Black Shadow: A Multi-Channel Sleep Study Device

In my experiments so far I've only looked at a few "channels" of data such as airflow, blood oxygen levels and sleep stage, but when you have a full sleep study in a hospital or sleep-lab many more channels are monitored, all of which provide insights into the cause of your sleep problem.

I've recently been using a hospital-grade multi-channel sleep study device that is new to the market. It's designed to be used at home or in a sleep lab / hospital. I've been using it at home. It's called the Black Shadow and it's made here in the UK by Stowood Scientific Instruments.

The Black Shadow is capable of monitoring:

• Nasal Airflow (via a nasal cannula)
• Oral & Nasal Airflow (via a thermal sensor)
• Respiratory effort (ie breathing movements) (via 2 inductance belts)
• Pulse oximetry
• Pulse rate (via pulse oximetry)
• Plethysmograph (pulse profile)
• Sound recording and snore detection (from a small microphone on the cannula)
• Actigraphy body movement
• Body position: Left, right, prone, supine & upright
• ECG (one channel)
• Separate leg movements (via EMG or movement sensors)
• Event marker (via a patient activated button)

The system also has provision for auxiliary inputs (4). It also has provision for ECG/ EEG/ EOG/ EMG data (via an additional unit).

The prospect of having all the data available in an automated unit really intrigued me. In subsequent blog-posts I'll show how I've used the Black Shadow to verify some of my previous experiments and create some new ones, but this post is really going to be an introduction to some of the things that the Black Shadow has revealed about my sleep.

In many sleep labs and hospitals, you are connected to several bedside monitors by long wires allowing you to move in bed, but making it hard if you need to get up in the night to use the bathroom. The Black Shadow overcomes this by being wearable, so once you're connected, you are free to move around should you need to.

I've no intention of posting a photo of me in pyjamas wearing the device on the internet, so I fitted it to a mannequin to illustrate how it's worn.

The next few photos illustrate some of the sensors in a bit more detail (click for larger images).

Under the shirt the mannequin is wearing three adhesive electrodes which are used to record ECG data.

The microphone and thermal sensor fix to the nasal cannula, which is then worn around the ears:

Heel / Ankle straps are also worn to detect leg movements in the night:

Also an oximeter sensor is worn on the finger which connects via a long lead to the top of the central unit. It's a flexible probe and actually a lot more comfortable than the plastic "crocodile-type" probes that I'm used to. My mannequin's hands are a bit like mittens, so I haven't fitted the oximeter probe to him.

I decided to start from scratch and take a baseline recording of my sleep (no vitamins, no supplements, no Rematee and no alcohol). After connecting myself to the various sensors I pressed the button on the front of the unit, waited for the recording light to flash and then went off to sleep.

Configuration

The Black Shadow is Bluetooth capable, which allows you to configure the device wirelessly, and to see live data via bluetooth at the beginning of the night in order to check that all the channels are configured correctly. So after pairing with my laptop, and checking that everything was fine, I was ready to sleep.

Analysing the Data

The data is recorded to a high speed SD Card.

In the morning I removed the SD Card from the Black Shadow and downloaded the data to the Visi-Download software.

Not only was there a lot more data than I'd had access to before, but I was able to manipulate it and interrogate it in ways that I've never been able to do previously.

Once downloaded, I was able to see a graph of all the channels on one page along with some additional channels derived from the data: "Pulse Transit Time" (an indicator of intrathoracic pressure and an indicator of autonomic arousals in sleep), "R-R interval" and "Flatness".

The channels are able to be moved up and down the screen so that you can, for example, put the SPO2 (oxygenation) on top of the airflow channel, making it easy to spot correlations.

The software will perform an analysis on the data (using customisable criteria) and display it in a multi-page report. Once the data has been analysed, markers appear on the graphs showing events such as apnoeas, hypopneas, pulse rate changes, oxygen desaturations, snores etc.

Zooming in on an event allows you to verify it, and if necessary disregard it.

Positional Data

As expected, it was easy to spot that I did indeed have respiratory events through the night. Once analysed, my AHI was calculated to be 7.24 which agrees nicely with my previous 5-day mean score of 7

Comparing selected channels with the body-position channel, it was also easy to see that the vast majority of my respiratory events occurred while I was sleeping on my back. (Click for larger image).

The body position channel (3rd down) shows the first half of the screen with me on my back turning onto my right side for the second half.

