Andreas, I have been watching your videos for years now (first video of yours I watched was somewhere around #50). I must tell you, tremendous gratitude over here for you putting this together. I would not have connected the dots on this without you covering it. I was able to help family in New Orleans that are infected monitor their readings because of this video! You have provided a wonderful public service by producing this video. Thank you!
I’m a retired nurse with some 40 years of experience in clinical, administrative and teaching roles. You explanations were as good as any I’ve heard. You are certainly to be commended for your diligent research and clear presentation. Thanks very much Andreas.
Thank you for your nice words. They are very valuable for me. This was not a "normal" topic for me and I was not completely sure if everything was right...
Thenk you for this video. As engineer and asthmatic patient I have one cheap device and I thought if it was good or not (for 20€). The one they use in hospital costs about 350€! I appreciate your work.
I also thought these are toys. This is one of the reasons for this video. When I started to dig into the matter I saw that the method is simple and can be done with standard LEDs and chips. And the accuracy seems to be quite good.
I wish this video had come out a few months ago. I am in the hospital for a Pulmonary Thromboendarterectomy. Like you I found a lot of deficiencies in many of these devices. Not to mention proprietary data formats and bad software. I found all the features you mentioned in the MTS OX-610A. It is also much more motion tolerant to use during the day. Thanks for the video.
In the Late 70's I worked for Hoffman La Roche as a Biomedical Engineer. We had a line of P02 Monitors Based on the Clark Electrode. The Clark Electrode was applied to the patient's Skin via adhesive and then connected via a wire to the monitor. The Clark Electrode needed to be set up with a Electrolyte and Membrane before use. We are very lucky to have optical P02 monitors now
@@AndreasSpiessThis technology is still used for babies as their fingers are too small. Normally you apply the sensor on the head but you cannot monitor it continuously and a lot of time you burn the skin (above the Fontanella) . Another technology for babies uses laser diodes to get a continuous signal to avoid burning.
hi andreas, I have been sleeping with the CPAP for about 15 years and I must say that my life has changed for the better almost instantly. I was always sleepy although I slept normally and often woke up with a headache, then I did a polysomnography and it resulted in more than 200 apneas, one of which for a minute! Since then, about once a year I go to a hospital for a series of check-ups, the last and most invasive is a blood gas analysis that consists in taking arterial blood (the arteries are deeper and more difficult to reach) just to see oxygen saturation. During the outpatient checks for some time they have used oximeters similar to those you have shown and just some months ago, thanks to the banggood advertisements I also bought one that I use every now and then as needed: sometimes, during a slightly restless sleep , the cpap mask moves and the system efficiency is reduced. fortunately the level of my syndrome is not high and therefore I am not in danger of life, but age always advances .... P.S. now I also know how it works! tnx ciao
i think there have been some experiments with doing this with *Blood Glucose* as well, to make a non-invasive continuous blood glucose monitor. I don't think any commercial devices exist yet, as the existing continuous blood glucose monitors that i know of all still use a very small electrode that must be injected into the skin. edit: Seems there is some interest in "near-infrared spectroscopy" with it, there's a 2013 paper which mentions the *1550nm wavelength* as a target for glucose reading. (Paper is free to read: _Non-invasive blood glucose monitoring using near-infrared spectroscopy, by M. Ahmad, A. M. Kamboh, and A. Khan, EDN, 2013.)_ Their prototype also reads blood oxygen levels, just as a happy sideeffect
I can imagine this is a huge market. We also have companies here in Switzerland which manufacture those (conventional) glucose testers. Good business AFAIK...
Very interesting video, and I was pleased that you mentioned OSCAR. I picked up thhe RD117 board with its MAX30102 chip and a couple of M0 and M4 boards, but the signal analysis was beyond me. I may get back to it when I am happier with the state of OSCAR software.
Thank you for your software! I think your time is better invested in the software. A lot of people like me do not want to run to the doctor all the time. Maybe you also make a list of oximeters which are supported or which ones you prefer. I hope you will ahve now a few users more...
Brilliant I love the extent you go too... you definitely get a +1 from here... you are very dedicated not many would have tried to sleep with nail polish to see if that tumble the reading.
That's very useful information about oximeters in a electronics perspective. My model is a rebranded Aeon A310L no bluetooth, uses 7-seg digit displays which are very easy to see. Always clean the device after usage for better results.
Interesting video as usual. I've learned something new about CPAP and sensors. Your videos sure do stimulate me to keep on learning and experimenting. Isn't electronics an wonderfull and awesome hobby ?!
