Introduction to Polysomnography Basics

Hi, this is Tom Tolbert. I work in the Sleep Medicine Department at Mount Sinai, and this is an orientation to polysomnography, to PSG, for people who are just starting to learn how to read sleep studies. In our lab, we use two programs to read sleep studies, either Compumetics or Minerva, and the examples I'm going to be giving here are through Minerva, but right here is a fairly standard display of an interface you might see in order to read a sleep study, and it is intimidating. There is a lot of information here, a lot of signals, a lot of stuff that's hard to understand the first time you look at this. It's easier when you remember that, in general, these interfaces are often divided up into three panels. And each panel has a different time scale. At the top is the sleep architecture for the entire study. So from this side to this side, this whole thing is the entire study. That's the whole length of the study from beginning to end of the night. And this is called the hypnogram, and it's sort of a graphical, pictorial, schematic representation of sleep architecture. It's a continuous line that moves up and down depending on what stage of sleep or wakefulness the patient is in at the time. So up here at the top you have wake, and you can see at the beginning of the study here there's wakefulness. And then there's stage N1 and N2. And this patient had some unstable sleep, so there was a lot of movement sort of wiggling around between stage N1 and N2. And then delta sleep, or stage 3, you can see some of that right here. And then also REM sleep is down here at the bottom. Sometimes it's actually at the top. These things can move around on different hypnograms. Right here, there's stage REM sleep. And then sometimes, not always, but often there's a position indicator on the hypnogram just showing what position they might be in. In this case, this blue line means supine sleep. Under that panel is another panel that's usually used for sleep staging. And this is often on a time scale of 30 seconds because by convention, sleep epics are 30 seconds long. And this panel is used generally to specifically score what stage of sleep each sleep epic is in. Under that panel is usually a slightly longer time scale. In this case it's two minutes. And this panel is used for limb movement and respiratory scoring. And you can see if this is a 30 second epic up here, down here these pink lines separate each 30 second epic. So there's one, two, three, four epics here. 30 seconds times 4 is 120 seconds, that's 2 minutes. So this window down here, this panel, is 2 minutes long. That's not always the case. Sometimes it could be shorter, a minute long, sometimes it could be up to 10 minutes longer. Limb movements and respiratory events like apneas and hypopneas are sort of harder to see on a 30 second time scale, so we tend to use a longer time scale down here. This whole screen is much easier to digest once you realize that you'd Generally only look at one of these at a time, one of these panels at a time, and it depends on what your question is. If you're asking like how was the whole night of sleep, what kind of sleep did they get, did they get a lot of REM, did they get a lot of stage 3, that's when you look up at the hypnogram. If you want to know what specific stage of sleep they're in, you look at the sleep staging panel, and then if you're interested in respiratory events and limb movements, you tend to look at this panel at the bottom. Now like I said, this is in Minerva, this program is called Minerva. There's also Compumetics, but there's lots of different ways to read sleep studies, and they can all look a little bit different just because they change the colors or the order of the channels might be different. And there's a lot of customization available, too. So depending on the preferences of whoever's reading the sleep study, they may move channels around, and you just have to pay attention to what channels you're looking at in order to not get too confused. Let's look specifically at sleep staging, and that's this panel here. Sleep staging requires three things. It requires a look at brain waves, and we do that with EEG. A look at eye movements, which is electrooculography, E-O-G, but just the eye channels is often what I'll call them. And then also a sense of muscle tone, and that's done through electromyography, E-M-G, of the chin. So the way I have this set up right now is off in a view you'll see with the EEG channels on top, and then the eye channels. and then the muscle tone in the chin channel here. But these can be moved around. Like I said, there's a lot of customization available, and some people prefer to have the eye channels on top. So you just have to pay attention to what's over here. to avoid getting confused about what you're looking at. Let's talk about EEG. In the 1960s, Rechstofen and Kales, two researchers, noticed that if you looked at brainwaves during sleep, there were some discrete patterns. And so they called these sleep stages, depending on what patterns they were seeing in the EEG. And conceptually, it's largely the same today. There have been some changes. We've standardized the way we do EEGs a little bit differently, and the sleep staging has changed a little bit. But for the most part since the 1960s it's been the same in that we're just looking for patterns in brain waves that correlate with what we understand about sleep. The standard way to take an EEG is with the 10-20 system and the way this works is you take the head from the nasion right at the front of the head at the forehead and the inion at the back of the head and you divide each half of the head into a hundred percent so this half of the head is 100%, this half of the head is 100%, and then at 10%, then 20, 20, 20, and another 10% positions, so you've divided this half of the head into these sections, you put your leads. Okay, so that's why it's called the 10-20 system, and there's nothing magical about this except that it's a convention that allows standardization of EEG recordings across multiple studies. The nomenclature is... It's pretty straightforward. You have F for frontal, C for central, and O for occipital. So it's front, center, back of the head. And then M is for mastoids. They used to be called ears, but now they're mastoids because you put them over the mastoid processes. By convention, even numbers refer to the right side of the head, and odd numbers refer to the left side of the head. In order to actually see a wave that corresponds with an electrical vector like a polarity or a electrical current, you have to reference these unipolar EEG leads to another lead. So in this case, we take the frontal lead on the right side, that's F4 here, reference it to the mastoid lead on the left side so that we're looking at current along this vector. Similarly with the central lead, it's also referenced to the mastoid lead. And then the occipital lead is also referenced to the mastoid on the left side of the head here. So if you're