Comparing the supine to side-sleeping data it is clear that (From the top down):

• My SPO2 (oxygen saturations) are higher and stable with side sleeping
• My pulse is lower and stable
• Change in body position from Supine to Right
• My snoring stops (filtered sound channel)
• My oral/nasal airflow is stable
• My leg isn't twitching
• My body movement is greatly reduced

The improvement is revealed in a table in the report:

It's possible to exclude periods of data from the analysis by highlighting them; so for example, by only looking at the periods when I slept in a supine position (on my back) I could see how my AHI was if I only slept on my back.

All of this is a great indicator that positional therapy such as the Rematee would help my sleep problem.

Obstructive vs Central Apnoeas

The two respiratory effort channels allowed me to see whether my apnoeas were obstructive or central in nature. Obstructive apnoeas are where the body still tries to breathe (ie the intercostal and diaphragm muscles still move as normal) but the airway is obstructed, rendering the effort useless. Central apnoeas are where the brain doesn't signal the muscles to breathe, or the signal is blocked for some reason.

Each respiratory effort graph relates to one of the inductance belts that are worn around the chest and stomach. The graph therefore shows the chest and stomach movements, and hence if an effort to breath was made how strong it was in relation to other breaths.

My graph showed that I still attempted to breathe and yet no airflow was recorded, this makes my apnoeas obstructive in nature. (Click for larger version)

Audio Recording & Snore Detection

The Black Shadow records sounds continuously through the night, so unlike my attempts with a voice-activated dictaphone, you are able to hear sounds leading up to an event rather than just a second or so after the event itself.

Because the microphone is calibrated, the sound channel is graphed according to sound amplitude in dB. The software also creates a second sound-channel of sound that is filtered to highlight snoring. The detected snores are then marked automatically, making it easy to click on the graph and actually hear the snore. Listening to sound can be done at any point of the night, not just snores, so it's possible to hear sleep-talking and other noises.

I'd like to say that I sleep silently, and that the Black Shadow didn't detect any snoring from me, but that would of course be a lie because I snored a lot. The report (above) showed that I snored at a level of above 55dB for 90% of the night! 158 snores were louder than 75dB, which is roughly the noise level generated by a lawnmower!

This is a screenshot of 23 seconds of my sleep. Four snores are shown which were detected and marked automatically, these correspond to the snores that have been isolated on the filtered channel. What I find interesting about this screenshot is that the vibrations caused by my snoring are visible on the airflow channel (top - orange line).

(Click for a larger image)

Again, I was able to see that most of my snoring took place while I slept on my back.

Leg Movements

When I took a look at the leg movement data I saw something that surprised me.

Like most people, as I drift off to sleep, I twitch a bit. If I'm laying on my back I know that my left leg is prone to having the odd twitch (since I injured a disc at L5 in my spine), and this also happens when I'm sitting using my computer late at night, but I didn't suspect it happened when I was asleep. Looking back, I probably should have done.

Sometimes arousals from respiratory events will cause a body or leg movement, but I was seeing these twitches in periods of sleep that were free of respiratory events.

This is what I saw when I looked at a 12 minute section of my data (click for larger image):

At 1:15 am my left leg (leg 1) started to twitch slightly, and as you can see the amplitude increased almost with each subsequent movement over the next 4 minutes. This then caused me to turn from my back onto my left side (shown by the Body Position channel in blue). This happened several times a night, in all sleeping positions and on more than one night. Although I haven't been able to eliminate them, maybe this is an indication of why the Vitamins and Minerals improved my sleep as Iron is thought to be beneficial to Periodic Limb Movements.

Those legs movements and the subsequent turning over could be a problem if it happened often enough because it is yet another thing that can cause a sleeper to awaken briefly. These small awakenings are called micro-awakenings (or micro-arousals). The higher the number of these there are in a given night, the more your sleep is disrupted and the greater the likelihood of waking in the morning still feeling tired. If I want to improve my sleep further then these are the sort of things that I need to be aware of.

Pulse Rate Increases

The Visi-Download software also shows other arousals which are calculated from my heart-rate data.

This section of my sleep shows how a cluster of respiratory events (marked by the blocks under the airflow line) affected my oxygen saturations (top red line). If you look at the other data shown in line with those events you'll see that not only was my heart rate affected, but also my Pulse Transit Time - a very good indicator that I was continuously being woken (brief)y by my respiratory disturbances, causing fragmented sleep.

The Pulse Plethysmograph also shows changes, even on this timescale because the arousals went on for a prolonged period of time. Shorter micro-arousals are able to be seen using the PTT and Pleth channels when looking at a smaller timeframe.