Great video Andreas, thank you! However please be aware that this method doesn't allow you to draw any conclusion on a potential infection (eg with the virus), it is just a parameter that gives you an indication about the amount of oxyhemoglobin in your blood. Skin, bones, light etc are impacting the reading and therefor the absolute number you read on the display is never precise. You can read a SO2 of 99% or 92% and be healthy, but in case the value over time drops you should further investigate the reason for it. So the sensor is used to detect abnormalities over time and software on intensive care units offer dashboards that detect any significant alterations. Obviously a significant drop of SO2 above 10% would be such an indicator.
TH-cam changed their policy, mentioning Chinese Wuhan Virus is completely ok and videos will be monetized. We will not forget Alphabet/Google/TH-cam censorship.
Excellent explanation, as usual! The Red/IR was also visible in the photo of the "LED" 7:44 --> There are two dice and two bond-wires placed back-to-back. Another clue as to the red/IR on the scope 10:14 is the forward voltage of each. The IR has lower energy per photon, and hence a lower forward-drop (~1.4V) whereas the red has ~2.0V.
That is what I also thought when I saw the two different voltages. But I did not lookup the foreward voltages of hte different diodes. I think these devices are not crappy at all and it would be very hard to do it better.
If you want the numbers to change, it does not take long if you hold your breath and do physical working like jumping up and down. Something like 30 seconds should drop oxygen levels quite a lot. That is much easier than holding your breath for 2 or 3 minutes at rest.
here's some general advice for the author of this great great video :) a) drink less (if you drink more than a glass) b) do more sports (if you rarely do) c) start practicing a wind instrument with a lot of air resistance (why not, interesting physics and probably you can order some from AliExpress) d) if you're not feeling well, go to the doctor, do not save money by not taking care of your health :)
@@AndreasSpiess Thanks. I'm getting a CPAP machine in the next few days and will investigate it's capabilities. Back in the 70s, my Uncle's BSc thesis was on the use of inductive sensing for SpO2 - using oxygen's paramagentism.
Interesting video.Thanks. I think it was nellcor who patented the method of putting a resistor in the sensor assembly. The value was dependent upon the calibration they did at the factory. So all the readout machine has to do is read that value at switch on and so be able to calculate much more accurate results. Nellcor sell many sensors :)
Thank you for all the effort you put into these videos Andreas, I found your channel by accident but I thoroughly enjoyed the clips I watched, you just got a new subscriber. I even got a few good laughs thanks to your Swiss engineer sense of humour :) Quick question for you, please: do you know of any oximeter that can be read continuously? USB would be ok, Bluetooth would be great. Thanks a lot again!
Very useful for pilots of non-pressurized general aviation aircraft who fly above 10kft or so. They must keep an eye on their oxygen levels. Good oxygen : can fly higher and avail of better winds and economy, not so go good : need to descend (if terrrain allows). Can add a warning light somewhere in the cockpit or maybe a display. Can also log over time to help understand your personal tolerance/stamina and plan future flight profiles accordingly. Sweet.
Good idea! If I remember right: There is also a law about how long you are alowed to fly above 10000 feet. My engines always gave up before my loung ;-)
By the way, there are now cheap boards with AD8232 chips for ECG amplification. Normally it's a mess to set up differential amplifiers, what with dual power supplies etc. But these boards do all the work, require only 3.3V and output a very arduino friendly signal between 0 and 3.3V. From what I can tell (I'm a Veterinarian) the signal should be good enough for most ECG based diagnostics, even if not certified. It seems to have a low-pass filter around 1000 hz. Not sure how useful the absolute amplitude is, there is also some high-pass filtering going on. When I say "ECG based diagnostics" I mean that the amplifier seems good enough, but you'd have to arrange enough of them in the right way, of course.
There are quite a few videos around with this sensor. So I decided not to cover it. I would not know what to add. Most oximeter videos were less technical.
Relly nice Andreas! I'm asking my self how do you find the right target for your videos and how you have your devices ready on time... Very instructive this time.
You can also get a remote probe for CMS50D+. It plugs into the mini USB port. Makes it a bit more comfortable to wear for longer and it provides a much better attachment. Got one from Contec's eBay shop for about £20.
You must be spying on me. I was just looking at oxygen sensing devices last week! Thanks very much for the thorough summary, especially for pointing out stuff that isn't worth bothering with. I would NOT recommend DIY instead of a proper medical consultation if you think you have sleep apnea. The first "symptom" of sleep apnea is often a car accident.
My last month went watching these devices. it was painful. but these devices are extremely useful tools for doctors. its good that they are now in getting cheap for not so wealthy people but at the same time, i think, its more about what one uses the data for, rather than collecting data (like my organisation). When i see my pulse rising, i get nervous and it rises even more due to positive feedback, so i remove all these!!
You are right with the heart rate. I am not so sure if you can influence SpO2 as easy. At least the application I showed seems to be quite common. At least in my country and at a certain age.
