ways, the so-called beat-to-beat format (at every beat, output sat and rate - but the computer program has to apply time heartbeat by heartbeat) might be the simplest of schemes to deal with, but there is good data being recorded into the device's easily accessible internal memory. If you can deal with five second averages over 12 hours and events being recorded for, say, some number of desaturations, it might be possible to get a lot of useful data out of this thing and not to have to hook it to your computer overnight.

I was just going to say that.....

Severeena, I'm not sure who you are going through to get your test, but I would RUN, not walk, to another Dr. / Sleeplab to get appropriate treatment. One month is too long, in my opinion, to put a patient off for testing for this condition.

Having said that, a pulse oximeter can provide some good information regarding apnea/hypopnea events. As I'm sure you are aware, it is by no means a complete picture of whats going on and gives no indications regarding a whole host of sleep disorders.

I also bought a Nellcor N-200 from E-Bay and use it along with the data from EncorePro/MyEncore to evaluate my treatment. For those who are soldering/wire stripping challenged, the cables can be ordered online from somewhere like cables to go. It will likely take two off the shelf cables, a 9 pin null modem cable and a 9 pin male-male cable. So far I've used a program called RealTerm, http://realterm.sourceforge.net/ to capture data. This program has the advantage that it is designed specifically for data capture and device control and can easily be kicked off completely from the command line or a shortcut to capture data to a file.

So far, I have been using the beat-to-beat format without a "real" time stamp. As Brasshopper said, this gives the O2 saturation and pulse rate once per detected heart beat. I also put the unit into fast averaging mode which I beleive is a 3 second average. I believe you hold the high sat and low sat and turn the dial until 2 is displayed to get this mode. This gives me a file with fixed length strings that I can easily import into Excel or OpenOffice Calc to graph. This gives a pretty clear picture of whats going on, but there are a few drawbacks. First, as already stated, you get no real correlation to absolute time. Second, if the probe gets displaced or shifted and loses O2 data, you get a single 0 level output line until it picks up again whether this is 1 second or 3 hours.

I haven't played with the trend data or the internal memory yet. The N-200 is capable of printing graphical data directly to ancient serial printers that support serial printer graphics format.

Two points:

1. Send the program the letter W. It will respond with what the contents of its so-called trend memory is - that is, as I understand it, a line for every five seconds it has been observing, to the limits of the memory.

2. Would you be interested in a simple program that read the beat-to-beat (x02RnnnSnnn\r\n) and put it out as
hh:mm:ss.sss,nnn,nnn
so that it would have a good solid timestamp? It bothers me not to share software, so long as someone can make their own cable.

Thanks!

I've played with the other formats some. Since I don't mind having the computer running along with the pulse-ox, the beat to beat seems to me to give the best information. As time goes on, I might play with the other info some more.

I would definately be interested in the program. I've been searching for a caputure program that would add a time stamp and haven't been able to find one so far. I had thought about writing something but I just haven't had time.