Event Marker

Pressing this button during recording causes a marker to be inserted into the data. This could be for any event, such as being woken, waking from a nightmare, feeling unable to breathe, or even (as I plan to use it) for marking lucid dreams and sleep paralysis. 

The Black Shadow opens up many additional ways for me to explore and hopefully improve my sleep and I'm looking forward to experimenting further with it.

Additional Links

Non-invasive Monitoring of Vital Signs Utilising Pulse Wave Transit Time

Use of Pulse Transit Time as a Measure of Autonomic Arousals in Patients with Obstructive Sleep Apnea

Pulse Transit Time Improves Detection of Sleep Respiratory Events and Microarousals in Children - Pepin et al

Obstructive Apneic Events Induce Alpha-receptor Mediated Digital Vasoconstriction - Zou et al

Lucid Dreaming - Wake & Back to Bed

A few nights ago I had a spontaneous lucid dream.

The curious thing about this is that I can pin-point when it happened on my hypnogram, so I thought I'd share it on here.

First let's take another look at a near-perfect hypnogram. This is from my son's sleep and is about as near to textbook as I think possible in real-life:

Adults will typically get less REM than this, but the pattern should still be the same. Slow-wave (deep) sleep in the first half of the night, decreasing as the night wears on then increasing amounts of REM into the morning.

In my previous post about Lucid Dreams I mentioned a method of inducing lucid dreams called "Wake and Back to Bed". Essentially it goes like this: 

1. Go to bed at your usual time
2. Set an alarm for the early hours (say 4am)
3. Wake when the alarm sounds
4. Get out of bed and read for an hour or so
5. Return to bed

This has the effect of causing your REM sleep to be concentrated in the period when you've returned to bed (the idea being that your body is now screaming out for the REM that it should have had when you were awake for the hour). The period of being awake should also "sharpen" your conscious mind so that it is more likely to be aware of the dreamworld when it next encounters it.

For my lucid dream, this happened by accident as is good evidence that the method works.

I spontaneously woke at my usual weekday alarm-time which was ridiculously early for a Saturday morning, so took off the Zeo headband, flicked through the news and Twitter on my phone (still in bed), then decided to have another go at sleeping.

I didn't get a mass of REM in this second sleep, and most of the "sleep" that I did get was light dozing, but there was an isolated period where I slipped into REM and had a lucid dream. This can be clearly seen on my Zeo hypnogram for the night:

I know that it's boring listening to other people's dreams but to give you an idea of how lucidity evolves I'll outline it...

I was in a restaurant in France. I was sitting with my extended family who took up most of the restaurant. Then I realised that people were "tutting" at me. Somehow I'd managed to offend the restaurant owners. 

It seemed odd. After a while the tutting turned to people talking about me as if I wasn't there. I looked down at my plate and realised that the tables were full of rain. Again, more oddness. Water was actually flowing around our plates as if we were eating in some sort of water feature.

It began to make less and less sense, then I looked at the faces of the people who I thought were family. I didn't recognise any of them. 

Then it hit me: people's faces changing, changed laws of physics, feeling of conspiracy... it was a dream. I still wasn't sure, so I calmly did a "reality check" (Something that I don't get to do in dreams that often)... I counted the fingers on one hand. Five... so it must be real. I looked at my hand again and saw an extra finger! 

Ha! Definitely a dream then. 

I got up to leave and (almost symbolically of the dream trying to re-capture me) found that the people in the restaurant were trying to prevent me from going. I managed to get out of the door which seemed to pull me back like a magnet, but eventually I was in the street outside. 

More symbolism: The world outside was definitely the dreamworld, it was a town mixed with cartoon imagery complete with giant cartoon people using skyscrapers as stepping stones. 

Anyway, I won't go on, but as lucidity came on slowly I was able to stay calm and stay in the dream, rather than waking. Eventually after a couple of periods of forgetting I was dreaming and becoming lucid again, the dream won and settled into a normal dream. The dream covered about 3 days, but as you can see from the solitary green bar on the hypnogram, it lasted not much more than 5 minutes.

This is definitely a method that I think is worth exploring further.

Sleeping Position: Get off your back!

When you have a multi-channel sleep study, one of the "channels" that is examined is the position that you sleep in.

At first this may seem a strange thing to monitor, after all if you manage to achieve sleep, who cares how your inert body is positioned? 