The Arduino sketch may not work with the max 30102. But a more powerful arduino type device may work using the sketch. A reflexive rather than translucent hemoglobin device is useful very useful in determining blood perfusion in less than translucentParts of the body.
About the bluetooth interface... perhaps you could simply use it to connect to a Pi 4, and poll data continuously? There _has_ to be some way to do that. Also, if this was to be used over a long time, is there no way to put a cutoff switch for the Oled display, and possibly run an external power source to the oximeter?
Maybe you are right. I think, spendig 10 dollars more and save on batteries is also a deal... I would not overcomplicate things with external batteries, but of course, it can be done.
@@AndreasSpiess my cpap had a SD card too but i cannot read data. the pc tell that SD card is empty, but maintenance people can read it! can you tell us why, maybe in a video? Hoping all the cpaps brands uses the same code tnx in advantage
I do not think there is a standard. So you have to google. Maybe you find information about your particular device. I selected mine because it is compatible with this OSCAR software. Maybe you place an issue there and ask for support? For this channel, this is too special.
@@AndreasSpiess Yes, each manufacturer probably uses a different standard, or personalizes the data so that only those in charge of the work: maintenance technicians or doctors, can read the data. Ultimately there may be some fool who wants to tamper their data for some sort of convenience. p.s. my CPAP is a Respironics: a medical brand of Philips Hi
I’d like to see a more in-depth look at decoding the data. I tried to make one once but trying to process the data to only read IR or red light was difficult. Afterwards I realised that having a background readout between the IR and red light pulses was important for self-calibration purposes, and that there were too many degrees of freedom otherwise. But I think using IR and red LEDs as photodiodes for seperate sensing would allow for better differential measurements in order to cut down on one of those degrees of freedom, I’d be able to have both LEDs on at once. Maybe with a third colour sensor I’d be able to ditch the switching signal altogether. The problem is that you get 100Hz or 120Hz noise from overhead lighting, so filtering that out becomes vital. I used a 2nd order sallen-key filter but I think a notch filter would have worked better, especially for strobing the red and IR LEDs at a similar frequency like 200Hz.
As I mentioned (And you experienced): It is not easy to extract those values. But this topic is very special and would not be interesting for most of my viewrs.
I got a good one($40) a couple of months ago, and I guess I'm too healthy because it doesn't go below 95% even when training. I got a cheap ebay blood pressure tester and it goes from medium BP to low after 20 mins of weight training. I go to free UNI lectures and training to fatigue is the only thing that partially helps with dementia.
The graph at 13:50 actually shows SpO2 levels rising during the time breathing stops. I assume this is because Oximeter doesn't keep track of time and software calculates time from the starting time that you enter manually and sample number, so this SpO2 plot is actually off by about 40 seconds.
an oximeter is VERY expensive up here in scandinavia, about 50euro, i got myself one anyway, but i dont undestand the difference between the two different waveforms shown at the bottom line, Levi
@@AndreasSpiess i have two different "waveform" lines BELOW the digital numbers, i now know that one is the STRENGHT of the heartbeat, but still dont know what the other "waveform" shows, maybe we have different machines
Nice review. Just after I watched your video yesterday I ordered a CONTEC CMS50D+ from AliExpress. Simply because I like gadgets and number-crunching in MS Excel ;) But after some googling i wonder if I shoud have ordered the CONTEC CMS50H instead. Just because of the additional Perfusion Index (PI) data. I have not figured out if this data is exported to the computer though, as I have not found the manual online for the CMS50H. Any thoughts on this?
The OSCAR software mentioned here accepts data from the CMS50 D+, F, and I models for sure. I don't think anybody has tried an H model, but believe the software should handle the Perfusion.Index. It is a work in progress.
Thank you for your excellent video. I learned a lot. Q: Given what you said about inaccuracy of mobile modules (like the MAX30102) , how do watches like the FitBit Sense achieve such accuracy? Have you considered doing a teardown?
Did you test if it provides accurate measurements? I did not find a comparison between the Fitbit and an accurate device. My Polar watch also (sometimes) measure heartrate quite accurately even if the glossy leaflet says it works.
...i was just thinking about this lastnight...along with body temp.. unfortunately i dont have the sensors on hand... thanks Andreas, some timely info!
Now let's get something like this working on the bike. I'm quite curious what the readings would look like during a solo ride versus group ride versus criterium race. I don't race any more but was always amazed by the computer metrics after a crit. I usually had 10-15% higher average heartrate/speed/cadence than anything I could do by myself. I wonder what the blood oxygen would look like.
