Thread: REM Detectors (again)

I've tried to follow this thread, but I seem to be really missing the overall idea. I want to make a Kvasar or some type of mask that can function similar to the NovaDreamer's REM detection.

Can someone point me to the posts in here that will get me to that point? I have a fairly decent understanding of basic electronics, so I just need some help figuring where to go next. Thanks.

Originally Posted by TheKing

I've tried to follow this thread, but I seem to be really missing the overall idea. I want to make a Kvasar or some type of mask that can function similar to the NovaDreamer's REM detection.

Can someone point me to the posts in here that will get me to that point? I have a fairly decent understanding of basic electronics, so I just need some help figuring where to go next. Thanks.

This link may get you started: Mathieu Mallet // Morpheus REM Monitor

Thanks! I'll research that.

Originally Posted by TheKing

Thanks! I'll research that.

Here is another one: www.bbelmont.com/pdfs/DreamMachine.pdf

On that one, is that also a mask? I went through the documents, and I just saw a lot of schematics for building the board, so I am guessing this just does the REM, but it's called the Dream Machine.

I did a quick scan before printing it, so I probably missed something.

Originally Posted by TheKing

On that one, is that also a mask? I went through the documents, and I just saw a lot of schematics for building the board, so I am guessing this just does the REM, but it's called the Dream Machine.

I did a quick scan before printing it, so I probably missed something.

Think it is some kind of mask or at least some type of signaling for LD cue: "After detecting that the user is in REM sleep, the machine will either wake up the user gradually (alarm mode) or give subtle hints to induce lucid dreams (blink LEDs only 5 times)."

Perfect! My speed reading isn't up to snuff...

I'll give it a go though. Thanks for the links.

Hi!
I've been thinking about making a DIY REM sleep detector and I found this topic. The topic is huge and I only got to speed read a few post, so if i'm giving old information and need to be updated, please tell me.

Based on my research, you could detect REM sleep with heart beat alone, but I just had the stinkiest time trying to figure out how to find the algorithm on Google. For what I found, it use as an input the Heart rate variability (Heart rate variability - Wikipedia, the free encyclopedia). Peripheral arterial tone can be used too and its detected by a wristwatch-like medical device used by paramedics.

But my most promising solution comes by searching about a kind of gadget called a bio-alarm clock. This stuff wakes you up at a light stage of sleep so you feel fresh and restful, as opposed of being groggy from waking up from a deep sleep stage. They do detect REM sleep and other sleep stages as part of their algorithm and sometimes they generate a nice graph of your sleep stages afterward.

Some of this stuff is real expensive and they are not made to be used for LDing. But I discovered an bio-alarm clock in the form of a $0.99 IPhone app (Sleep Cycle alarm clock) that does all the stuff only with the IPhone accelerometer! You just have to put the IPhone on your mattress and, as it is claimed, detect the movement you make in bed. As the site says: "Since you move differently in bed during the different phases, Sleep Cycle uses the accelerometer in your iPhone to monitor your movement to determine which sleep phase you are in." The app even gives you the graph of your sleep stages. When I saw that my jaw dropped and I said: wow... darn impressive.

This make me think that:
1-One could make a cheap NovaDreamer-like IPhone app for LDing
2-For people like me who don't own a IPhone, the only sensor I would need for a DIY solution is an accelerometer. (Like a Wiimote connected on a PC via bluetooth placed on my mattress, for example)

Now I just need to find an algorithm, since the damn IPhone app isn't open-source.

Ive learnt about eye tracking in college. There are devices + technology out there that can track your eye movements, ive seen a video where a disabled person is controlling a computer by moving there eyes, so in theory if you could make the devices more precise and look for smaller eye movements. And make a program where it tracks your eyemovements over a certain movement (over 10 moves in 30 seconds?) then activate something which informs you that you are dreaming i think its possible guys

Originally Posted by philquiet

Hi!
Now I just need to find an algorithm, since the damn IPhone app isn't open-source.

That might be something I'll have to look into. I'm sure there are a couple of articles in the good ol' peer reviewed journals lying around on the topic.

You could also use a small PIC (like the 16F88) to sample readings from an accelerometer, and send the serial data back to the PC using a recycled 433MHz transmitter from a discarded wireless thermometer (4800bps)...

I've been reviewing some of those IR sensing goggles for REM detection. It seems simple enough. I have several IR LEDs and phototransistors lying around.

If it's as simple as shining IR on the eyelid and measuring the frequency of voltage change at the photo-transistor, then this is so simple it's not even funny! The Mathieu Mallet version shows a very simple configuration. The photo transistor could modulate the VCO section of a 4046 PLL or simple a-stable 555 circuit running at <=10kHz, and send the output to the MIC in of a laptop. A simple frequency count can be done in software, and simple calculations can show presence of REM.

This simple setup would work for experimental purposes to get all the bugs worked out. It would also work great for those who don't mind sleeping with a laptop running next to their bed. Use a simple pair of wires to carry You could have a plethora of signals to alert REM from a PC, such as subtle audio files to play, rather than a harsh buzzer that's more likely to awaken than cue you into lucidity.

After figuring out the best algorithm for detecting, then it could be ported to a PIC or AVR for a portable version.

I'll whip up an IR detector in some goggles then see what kind of results I get. This should prove quite interesting!