Sleep position affects several aspects of our sleep, for example when I sleep on my back I find that:

• I experience more sleep paralysis
• I experience more body twitches as sleep begins

• It's also known that breathing is harder when sleeping flat on your back (supine), it is for this reason that patients with higher level spinal injuries are recommended to sleep with their head inclined and foot of their bed lowered as this can actually increase their vital capacity
• GERD / Acid Reflux / Heartburn is affected by sleeping position. Raising the head of your bed can prevent or at least reduce reflux episodes.

However, in this blog-post I'm going to concentrate on another major effect that supine sleeping has on the sleeper...

Supine sleeping can dramatically increase the number of obstructive apnoeas and hypopneas a sleeper has through the night!

The reason for this is that gravity has a greater effect on the muscles and soft tissues that relax in sleep and cause obstructive apnoeas. If you can lie on your side then gravity doesn't affect them in the same way, and can go some way to keeping the airways open during sleep.

Invariably we move around in our sleep so I needed a way to monitor my sleeping position as my monitor doesn't have this ability.

More sophisticated sleep study devices do this automatically, and there is actually an iPhone app that can do this for you too: Somnopose, after a couple of night of experimenting with the app you should be able to get a fairly accurate reading. You will probably be surprised how many times you actually turn over in your sleep.

This was actually quite easy for me to do manually as I already had my infra-red camera watching me for several nights during sleep. I then analysed the video to determine my sleep position. As well as recording continuously, the camera creates a new file every time it detects movement, so all I had to do was flick through the files and record the position on a chart alongside my oxygen desaturations. I did also plot a chart of body position against detected apnoeas, but the chart below shows things clearly enough.

 s=supine (back), p=prone (front), l=left down, r=right down

It was no surprise to see that the majority of my apnoeas and hypopneas occurred when I was sleeping on my back.

On several nights I deliberately began sleep on my side, but always woke to find myself flat on my back again and with a high AHI recorded. I needed a way to prevent myself from turning over in the night.

Years ago, a solution to positional obstructive sleep apnoea was to "sew a tennis ball into the back of your pyjamas", this had a "Princess and the Pea" type effect on the sleeper and caused them to be uncomfortable if they turned over and slept on their back.

Nowadays there are more sophisticated solutions, and the one that I put to the test is called the "REMATEE Bumper Belt". It's a neoprene belt that velcro fastens around your chest. At the back of the belt are pockets into which you place inflatable "bumpers". Sounds strange?

It may be clearer if I show you a photo of what I'm describing. Rather then subjecting you to a screen-capture from my camera, I'll use a product photo instead (click for a larger version).

It's surprising how quickly you get used to it.

My belt also has shoulder straps to stop it turning round in the night although when it first arrived I was slightly daunted by the idea of being strapped in, so I decided to tuck the shoulder straps into the pouches that held the bumpers.

I found that if the belt was tight enough then I didn't need to use the shoulder straps, but it does take a few nights of trial and error to get used to how tight to wear it and how much air to put in the bumpers.

When the belt first arrived I tried it out for a few nights without monitoring my breathing and found that it did affect my Zeo stats. It increased my REM sleep. It also reduced the number of awakenings that I had through the night.

Initially, it seemed to decrease Deep (SWS) Sleep but this stopped and my SWS actually increased after a few nights of using it. (See graphs below).

Then I began my five nights of recording my breathing...

Again I followed my rules for determining my 5-night mean data with the obvious exception of not beginning the night by sleeping on my back. I also didn't take any of the supplements discussed in this blog, and I stopped the vitamins and minerals on the Friday beforehand to try to eliminate the effects of those on my sleep.

The results were clear:

The REMATEE belt kept me off my back and reduced my oxygen desaturations:

 s=supine (back), p=prone (front), l=left down, r=right down

My AHI each night was dramatically reduced:

This brought my 5-day mean AHI down to 1.70.

This makes the REMATEE the most effective method that I've found so far to reduce my AHI. 

The reduction in respiratory events (reflected by the reduced AHI) also had a positive knock-on effect on my Zeo data...

Considering that most of my respiratory events happen in REM sleep, it makes sense that when the number of events was reduced, my REM increased (because it was not being broken by me waking up to breathe). As already discussed, my Slow Wave Sleep (deep) also increased.

Will I carry on using the REMATEE? Definitely.

As a follow up, I plan to combine the REMATEE belt with my 5-HTP routine as both seem to reduce my AHI through different mechanisms. Maybe their combined effect will be greater than as individual methods.

Additional Links

The REMATEE Bumper Belt is available in the UK from Intus Healthcare: 
http://www.eu-pap.co.uk/rematee-neoprene-bumper-belt.html

REMATEE http://www.antisnoreshirt.com/