Great technical review !!!!! You've just convinced not to waste money on those smartbands and pay a bit more on a real finger oximeter !!!!! Thnks !!!! +1
Thank you for this vid. I'm interested in measuring this but I always thought such medical sensors would cost an arm and a leg, or even more. Do you think it would be possible to receive the data live by BLE, all through the night, using an ESP32-based project? Such project could pass the data to MQTT/NodeRed/Influx/Grafana and it could administer a small electric shock at an APNEU event, or, start beeping when the sensor falls off from your finger during sleep. Another way to detect breathing would be to put a loop of flexible resistive wire around your chest and measure the resistance of that continuously. That would be a good alternative for people that would pull off the finger sensor in their sleep. You should buy some AAA-sized 3.2V LiFePO4 cells for your oximeter. The ones I bought come with a short circuit dummy, for use in 3 Volt devices that hold two batteries.
yes this is very possible. You need to be careful with power requirements for discharge you would want to deliver. Just esp32 ble with i2c Max chip is very very simple but can be very noisy as well.
Andreas, at 9:40 you show that the polarity of the diode gets reversed to alternative between iR and Red light from the LED. Does this mean that they're using a module with two LEDs in it; one forward biased and the other, reverse biased?
I bought the CMS50DA+ from Walmart. I use it with Eneloop NI-MH batteries, it's good for one night, probably not two, on each charge. When the batteries are low it can give nonsensical results rather than shutting down. I tried reading the data via USB with SleepyHead software with no luck. It does work with the supplied software running on Windows XP (i have no plans to ever get a newer windows machine).
Question. That red led in MAX30102 is part where it should contact with our skin? If i need to make casing box out of it.. will the function still work properly if that LED closed by kind of transparent plastic?
The 2nd fastest way to detect bad sleep (snoring, stopped breathing) is to use an audio recorder like audacity, set the sample rate to 8000 and a single channel (mono) to not waste storage space, and let it record the whole night. In the morning, look at the waveform where it was loud. If there is no snoring or sudden gasping for air, there is probably no sleep apnea problem. ;) (The fastest way is getting told so by your significant other)
You are right! I was in military service in large rooms an I know what you are talking about. But there is a second indication which is quiet: When the brain stops to create signals.
The brain is quite robust. If there is something wrong with breathing, the CO2 levels rise and then it initiates some emergency functions. Rising heart rate, increased breathing reflex and partial wakeup. The amount of signals created by the brain will increase to stop the dangerous situation. The biggest problem with this is, it is a lot of stress for the body instead of regenerative sleep. Other health problems, like cardiovascular diseases might follow and can really cause the brain to stop creating signals... A CPAP machine can help with sleep apnea, so can a change in sleeping position or weight-loss. My Grandmother didn't want to hear anything about sleep apnea or snoring. The moment when the partial wake-up with forceful inhalation happened was extremely loud. My bedroom was on another floor and I still woke up from it. I wish I had known about CPAP devices when she was still relatively healthy.
How do you see before buying it if it is a reflective model or not? We can find some oxymeter on Aliexpress below 5€, without BLE which is not so useful, what do you think about that? Thanks
The Max3102 is pretty accurate, you need the correct algorithm, a proper encloser. It's unrealistic to expect the raw values to be reflective and that too without an enclosure with stray light entering the sensor.
@@AndreasSpiess I guess I will have to make a video :-) the Max30102 is made by Maxim integrated, this chip is used in 1000s of professional medical devices. The issues you are seeing are software related.
Hi Andreas, just a quick heads up , is it possible to monitor multiple bluetooth oximeters with a single ESP32 ? And another really silly query, is it possible to reduce the SP02 for a few second. I tried holding my breath till I could but could not see much change :) , just to check if the oximeter is really working
6:40 To measure our dog’s SPO2 the vet clipped the sensor to his tongue, poor guy couldn’t close his mouth but didn’t mind at the time. He had nearly “dry drowned” from drinking too much water whilst swimming and was unconscious. He recovered really well.
I liked this video. I do not understand how the processor measures absorption. Does the detector output an analog voltage. I say this because when the red light is on the detector output dips slightly not a 1 or 0. I will have to read up more on this. Thank you.
I showed the sensor which measures the remeining light emitted by the LEDs on the other side of the finger. It uses an Analog to Digital converter for that reason.
About a month ago I bough one of those things from the German company Beurer for about 50EUR. I have most of the time about 90%-92% oxygen level. That is far away from optimal but not as bad as it sounds. I also visited a doctor and he confirmed those values. I write this here so people don't panic if they see low values. My girlfriend has constantly 99% measured with the same device.
Please be aware that the impact of skin color, segmentation of the skin etc etc are impacting the measurement. Basically in med applications the trend is of essence, so if you have a drop in saturation compared to your 'standard' value there is an issue, otherwise you should not become concerned!
@@juergwyser6520 In my case the values from my measurement and the values from the hospital match. I have "white" skin. My girlfriend has brown skin and her measurement is constantly 99%.