looking at these three channels like this, you're really looking at the right side of the head with these even numbers from front, center, to back. And what are we actually looking for? Well it turns out it's not too complicated. If you just remember diet tab, you can remember all the different wavelengths that practitioners look for. because we're not looking for anything as complicated as an electrophysiologist might look for. There's really just five sort of wave frequencies we look for. The slowest are delta waves, and they're at a frequency of 0 to 4 hertz, or they occur 0 to 4 times per second, same thing. Each one of these boxes sort of delineated by these faint green lines is a second in this view. So if you count the number of waves in each one of these boxes, you get the frequency. and there's like pretty much half of one of these waves in each one of these boxes here. So this is going at a frequency of about 0.5 hertz or half a wave per second. The next fastest or the second slowest depending on how you look at it are theta waves and these are at a frequency of 4 to 8 hertz. They're sometimes described as sawtooth waves because they kind of look like a sawtooth. Just faster than theta waves are alpha waves and these are at 8 to 13 hertz. If you were to count the number of spikes number of little waves in this one second box here you get something between 8 and 13 Hertz and then faster than alpha waves are beta waves and these are pretty much anything over 13 Hertz. Beta waves are often I think the toughest to see is discrete waves because this is what our brain looks like when we're awake with our eyes open okay so we have these really fast waves and a lot of just noise that all these little neural vectors, these electrical vectors, sort of cancel each other out. So you've got almost like a much flatter line than the other waves that, if you look carefully, has very high frequencies in here. So I got my mnemonic a little bit out of order the way I've put it on the slide here, right? So it's D-T-A-B. But why is this backward? Well, you see both alpha waves and beta waves in wakefulness. Okay, so these are both wavelengths, wave frequencies that you will see. when a patient is awake. The difference is eye closure. When you close your eyes, and this is eye closure right here, we'll talk about the eyelids in a sec, but just take my word for it, this is the patient closing their eyes. That's what these wavelengths are right here. When you close your eyes, you stop processing visual information in the occiput, in the occipital cortex on the back of the head. And when you stop processing visual information, there's less neural activity, meaning there are less vectors to cancel each other out. And so you start to see the emergence of this background pattern, these alpha waves. So instead of having just a lot of noise, you can see these little waves pop out. And about 90% of people do this where when they close their eyes they produce this alpha rhythm at the back of their heads from that phenomenon. And so this is sometimes called a posterior dominant rhythm. So this whole period here is wakefulness, but right here the eyes are open and when the eyes close you see these alpha waves appear. Now the S stands for sleep spindles, also called sigma waves, which are kind of these... little phenomenon, these discrete little waves that happen over just a short period. They just kind of appear and go away just as quickly. And they're at 11 to 16 Hertz and they characteristically look like a spindle, like string wrapped around a rod or a peg, in that they taper from small to get bigger and then they taper down again. So you can kind of see that shape here. This is a sleep spindle. So this is pretty much everything. There's one wavelength that doesn't really fit into the diet tabs mnemonic and that's the k-complex which is kind of a special case of a delta wave it's like a delta wave that's isolated that sort of stands by itself and it's actually from 0.5 to 2 hertz so it's just kind of in a narrower range than typical delta waves and by convention negative deflection is up so this is a characteristic k-complex in that It has this initial negative deflection, then a positive deflection, then returns back to baseline. So it has this sort of canonical shape, the normal shape of a K-complex. And if they don't look like this, they could be something else. So don't get tripped up. This is what a K-complex is really supposed to look like. Let's move on to the I-leads or the electrooculography, the EOG leads. These are electrodes that are placed outside the canthus. of either side of the eyes. And by convention again, left is odd, I1, E1, and right is even, E2, I2. The way this is set up is pretty ingenious because eyes are like batteries. They have dipoles. The cornea of your eye is positive while the retina has a relatively negative charge. And that means that if your cornea moves towards an electrode, it's going to cause a positive deflection which is down. If it moves away from an electrode, it's going to cause a negative deflection which is up. By putting the right eye lead above the eye and on the outside, and the left eye lead below the eye and on the outside, it means that any eye movement is going to cause these waveforms to go in opposite directions. So I can tell that these are true eye movements because the waves are basically mirror images of each other. Here's an example, all right? There are these big waves in the EEG leads. How do I know if these come from the brain or from the eyes? Because the EEG leads are close enough to the eyes that they can detect eye movements. Well, I just look down in the eye channels and I see that since these waveforms are going away from each other, they're like mirror images of each other, I know that these are eye movements and the eye movements are being caught in the EEG leads. To give a counter example, here are some big EEG waves again. And if we look down below here, here the eye channels are moving together. They're no longer mirror images. So I know this can't be an eye movement because with a setup that looks like this, eye movements will not move together like this. They'll move apart from each other. So I know that this is truly a EEG wave. Let's look at some actual sleep staging. I'm going to go through just a few example epics here. You saw this one previously. This is wake and this is very fast. noisy rhythm that you see when somebody's awake and their eyes are open. When the eyes close, and you can see this big wavelength in the EEG here, and not really sure what that is, but if you look down you can see that the eye channels are moving apart. They're mirroring each other. So this is actually eye closure here. And I know that in part because after this happens, I see alpha rhythm. And alpha rhythm is what happens when you're awake and you close your eyes. You get this posterior dominant rhythm that you see best on the back of the head in the occipital leads. So when somebody's awake and their eyes are open, often their