A great place to start Disorder information source Rated and Reviewed Sleep Links Entrance to all monitored Sleep Forums Search all of Sleepnet.com


Sleep Apnea
Forum Posting


Archived Apnea Forum83 viewing only. To post a new topic go to the Apnea Forum Homepage.

Re: Accuracy of Sleep reports

Posted by Kevin (in Walla Walla) on February 01, 2002 at 13:44:05:

In Reply to: Accuracy of Sleep reports posted by tominjax on February 01, 2002 at 04:56:33:

Hi tom,
I'll try and go through the process we use in my Center. We use a digital acquizition and storage system with "computer assisted" scoring. All the analog signals from the patients are transmitted by cable to the control room. Next we amplify and digitize the signals (brainwaves are amplified several million times). The data is displayed on large monitors and recorded to hard drives. We sample at 100 Hz for routine studies. The recording technologists monitor the patients during the night, adjusting the amplifiers for sensitivity, filters, polarity, etc. Also titrate xPAP in response to abnormal respiratory events. Once the patients are awake we can begin scoring the studies. I use a 3 pass system. Our average study length is about 1050 epoches, or ~500 minutes. In the first pass we go through in 30 second screens in order to stage sleep. This involves looking at the frequency and amplitude of EEG, EOG, and EMG signals. Next we go through all the data while compressed, anywhere from 90 to 240 seconds per screen. This is when we score respiratory events by examining the airflow, chest effort, abdominal effort, and oxygen saturation channels. If the patient is on xPAP we also have a PAP flow signal. If its a PES recording we also analyze the flow signal looking for increasingly negative waveforms followed by an arousal. Finally we go through it a 3rd time, back at 30 seconds per screen, looking at arousals and PLM data. I like to wait and do these after scoring the respiratory events because it allows us to then attribute an arousal to a respiratory event, a PLM, a movement, etc.

We use the computer to help in saturation analysis and heart rate analysis, and to cruch the numbers. I've played around with computerized analysis of EEG, PLM and respiratory data, and it actually takes longer because you have to go back through and correct all the things the computer missed, or got wrong.

We generally spend about 2 hours scoring a routine PSG. In really bad ones we can take 4 or 5 hours.

What most concerns me are the places with untrained people recording and scoring sleep studies. If you've go garbage going in, you're going to get garbage coming out.

Follow Ups:


Archived Apnea Forum83 viewing only. To post a new topic go to the Apnea Forum Homepage

  • IMPORTANT : Information not intended as medical advice. If you suspect that you have a sleep disorder you should seek care from a qualified professional. Read Terms of Use.
  • The Sleep Forums are not to be used for commercial purposes.
  • Commercial products and services are not endorsed by Sleepnet.com.
  • Sleep Deprivation due to Sleep Apnea and insufficient sleep are common and can present as insomnia, narcolepsy, or idiopathic hypersomnia. In infants and children sleep problems commonly present themselves as ADD or ADHD.


Copyright ©1995-2005 Sleepnet.com., All rights reserved