Another very interesting video from the old guy with the Swiss accent who is not a doctor, but plays one on TV, or at least on TH-cam. I'm surprised that they don't use the blood oxygen meter as an additional test for detecting people with corona virus. Maybe it could detect people with 'shortness of breath'? Final question: Do you really have to pay in Switzerland to see a doctor? Of course we pay quite a lot here in the US, but we are always told that people in really nice countries like Switzerland have free access to unlimited healthcare. Of course we also have to buy our own guns here and only the wealthy can afford their own fallout shelters...
Great video as always! I purchased the CMS50D+ and cannot get OSCAR to show anything. I can connect directly via Oximetry wizard->start->choose CMSF37 (strange, CMS50D asks for upload file option in menu which is not there) Then communication starts (also visible in the log) Then nothing happens! Also by importing the . SpO2 file from SpO2 assistant nothing happens. I probably missed something. Any suggestions?
I would log an issue with Oscar. And I use the application which comes with the oximeter, store the results, and upload the file to Oscar. Much more reliable.
Yes. But it had its problems (see my issues on the Oscar repo). This is why I use the app of the oximeter. One minute more work, but so far it always worked.
Wow this video came at just the right moment. Over the last year I have had a few heart and lung problems and this covid-19 pandemic puts me in the "at risk" group.I struggle to reach 94% oxygen during the day and drop below 50%, according to the hospital monitor, at night = sleep apnoea. The hospital monitor was great but I was going to see if I could get an arduino to monitor my oxygen levels and BPM ( it's what we nerds do during our last days). But now I shall be buying something a bit more professional off banggood or aliexpress. Hopefully it will arrive within a reasonable timeframe
Sorry - bus was late so I'm a fair way to the back today! It seems that the meters are OK in the UK but we only have 5000 ventilators in total - hence the governments panic this week to try and get some built (despite having a long warning). Tempted to try this on my feet - I have a circulation problem so any rough guide would help me monitor the effect of the tablets rather than a 'gut feel'. Thanks for the video - maybe a project for the lock down...
I think I stated my opinion. If you follow the link of hte project in the video description you see that this guy used FFT to extract the pulse. You can definitively do more with a better CPU.
should be nice to review body thermometers from aliexpress also. Big price differences and forehead thermometers are really popular now. I have now a forehead contact thermoemeter from Lidl and it works really perfect. But you need to read clearly the manual how to use and know the difference between contactless and contacted forehead thermometers. ;) But with Corona now all the thermometers are sold out and the price is very high.
The coronavirus came here. So I guess I sit in the first row of the balcony just in case. Anyway I like taking medical machines apart. To be honest, CPAP machines are very interestingly engineerd. I repaired quite a few of those
I wondered what sensors they use to get such precise airflow and pressure values... I am sitting in the basement. But for me, this is not a big change ;-)
@@AndreasSpiess Some machines have a custom built sensor array. Others use a special version of a off the shelf sensor. It goes between the pump and the output. But the machine only measures airflow and pressure. The rest of the parameters are calculated. It is also common to see the machine monitor the pump itself.
Hello Sir , Will you please suggest the sensor name to use on wrist because i want to use reflectance method to measure the oxygen level and want to implement on wrist . because now days to many smart watches are coming with this feature and I want to test it and if you have idea on reflectace method and if you know the sensor name which can be used on wrist or forehead please suggest .
@@AndreasSpiess Yes sir i know that you have mentioned the sensor name and i am already aware of that sensor max30102 and but my point is that can i use it on wrist and will it give accurate reading . because to use on wrist i have to use reflectance oxymeter method .
Great video! , i was wondering if i can read values of a normal oximeter using an arduino or esp32? , i do not wanna use a sensor but a real oximeter that i can buy , with a modification or idk! would i be able to do it?
Did you try a restrictive bandage to reduce fresh blood flow to your fingers? Depending on your cardio-vascular systerm, holding your arms up above your head for a couple of minutes ought to reduce the flow. There's not much hydrostatic head reserve in the human body.
having (a lot of) COPD, i use an Innovo INV-4301/PE . . . it seems stable, has long battery life and usually gives the same readings as my doctor's office machine. just tested it putting my finger in sideways . . . it worked . . . might be a work-around for nail polish enthusiasts?
Andreas, thanks for another excellent video, I was wondering whether this sensor could be associated with a LoRa device in order to update patients data to a LoRa gateway, I am asking it because in Emilia Romagna (Italy) there are gateways already in place in several City Halls, including my town, these collect all kind of data, from private and public sources, the access is free and the entire network is covered by a local public company which only reserves to mamage and store the big data. Given the critical time, this infrastructure could serve the purpose of gathering medical data from patients within the Region. Thanks Claudio
I assume you could connect it to a lora node. But the question is more: What problem do you want to solve which cannot be solved by a Smartphone. A smartphone is simple to use and already in all households.