eyes are going to be moving around a lot and you'll have a very noisy signal. Whereas if their eyes close you'll see this alpha rhythm and most of wake during a sleep study is people with their eyes closed and you see this this alpha rhythm. In the transition from wakefulness to sleep you get a kind of slowing down, a further slowing down. So this alpha rhythm has kind of decreased in frequency and it's harder to pull out waves that are this fast starting right about here. So here maybe there's some theta in there a little bit but there's slower waves. They're not quite as fast as the alpha rhythm here and this is called low amplitude mixed frequency. And this is a sign of drowsiness or even of sleep onset if this continues into the next epoch. Additionally, on this channel you can see the eyes are moving here, but they're moving very slowly. There's kind of curved waves here as opposed to the more rapid things you saw here. This is blinking, this is eye closure, those are pretty quick. Whereas here, this is slow rolling eye movements, or SREMs. And that's a sign of drowsiness and even of stage N1 sleep. So this is stage N1, and again you see those slow rolling eye movements, and you also see low amplitude mixed frequency. And drowsiness and stage N1 sleep are sort of the same thing in a lot of ways, and that's why stage N1 is considered a transitional state. This is the stage you enter first from wakefulness as you start trying to reach those lower stages of sleep like N2 and N3. This is stage N2 and remember we talked about sleep spindles. This is a sleep spindle right here. That's characteristic of stage N2. So if you see sleep spindles you're very likely in stage N2. And then the other thing here and this is not quite as pretty a K-complex but I think this is still a K-complex. You can see there's an initial negative deflection and then there's a positive deflection that follows. So this is a K-complex. and spindles and k-complexes characterize stage N2 sleep. So first you see the low amplitude mixed frequency of stage N1, and as soon as you start seeing sleep spindles and k-complexes, you know you're in stage N2. Stage N3 is also called slow wave sleep or even delta sleep or delta wave sleep, and that's because it's characterized by delta waves. Instead of seeing the discrete single delta wave of a k-complex that we see in stage N2, we start seeing delta waves happening over and over again. They come together in clusters and might even cover the entire epic. But as long as they take up six seconds or 20% of the epic, it's considered a stage N3. Now scoring REM is complicated because it can be difficult to differentiate it from stage N1 because it tends to have kind of this slow, low amplitude mixed frequency signal. It can sometimes have alpha in it, but it This looks a lot like N1 to me. The only way I can tell that it's not N1, all right, is that there's this big eye movement right here. This is a rapid eye movement. There was a quick saccade with both eyes moving to one side of the head. If I look, this is the left lead. So this is actually looking to the right because the cornea is moving away from the left eye. So this is a big saccade to the right, and this is a rapid eye movement. And then there's some less obvious, more subtle. Rapid eye movements in here too One right here. So those rapid eye movements plus this kind of low amplitude mixed frequency And there's even some theta waves in here, which which are also characteristic of REM This is suggestive of REM sleep The only other thing I've got to look for is to make sure the chin tone is low because in REM sleep You're supposed to have your muscles essentially with with very low tone almost paralyzed that can be tough to tell sometimes just by looking at a the chin tone alone like i don't know is this really low tone i'm seeing something here some wiggling so sometimes i'll look back at an epic when the patient was awake so this is like a previous epic you can see this is epic 192 and this is epic 11 right here and there's much more activity in the chin tone here a lot more alpha you can't have alpha and rem but this is a lot of alpha this is overall a faster frequency than you're seeing in rem and there's theta waves here So this is a REM epic with low chin tone. This is a wakefulness epic with higher chin tone. And in both cases, you can have these rapid eye movements. Now, the last thing we look for in the EEG leads are arousals. And arousals are evidence of going from either one sleep stage to a less deep sleep stage transiently or even going completely to wakefulness. And that's signaled by you have what looks like stable sleep. and I think this is either rem or n1, I believe. This could be a k-complex even, so you could argue this is n2, but I think this is probably n1 or rem, because I think these are theta waves here. But anyway, you see... Suddenly there's this huge change in the EEG leads. There's a lot higher frequency here, and there was a big movement that caused a big deflection in the EEG lead. So the patient probably moved their head here, turned it to one side or another, and at the same time there's a big increase in muscle tone. So this is an arousal. Sometimes they're this obvious. A lot of times they're more subtle. You don't see as much actual movement that causes all the EEG leads to deflect like this. Sometimes you just see... increase in frequency. In this arousal here you can see this is probably low amplitude mixed frequency of either I think this is probably N1 because you can see maybe there's some slow rolling eye movements here but suddenly the frequency increases and you get some alpha in here and even some increase in the chin tone so this is a another arousal and there's even some movement of the patient's head probably because you get a big deflection in the EUG leads. And lastly here, this is an example of what looks like stage and three sleep. So these are delta waves. This is slow wave sleep. And then suddenly there's a big increase in chin tone again and an increase in the frequency, also with a movement of the EEG leads. All you need is an increase in the frequency and the amplitude to say it's an arousal. You don't always get this movement in the EEG leads, but certainly if you see something like this, this should make you be thinking there's probably an arousal, an increase in the frequency. the waves here. So that's sleep staging where you looked at your waveforms to score what specific stage of sleep you are for each 30-second epic as well as any evidence of arousals. Let's move on to talk about limb and respiratory scoring and in order to describe the several channels you'll see here I'll give a picture of myself wearing some sleep equipment and blow up the channels right here. So this is the nasal pressure channel that comes from a nasal cannula. I'm wearing right here And by convention, inspiration or breathing in is up. Right here is pretty much what a normal breath looks like. These two right here are pretty normal. They're nice rounded on top. So