@@AndreasSpiess purpose may be of any type... but my section of question that is Which Sensor do U recommend to trace Sulphur Dioxide is in electronics section... which one is low-cost or how can I detect... and my use is in Exhaust Gas Analyzer that is used by Mechanical Engineers to find the usefulness of Engine Exhaust Gasses to use them or not decision like in Power Plants like in economizer, turbo engines etc to use exhaust gases etc.... my and your concern is only from sensor... remaining concern is by the user who want to analyze them..
Hi Andreas, I'm using the "Sleep As Android" app to track my sleep for multiple years now. While it's not opensource, the data it records is not closed down at all. I always saw the "Pulse Oximeter" option in the Wearable section of the app, but I thought that it only works with expensive devices. Looks like it expects to connect to any random Bluetooth device, and can record the data from them. If you have energy for some further investigation, you can check out the app and see if either the "CMS50D+" or the others can supply data to the app. I would be very interested to know, as I would buy one of them if I could add another data point to my sleep tracking (which works either with the phone's accelerometers, with a "sonar like" motion tracking, or with wearables like the Mi band, Garmin, etc.) Whether you look into this or not, thanks for the informative video about the subject! Here's the program's documentation page about the subject, if you want to check it out: docs.sleep.urbandroid.org/devices/oximeter.html
Thank you for the information. I was more interested in the technology of he sensors. So I think I will not cover the topic in the future. This is more a "maker" channel...
@@AndreasSpiess Sure, it just seemed you might be also interested in sleep tracking, so I wanted to point out this app in case you want to gather more data easily with your existing SpO2 sensors.
You said it is hard to manipulate oxygen levels. I was testing myself a few days ago and got 94-93%, just like other family members. "It is crap" that's what I told to myself. But then we went to the shop for a little apocalypse hoarding and when I got back it was 97%. Maybe just going to fresh air can increase it by a few percent?
"the Arduino doesn't have enough SRAM to store 100 samples of IR and red data in 32-bit format" This only shows that most Arduino-libraries are very poor at resource management. 200 values in 32-bit format only need 800 bytes and the ATmega328 has 2 kByte of SRAM...
I did not investigate furter When I saw the values jumping I started to doubt that you get reliable measurements even with more "power". But I might be wrong, of course.
@@AndreasSpiess I would just assume a very badly written library... It is for sure possible to implement the complete detection reliably in an ATmega328 including the control of the two LEDs. Yes we are talking data analysis here, but we are looking at very slow signals, data rates of 200 samples per second and in the case of the Uno about 16 MIPS
Andreas, I have been watching your videos for years now (first video of yours I watched was somewhere around #50). I must tell you, tremendous gratitude over here for you putting this together. I would not have connected the dots on this without you covering it. I was able to help family in New Orleans that are infected monitor their readings because of this video! You have provided a wonderful public service by producing this video. Thank you!
You are welcome. I hope, everything is ok with your family!
I’m a retired nurse with some 40 years of experience in clinical, administrative and teaching roles. You explanations were as good as any I’ve heard. You are certainly to be commended for your diligent research and clear presentation. Thanks very much Andreas.
Thank you for your nice words. They are very valuable for me. This was not a "normal" topic for me and I was not completely sure if everything was right...
Thenk you for this video. As engineer and asthmatic patient I have one cheap device and I thought if it was good or not (for 20€). The one they use in hospital costs about 350€! I appreciate your work.
I also thought these are toys. This is one of the reasons for this video. When I started to dig into the matter I saw that the method is simple and can be done with standard LEDs and chips. And the accuracy seems to be quite good.
I wish this video had come out a few months ago. I am in the hospital for a Pulmonary Thromboendarterectomy. Like you I found a lot of deficiencies in many of these devices. Not to mention proprietary data formats and bad software. I found all the features you mentioned in the MTS OX-610A. It is also much more motion tolerant to use during the day. Thanks for the video.
Your seem to be a more professional device. But also a little more expensive...
In the Late 70's I worked for Hoffman La Roche as a Biomedical Engineer. We had a line of P02 Monitors Based on the Clark Electrode. The Clark Electrode was applied to the patient's Skin via adhesive and then connected via a wire to the monitor. The Clark Electrode needed to be set up with a Electrolyte and Membrane before use. We are very lucky to have optical P02 monitors now
I was also fascinated about this simple but effective method.
@@AndreasSpiessThis technology is still used for babies as their fingers are too small. Normally you apply the sensor on the head but you cannot monitor it continuously and a lot of time you burn the skin (above the Fontanella) . Another technology for babies uses laser diodes to get a continuous signal to avoid burning.
@@juergwyser6520 Great info, Thanks !
hi andreas,
I have been sleeping with the CPAP for about 15 years and I must say that my life has changed for the better almost instantly.
I was always sleepy although I slept normally and often woke up with a headache, then I did a polysomnography and it resulted in more than 200 apneas, one of which for a minute!
Since then, about once a year I go to a hospital for a series of check-ups, the last and most invasive is a blood gas analysis that consists in taking arterial blood (the arteries are deeper and more difficult to reach) just to see oxygen saturation.
During the outpatient checks for some time they have used oximeters similar to those you have shown and just some months ago, thanks to the banggood advertisements I also bought one that I use every now and then as needed: sometimes, during a slightly restless sleep , the cpap mask moves and the system efficiency is reduced. fortunately the level of my syndrome is not high and therefore I am not in danger of life, but age always advances ....
P.S. now I also know how it works! tnx
ciao
It seems that a lot of such devices are used everywhere. This is also why I used this example. I have a colleagua who reports similar things like you.
i think there have been some experiments with doing this with *Blood Glucose* as well, to make a non-invasive continuous blood glucose monitor. I don't think any commercial devices exist yet, as the existing continuous blood glucose monitors that i know of all still use a very small electrode that must be injected into the skin.
edit: Seems there is some interest in "near-infrared spectroscopy" with it, there's a 2013 paper which mentions the *1550nm wavelength* as a target for glucose reading. (Paper is free to read: _Non-invasive blood glucose monitoring using near-infrared spectroscopy, by M. Ahmad, A. M. Kamboh, and A. Khan, EDN, 2013.)_ Their prototype also reads blood oxygen levels, just as a happy sideeffect
I can imagine this is a huge market. We also have companies here in Switzerland which manufacture those (conventional) glucose testers. Good business AFAIK...
Very interesting video, and I was pleased that you mentioned OSCAR. I picked up thhe RD117 board with its MAX30102 chip and a couple of M0 and M4 boards, but the signal analysis was beyond me. I may get back to it when I am happier with the state of OSCAR software.
Thank you for your software! I think your time is better invested in the software. A lot of people like me do not want to run to the doctor all the time. Maybe you also make a list of oximeters which are supported or which ones you prefer.
I hope you will ahve now a few users more...
Brilliant I love the extent you go too... you definitely get a +1 from here... you are very dedicated not many would have tried to sleep with nail polish to see if that tumble the reading.
Not only doing it, but confessing it in public ;-)
That's very useful information about oximeters in a electronics perspective. My model is a rebranded Aeon A310L no bluetooth, uses 7-seg digit displays which are very easy to see. Always clean the device after usage for better results.
Interesting video as usual. I've learned something new about CPAP and sensors. Your videos sure do stimulate me to keep on learning and experimenting. Isn't electronics an wonderfull and awesome hobby ?!
And for the moment much saver than drinking in bars ;-)
Great video Andreas, thank you! However please be aware that this method doesn't allow you to draw any conclusion on a potential infection (eg with the virus), it is just a parameter that gives you an indication about the amount of oxyhemoglobin in your blood. Skin, bones, light etc are impacting the reading and therefor the absolute number you read on the display is never precise. You can read a SO2 of 99% or 92% and be healthy, but in case the value over time drops you should further investigate the reason for it. So the sensor is used to detect abnormalities over time and software on intensive care units offer dashboards that detect any significant alterations. Obviously a significant drop of SO2 above 10% would be such an indicator.
It had a reason I did not "praise" it for the virus application. The CPAP application is well known and AFAIK accepted.
Oh no, you mentioned the human malware in the video now it gets demonetized.
:-))
I wonder if they can tell the difference between Corona Virus and Corona Beer?
TH-cam changed their policy, mentioning Chinese Wuhan Virus is completely ok and videos will be monetized. We will not forget Alphabet/Google/TH-cam censorship.
Excellent explanation, as usual! The Red/IR was also visible in the photo of the "LED" 7:44 --> There are two dice and two bond-wires placed back-to-back. Another clue as to the red/IR on the scope 10:14 is the forward voltage of each. The IR has lower energy per photon, and hence a lower forward-drop (~1.4V) whereas the red has ~2.0V.
A project-idea? --> Create an open-source oximeter better than the crappy and unstable cost-down-rubbish with a (say) Teensy of STM32 M4 or ESP32
That is what I also thought when I saw the two different voltages. But I did not lookup the foreward voltages of hte different diodes. I think these devices are not crappy at all and it would be very hard to do it better.
If you want the numbers to change, it does not take long if you hold your breath and do physical working like jumping up and down. Something like 30 seconds should drop oxygen levels quite a lot.