starting here, this is a breath in, all right? And then down is exhalation. So down is out. This isn't a perfect drawing of exhalation, but it's pretty good. This is what exhalation should look like. So inspiration, exhalation. You can see that coming in and out of the nose with the nasal cannula. Underneath the nasal cannula I'm wearing an oronasal thermistor, or sometimes we just say thermistor, and that is a detector that measures changes in temperature. So as you breathe in, air gets cooler, and when you breathe out, the air gets warmer. And in contrast to the nasal cannula, which can only see breathing in and out through the nose, the oronasal thermistor can see breathing in and out through the nose or through the mouth. So that's why it's called an oronasal thermistor, because it detects both nose breathing and mouth breathing. Why do you need both? Well, one is to detect mouth breathing, right? Because if you don't see any flow on the nasal channel, you have to make sure they're not just breathing in and out through their mouth by looking at a thermistor. But the other is, the thermistor channel is essentially qualitative. All it can tell you is if a breath is present or not. It can't really tell you how big a breath is. So you need the nasal pressure channel to see how big breaths are, whereas you'd use the thermistor channel just to say if breaths are there or not there. So one way to put it is that to score apneas, you look at the thermistor and you're looking for just nothing. Whereas to score hypopneas, you look at the nasal pressure and you're looking for small breaths. Breaths that are smaller than the normal size breaths. As you can see down here, I'm also wearing a band around my chest. This is for respiratory inductance plethysmography. That's just a fancy way of saying that this band measures changes in the volume of my chest as I breathe in and out. Same thing on my abdomen here. These are just what are called effort channels, and these waves up and down show evidence of inspiratory effort. So you can tell that I'm attempting to make breaths, and since I have flow on the thermistor and the nasal pressure, that I'm actually getting breaths in and out. So here's a patient who's having pretty severe obstructive sleep apnea. An apnea is defined as basically an absence of any breathing for a period of at least 10 seconds. And you can see here these pink lines again are dividing each epic. And then the fine green lines here in this wider view, they're 10 seconds each. So you can see that this absence of flow right here is much more than 10 seconds. It's about 20 seconds, maybe even more. And you can see there's just no flow in the nasal channel. And then there's very little change in temperature in the thermistor channel. Now whether it's an obstructive apnea, or a central apnea depends on what you see in the effort channels. So that's the thorax and the abdomen here, the size change in the chest and the size change in the abdomen. The thoracic channel for this particular patient doesn't look like it's working. It's not. in the right place on the patient's chest probably, but at least the abdominal channel is moving up and down with the breaths here, and you can see that it's not moving as much here, but there's still some movement that looks like attempted breaths. So that's evidence of effort, and that implies that up here is an obstructive apnea. If these events have already been scored or tagged by a sleep technologist, a sleep scorer, you'll see the events displayed kind of like this, usually with boxes drawn around them and often they'll be already classified as either obstructive or central. So these are obstructive apneas. Apneas don't require any consequence to be called apneas. That is if you just don't have breaths for 10 seconds or more, if there's just no flow and no change in the thermistor channel, that's an apnea. Nothing else needs to happen. Hypopneas on the other hand are a little more subtle, a little more complicated. Because they require a consequence. They require that something happens or is attributable to the hypopnea. I'm only going to talk about the American Academy of Sleep Medicine recommended criteria for scoring hypopnea, but there are other ways to score hypopneas. But the AASM definition of a hypopnea is a discrete reduction in airflow to 70% or less of baseline airflow that lasts at least 10 seconds. So... You know, we're seeing big breaths here, big breaths here, pretty big breaths over here. And then right here in this area, there's a kind of discrete reduction in flow. So I would say that there's a hypopnea right here, or at least for these four breaths right here, for sure. And this probably is at least 10 seconds. You can see, yeah, here to here is supposed to be 10 seconds. So this is a reduction in flow that lasts at least 10 seconds. But that's not enough. Like I said, there has to be a consequence. And the consequence can be either a 3% or greater desaturation. And that doesn't appear to be present here. So we have sort of a maximum of 97%. And get down to a minimum of maybe 95% somewhere in here. Yeah, I guess there's a 95 right here. The other thing that can happen, though, is you can have an arousal that is attributable to this hypopnea. So in order to see if there's an arousal, we have to go up back to the sleep scoring panel. And you can see right here that... that the waves are kind of small and a little slow. They're not as fast as they are over here. So I think what happened here is you were probably in maybe stage N1 here, maybe N2, and then there appears to be an arousal. And that arousal is associated with this reduction in flow. So I would call this a hypopnea associated with arousal, and that makes it a true hypopnea. Now lastly, limb movements. You can't see it in this picture, but on either leg I have a piece. EMG electrode on my anterior tibialis on both sides. And those EMG leads are to detect limb movements, like periodic limb movements. So these are the leg channels. You can see that the left leg is here, the right leg is right here, and you can see that there are spikes, increases in the voltage here that are indicative of movements of both legs in most of these cases. Normally I'd also have a video where I could actually see the patient sleeping, and I can see if there's any evidence of... the limb actually moving on the video. Sometimes you won't see it and there's still muscle tone there it's just not enough to actually cause a movement of the leg but sometimes it's helpful to actually see a limb movement to make sure that there isn't some kind of artifact making this signal. I don't want to go through all the technical details of scoring periodic limb movements because there's a lot of subtlety sort of funny rules to them. This is just something you should be aware of and look for and you'll learn the more technical aspects of them as you do it more. So thanks for watching. I hope that's helpful and hopefully this is enough to get you started on reading some of your first sleep studies.

There is a lot of information here, a lot of signals, a lot of stuff that's hard to understand the first time you look at this. It's easier when you remember that, in general, these interfaces are often divided up into three panels. And each panel has a different time scale.

At the top is the sleep architecture for the entire study. So from this side to this side, this whole thing is the entire study. That's the whole length of the study from beginning to end of the night. And this is called the hypnogram, and it's sort of a graphical, pictorial, schematic representation of sleep architecture. It's a continuous line that moves up and down depending on what stage of sleep or wakefulness the patient is in at the time.

So up here at the top you have wake, and you can see at the beginning of the study here there's wakefulness. And then there's stage N1 and N2. And this patient had some unstable sleep, so there was a lot of movement sort of wiggling around between stage N1 and N2.

And then delta sleep, or stage 3, you can see some of that right here. And then also REM sleep is down here at the bottom. Sometimes it's actually at the top. These things can move around on different hypnograms.

Right here, there's stage REM sleep. And then sometimes, not always, but often there's a position indicator on the hypnogram just showing what position they might be in. In this case, this blue line means supine sleep. Under that panel is another panel that's usually used for sleep staging.

And this is often on a time scale of 30 seconds because by convention, sleep epics are 30 seconds long. And this panel is used generally to specifically score what stage of sleep each sleep epic is in. Under that panel is usually a slightly longer time scale.

In this case it's two minutes. And this panel is used for limb movement and respiratory scoring. And you can see if this is a 30 second epic up here, down here these pink lines separate each 30 second epic. So there's one, two, three, four epics here. 30 seconds times 4 is 120 seconds, that's 2 minutes.

So this window down here, this panel, is 2 minutes long. That's not always the case. Sometimes it could be shorter, a minute long, sometimes it could be up to 10 minutes longer. Limb movements and respiratory events like apneas and hypopneas are sort of harder to see on a 30 second time scale, so we tend to use a longer time scale down here.

This whole screen is much easier to digest once you realize that you'd Generally only look at one of these at a time, one of these panels at a time, and it depends on what your question is. If you're asking like how was the whole night of sleep, what kind of sleep did they get, did they get a lot of REM, did they get a lot of stage 3, that's when you look up at the hypnogram. If you want to know what specific stage of sleep they're in, you look at the sleep staging panel, and then if you're interested in respiratory events and limb movements, you tend to look at this panel at the bottom. Now like I said, this is in Minerva, this program is called Minerva. There's also Compumetics, but there's lots of different ways to read sleep studies, and they can all look a little bit different just because they change the colors or the order of the channels might be different.

And there's a lot of customization available, too. So depending on the preferences of whoever's reading the sleep study, they may move channels around, and you just have to pay attention to what channels you're looking at in order to not get too confused. Let's look specifically at sleep staging, and that's this panel here. Sleep staging requires three things.

It requires a look at brain waves, and we do that with EEG. A look at eye movements, which is electrooculography, E-O-G, but just the eye channels is often what I'll call them. And then also a sense of muscle tone, and that's done through electromyography, E-M-G, of the chin.

So the way I have this set up right now is off in a view you'll see with the EEG channels on top, and then the eye channels. and then the muscle tone in the chin channel here. But these can be moved around.

Like I said, there's a lot of customization available, and some people prefer to have the eye channels on top. So you just have to pay attention to what's over here. to avoid getting confused about what you're looking at.

Let's talk about EEG. In the 1960s, Rechstofen and Kales, two researchers, noticed that if you looked at brainwaves during sleep, there were some discrete patterns. And so they called these sleep stages, depending on what patterns they were seeing in the EEG.

And conceptually, it's largely the same today. There have been some changes. We've standardized the way we do EEGs a little bit differently, and the sleep staging has changed a little bit. But for the most part since the 1960s it's been the same in that we're just looking for patterns in brain waves that correlate with what we understand about sleep. The standard way to take an EEG is with the 10-20 system and the way this works is you take the head from the nasion right at the front of the head at the forehead and the inion at the back of the head and you divide each half of the head into a hundred percent so this half of the head is 100%, this half of the head is 100%, and then at 10%, then 20, 20, 20, and another 10% positions, so you've divided this half of the head into these sections, you put your leads.

Okay, so that's why it's called the 10-20 system, and there's nothing magical about this except that it's a convention that allows standardization of EEG recordings across multiple studies. The nomenclature is... It's pretty straightforward.

You have F for frontal, C for central, and O for occipital. So it's front, center, back of the head. And then M is for mastoids.

They used to be called ears, but now they're mastoids because you put them over the mastoid processes. By convention, even numbers refer to the right side of the head, and odd numbers refer to the left side of the head. In order to actually see a wave that corresponds with an electrical vector like a polarity or a electrical current, you have to reference these unipolar EEG leads to another lead.

So in this case, we take the frontal lead on the right side, that's F4 here, reference it to the mastoid lead on the left side so that we're looking at current along this vector. Similarly with the central lead, it's also referenced to the mastoid lead. And then the occipital lead is also referenced to the mastoid on the left side of the head here. So if you're looking at these three channels like this, you're really looking at the right side of the head with these even numbers from front, center, to back.

And what are we actually looking for? Well it turns out it's not too complicated. If you just remember diet tab, you can remember all the different wavelengths that practitioners look for.

because we're not looking for anything as complicated as an electrophysiologist might look for. There's really just five sort of wave frequencies we look for. The slowest are delta waves, and they're at a frequency of 0 to 4 hertz, or they occur 0 to 4 times per second, same thing. Each one of these boxes sort of delineated by these faint green lines is a second in this view. So if you count the number of waves in each one of these boxes, you get the frequency.

and there's like pretty much half of one of these waves in each one of these boxes here. So this is going at a frequency of about 0.5 hertz or half a wave per second. The next fastest or the second slowest depending on how you look at it are theta waves and these are at a frequency of 4 to 8 hertz. They're sometimes described as sawtooth waves because they kind of look like a sawtooth. Just faster than theta waves are alpha waves and these are at 8 to 13 hertz.

If you were to count the number of spikes number of little waves in this one second box here you get something between 8 and 13 Hertz and then faster than alpha waves are beta waves and these are pretty much anything over 13 Hertz. Beta waves are often I think the toughest to see is discrete waves because this is what our brain looks like when we're awake with our eyes open okay so we have these really fast waves and a lot of just noise that all these little neural vectors, these electrical vectors, sort of cancel each other out. So you've got almost like a much flatter line than the other waves that, if you look carefully, has very high frequencies in here. So I got my mnemonic a little bit out of order the way I've put it on the slide here, right?

So it's D-T-A-B. But why is this backward? Well, you see both alpha waves and beta waves in wakefulness. Okay, so these are both wavelengths, wave frequencies that you will see. when a patient is awake.

The difference is eye closure. When you close your eyes, and this is eye closure right here, we'll talk about the eyelids in a sec, but just take my word for it, this is the patient closing their eyes. That's what these wavelengths are right here.

When you close your eyes, you stop processing visual information in the occiput, in the occipital cortex on the back of the head. And when you stop processing visual information, there's less neural activity, meaning there are less vectors to cancel each other out. And so you start to see the emergence of this background pattern, these alpha waves.

So instead of having just a lot of noise, you can see these little waves pop out. And about 90% of people do this where when they close their eyes they produce this alpha rhythm at the back of their heads from that phenomenon. And so this is sometimes called a posterior dominant rhythm.

So this whole period here is wakefulness, but right here the eyes are open and when the eyes close you see these alpha waves appear. Now the S stands for sleep spindles, also called sigma waves, which are kind of these... little phenomenon, these discrete little waves that happen over just a short period. They just kind of appear and go away just as quickly. And they're at 11 to 16 Hertz and they characteristically look like a spindle, like string wrapped around a rod or a peg, in that they taper from small to get bigger and then they taper down again.

So you can kind of see that shape here. This is a sleep spindle. So this is pretty much everything.

There's one wavelength that doesn't really fit into the diet tabs mnemonic and that's the k-complex which is kind of a special case of a delta wave it's like a delta wave that's isolated that sort of stands by itself and it's actually from 0.5 to 2 hertz so it's just kind of in a narrower range than typical delta waves and by convention negative deflection is up so this is a characteristic k-complex in that It has this initial negative deflection, then a positive deflection, then returns back to baseline. So it has this sort of canonical shape, the normal shape of a K-complex. And if they don't look like this, they could be something else. So don't get tripped up.

This is what a K-complex is really supposed to look like. Let's move on to the I-leads or the electrooculography, the EOG leads. These are electrodes that are placed outside the canthus.

of either side of the eyes. And by convention again, left is odd, I1, E1, and right is even, E2, I2. The way this is set up is pretty ingenious because eyes are like batteries.

They have dipoles. The cornea of your eye is positive while the retina has a relatively negative charge. And that means that if your cornea moves towards an electrode, it's going to cause a positive deflection which is down.

If it moves away from an electrode, it's going to cause a negative deflection which is up. By putting the right eye lead above the eye and on the outside, and the left eye lead below the eye and on the outside, it means that any eye movement is going to cause these waveforms to go in opposite directions. So I can tell that these are true eye movements because the waves are basically mirror images of each other. Here's an example, all right? There are these big waves in the EEG leads.

How do I know if these come from the brain or from the eyes? Because the EEG leads are close enough to the eyes that they can detect eye movements. Well, I just look down in the eye channels and I see that since these waveforms are going away from each other, they're like mirror images of each other, I know that these are eye movements and the eye movements are being caught in the EEG leads. To give a counter example, here are some big EEG waves again. And if we look down below here, here the eye channels are moving together.

They're no longer mirror images. So I know this can't be an eye movement because with a setup that looks like this, eye movements will not move together like this. They'll move apart from each other.

So I know that this is truly a EEG wave. Let's look at some actual sleep staging. I'm going to go through just a few example epics here.

You saw this one previously. This is wake and this is very fast. noisy rhythm that you see when somebody's awake and their eyes are open.

When the eyes close, and you can see this big wavelength in the EEG here, and not really sure what that is, but if you look down you can see that the eye channels are moving apart. They're mirroring each other. So this is actually eye closure here.

And I know that in part because after this happens, I see alpha rhythm. And alpha rhythm is what happens when you're awake and you close your eyes. You get this posterior dominant rhythm that you see best on the back of the head in the occipital leads. So when somebody's awake and their eyes are open, often their eyes are going to be moving around a lot and you'll have a very noisy signal. Whereas if their eyes close you'll see this alpha rhythm and most of wake during a sleep study is people with their eyes closed and you see this this alpha rhythm.

In the transition from wakefulness to sleep you get a kind of slowing down, a further slowing down. So this alpha rhythm has kind of decreased in frequency and it's harder to pull out waves that are this fast starting right about here. So here maybe there's some theta in there a little bit but there's slower waves. They're not quite as fast as the alpha rhythm here and this is called low amplitude mixed frequency.

And this is a sign of drowsiness or even of sleep onset if this continues into the next epoch. Additionally, on this channel you can see the eyes are moving here, but they're moving very slowly. There's kind of curved waves here as opposed to the more rapid things you saw here. This is blinking, this is eye closure, those are pretty quick.

Whereas here, this is slow rolling eye movements, or SREMs. And that's a sign of drowsiness and even of stage N1 sleep. So this is stage N1, and again you see those slow rolling eye movements, and you also see low amplitude mixed frequency. And drowsiness and stage N1 sleep are sort of the same thing in a lot of ways, and that's why stage N1 is considered a transitional state. This is the stage you enter first from wakefulness as you start trying to reach those lower stages of sleep like N2 and N3.

This is stage N2 and remember we talked about sleep spindles. This is a sleep spindle right here. That's characteristic of stage N2.

So if you see sleep spindles you're very likely in stage N2. And then the other thing here and this is not quite as pretty a K-complex but I think this is still a K-complex. You can see there's an initial negative deflection and then there's a positive deflection that follows. So this is a K-complex. and spindles and k-complexes characterize stage N2 sleep.

So first you see the low amplitude mixed frequency of stage N1, and as soon as you start seeing sleep spindles and k-complexes, you know you're in stage N2. Stage N3 is also called slow wave sleep or even delta sleep or delta wave sleep, and that's because it's characterized by delta waves. Instead of seeing the discrete single delta wave of a k-complex that we see in stage N2, we start seeing delta waves happening over and over again.

They come together in clusters and might even cover the entire epic. But as long as they take up six seconds or 20% of the epic, it's considered a stage N3. Now scoring REM is complicated because it can be difficult to differentiate it from stage N1 because it tends to have kind of this slow, low amplitude mixed frequency signal.

It can sometimes have alpha in it, but it This looks a lot like N1 to me. The only way I can tell that it's not N1, all right, is that there's this big eye movement right here. This is a rapid eye movement. There was a quick saccade with both eyes moving to one side of the head. If I look, this is the left lead.

So this is actually looking to the right because the cornea is moving away from the left eye. So this is a big saccade to the right, and this is a rapid eye movement. And then there's some less obvious, more subtle.

Rapid eye movements in here too One right here. So those rapid eye movements plus this kind of low amplitude mixed frequency And there's even some theta waves in here, which which are also characteristic of REM This is suggestive of REM sleep The only other thing I've got to look for is to make sure the chin tone is low because in REM sleep You're supposed to have your muscles essentially with with very low tone almost paralyzed that can be tough to tell sometimes just by looking at a the chin tone alone like i don't know is this really low tone i'm seeing something here some wiggling so sometimes i'll look back at an epic when the patient was awake so this is like a previous epic you can see this is epic 192 and this is epic 11 right here and there's much more activity in the chin tone here a lot more alpha you can't have alpha and rem but this is a lot of alpha this is overall a faster frequency than you're seeing in rem and there's theta waves here So this is a REM epic with low chin tone. This is a wakefulness epic with higher chin tone.

And in both cases, you can have these rapid eye movements. Now, the last thing we look for in the EEG leads are arousals. And arousals are evidence of going from either one sleep stage to a less deep sleep stage transiently or even going completely to wakefulness.

And that's signaled by you have what looks like stable sleep. and I think this is either rem or n1, I believe. This could be a k-complex even, so you could argue this is n2, but I think this is probably n1 or rem, because I think these are theta waves here. But anyway, you see...

Suddenly there's this huge change in the EEG leads. There's a lot higher frequency here, and there was a big movement that caused a big deflection in the EEG lead. So the patient probably moved their head here, turned it to one side or another, and at the same time there's a big increase in muscle tone. So this is an arousal.

Sometimes they're this obvious. A lot of times they're more subtle. You don't see as much actual movement that causes all the EEG leads to deflect like this. Sometimes you just see... increase in frequency.

In this arousal here you can see this is probably low amplitude mixed frequency of either I think this is probably N1 because you can see maybe there's some slow rolling eye movements here but suddenly the frequency increases and you get some alpha in here and even some increase in the chin tone so this is a another arousal and there's even some movement of the patient's head probably because you get a big deflection in the EUG leads. And lastly here, this is an example of what looks like stage and three sleep. So these are delta waves.

This is slow wave sleep. And then suddenly there's a big increase in chin tone again and an increase in the frequency, also with a movement of the EEG leads. All you need is an increase in the frequency and the amplitude to say it's an arousal. You don't always get this movement in the EEG leads, but certainly if you see something like this, this should make you be thinking there's probably an arousal, an increase in the frequency.

the waves here. So that's sleep staging where you looked at your waveforms to score what specific stage of sleep you are for each 30-second epic as well as any evidence of arousals. Let's move on to talk about limb and respiratory scoring and in order to describe the several channels you'll see here I'll give a picture of myself wearing some sleep equipment and blow up the channels right here.

So this is the nasal pressure channel that comes from a nasal cannula. I'm wearing right here And by convention, inspiration or breathing in is up. Right here is pretty much what a normal breath looks like.

These two right here are pretty normal. They're nice rounded on top. So starting here, this is a breath in, all right? And then down is exhalation.

So down is out. This isn't a perfect drawing of exhalation, but it's pretty good. This is what exhalation should look like.

So inspiration, exhalation. You can see that coming in and out of the nose with the nasal cannula. Underneath the nasal cannula I'm wearing an oronasal thermistor, or sometimes we just say thermistor, and that is a detector that measures changes in temperature. So as you breathe in, air gets cooler, and when you breathe out, the air gets warmer. And in contrast to the nasal cannula, which can only see breathing in and out through the nose, the oronasal thermistor can see breathing in and out through the nose or through the mouth.

So that's why it's called an oronasal thermistor, because it detects both nose breathing and mouth breathing. Why do you need both? Well, one is to detect mouth breathing, right?

Because if you don't see any flow on the nasal channel, you have to make sure they're not just breathing in and out through their mouth by looking at a thermistor. But the other is, the thermistor channel is essentially qualitative. All it can tell you is if a breath is present or not. It can't really tell you how big a breath is. So you need the nasal pressure channel to see how big breaths are, whereas you'd use the thermistor channel just to say if breaths are there or not there.

So one way to put it is that to score apneas, you look at the thermistor and you're looking for just nothing. Whereas to score hypopneas, you look at the nasal pressure and you're looking for small breaths. Breaths that are smaller than the normal size breaths. As you can see down here, I'm also wearing a band around my chest. This is for respiratory inductance plethysmography.

That's just a fancy way of saying that this band measures changes in the volume of my chest as I breathe in and out. Same thing on my abdomen here. These are just what are called effort channels, and these waves up and down show evidence of inspiratory effort. So you can tell that I'm attempting to make breaths, and since I have flow on the thermistor and the nasal pressure, that I'm actually getting breaths in and out. So here's a patient who's having pretty severe obstructive sleep apnea.

An apnea is defined as basically an absence of any breathing for a period of at least 10 seconds. And you can see here these pink lines again are dividing each epic. And then the fine green lines here in this wider view, they're 10 seconds each.

So you can see that this absence of flow right here is much more than 10 seconds. It's about 20 seconds, maybe even more. And you can see there's just no flow in the nasal channel. And then there's very little change in temperature in the thermistor channel.

Now whether it's an obstructive apnea, or a central apnea depends on what you see in the effort channels. So that's the thorax and the abdomen here, the size change in the chest and the size change in the abdomen. The thoracic channel for this particular patient doesn't look like it's working. It's not.

in the right place on the patient's chest probably, but at least the abdominal channel is moving up and down with the breaths here, and you can see that it's not moving as much here, but there's still some movement that looks like attempted breaths. So that's evidence of effort, and that implies that up here is an obstructive apnea. If these events have already been scored or tagged by a sleep technologist, a sleep scorer, you'll see the events displayed kind of like this, usually with boxes drawn around them and often they'll be already classified as either obstructive or central.

So these are obstructive apneas. Apneas don't require any consequence to be called apneas. That is if you just don't have breaths for 10 seconds or more, if there's just no flow and no change in the thermistor channel, that's an apnea.

Nothing else needs to happen. Hypopneas on the other hand are a little more subtle, a little more complicated. Because they require a consequence.

They require that something happens or is attributable to the hypopnea. I'm only going to talk about the American Academy of Sleep Medicine recommended criteria for scoring hypopnea, but there are other ways to score hypopneas. But the AASM definition of a hypopnea is a discrete reduction in airflow to 70% or less of baseline airflow that lasts at least 10 seconds. So...

You know, we're seeing big breaths here, big breaths here, pretty big breaths over here. And then right here in this area, there's a kind of discrete reduction in flow. So I would say that there's a hypopnea right here, or at least for these four breaths right here, for sure.

And this probably is at least 10 seconds. You can see, yeah, here to here is supposed to be 10 seconds. So this is a reduction in flow that lasts at least 10 seconds. But that's not enough.

Like I said, there has to be a consequence. And the consequence can be either a 3% or greater desaturation. And that doesn't appear to be present here. So we have sort of a maximum of 97%.

And get down to a minimum of maybe 95% somewhere in here. Yeah, I guess there's a 95 right here. The other thing that can happen, though, is you can have an arousal that is attributable to this hypopnea. So in order to see if there's an arousal, we have to go up back to the sleep scoring panel.

And you can see right here that... that the waves are kind of small and a little slow. They're not as fast as they are over here. So I think what happened here is you were probably in maybe stage N1 here, maybe N2, and then there appears to be an arousal. And that arousal is associated with this reduction in flow.

So I would call this a hypopnea associated with arousal, and that makes it a true hypopnea. Now lastly, limb movements. You can't see it in this picture, but on either leg I have a piece.

EMG electrode on my anterior tibialis on both sides. And those EMG leads are to detect limb movements, like periodic limb movements. So these are the leg channels.

You can see that the left leg is here, the right leg is right here, and you can see that there are spikes, increases in the voltage here that are indicative of movements of both legs in most of these cases. Normally I'd also have a video where I could actually see the patient sleeping, and I can see if there's any evidence of... the limb actually moving on the video. Sometimes you won't see it and there's still muscle tone there it's just not enough to actually cause a movement of the leg but sometimes it's helpful to actually see a limb movement to make sure that there isn't some kind of artifact making this signal. I don't want to go through all the technical details of scoring periodic limb movements because there's a lot of subtlety sort of funny rules to them.

This is just something you should be aware of and look for and you'll learn the more technical aspects of them as you do it more. So thanks for watching. I hope that's helpful and hopefully this is enough to get you started on reading some of your first sleep studies.

Transcript for:Introduction to Polysomnography Basics

Transcript for:
Introduction to Polysomnography Basics