That is much easier than holding your breath for 2 or 3 minutes at rest.
Good information. I was just sitting...
great elaboration, thanks. i have one of the max modules too, and yet to test one out.
I am curious what results you get...
here's some general advice for the author of this great great video :)
a) drink less (if you drink more than a glass)
b) do more sports (if you rarely do)
c) start practicing a wind instrument with a lot of air resistance (why not, interesting physics and probably you can order some from AliExpress)
d) if you're not feeling well, go to the doctor, do not save money by not taking care of your health :)
Thank you for your advice. Last year I made a 3000km bicycle trip along the danuber river to the black sea ;-)
@@AndreasSpiess keep it up! Somehow I have a hard time imagining that wasn't with a home constructed ebike!
Indeed on the interesting physics. My textbook related the uncertainty principle to how quickly you can change notes on a tuba.
Thanks Andreas, very informative and topical for me. I have recently been diagnosed with severe apnea and was researching this myself.
I hope you get it under control. And maybe you can use the software to learn...
@@AndreasSpiess Thanks. I'm getting a CPAP machine in the next few days and will investigate it's capabilities.
Back in the 70s, my Uncle's BSc thesis was on the use of inductive sensing for SpO2 - using oxygen's paramagentism.
Interesting video.Thanks. I think it was nellcor who patented the method of putting a resistor in the sensor assembly. The value was dependent upon the calibration they did at the factory. So all the readout machine has to do is read that value at switch on and so be able to calculate much more accurate results. Nellcor sell many sensors :)
I googled them. They seem to be specialized in this technology. I assume their price tags are a little higher ;-)
Thank you for all the effort you put into these videos Andreas, I found your channel by accident but I thoroughly enjoyed the clips I watched, you just got a new subscriber. I even got a few good laughs thanks to your Swiss engineer sense of humour :)
Quick question for you, please: do you know of any oximeter that can be read continuously? USB would be ok, Bluetooth would be great.
Thanks a lot again!
Welcome aboard the channel! I only have the meters shown in the video. The BLE meters read continuously.
Very useful for pilots of non-pressurized general aviation aircraft who fly above 10kft or so. They must keep an eye on their oxygen levels. Good oxygen : can fly higher and avail of better winds and economy, not so go good : need to descend (if terrrain allows). Can add a warning light somewhere in the cockpit or maybe a display. Can also log over time to help understand your personal tolerance/stamina and plan future flight profiles accordingly. Sweet.
Good idea! If I remember right: There is also a law about how long you are alowed to fly above 10000 feet. My engines always gave up before my loung ;-)
By the way, there are now cheap boards with AD8232 chips for ECG amplification. Normally it's a mess to set up differential amplifiers, what with dual power supplies etc. But these boards do all the work, require only 3.3V and output a very arduino friendly signal between 0 and 3.3V. From what I can tell (I'm a Veterinarian) the signal should be good enough for most ECG based diagnostics, even if not certified. It seems to have a low-pass filter around 1000 hz. Not sure how useful the absolute amplitude is, there is also some high-pass filtering going on. When I say "ECG based diagnostics" I mean that the amplifier seems good enough, but you'd have to arrange enough of them in the right way, of course.
There are quite a few videos around with this sensor. So I decided not to cover it. I would not know what to add. Most oximeter videos were less technical.
Relly nice Andreas! I'm asking my self how do you find the right target for your videos and how you have your devices ready on time... Very instructive this time.
This time, if I am honest, it was a little luck...
@@AndreasSpiess Hahaha rogue but sincere! I'm joking, I admire your work as maker and TH-camr
You can also get a remote probe for CMS50D+. It plugs into the mini USB port. Makes it a bit more comfortable to wear for longer and it provides a much better attachment. Got one from Contec's eBay shop for about £20.
Thank you for your information!
2:26 cracked me up. "As an engineer i could say, let's turn the logic and save some money" => "You stop breathing."
can't argue with that.
:-))
You must be spying on me. I was just looking at oxygen sensing devices last week! Thanks very much for the thorough summary, especially for pointing out stuff that isn't worth bothering with.
I would NOT recommend DIY instead of a proper medical consultation if you think you have sleep apnea. The first "symptom" of sleep apnea is often a car accident.
I had the topic prepared before Corona. But I was catched by reality...
This is a very detailed analysis.
Thanks!
My last month went watching these devices. it was painful. but these devices are extremely useful tools for doctors. its good that they are now in getting cheap for not so wealthy people but at the same time, i think, its more about what one uses the data for, rather than collecting data (like my organisation). When i see my pulse rising, i get nervous and it rises even more due to positive feedback, so i remove all these!!
You are right with the heart rate. I am not so sure if you can influence SpO2 as easy. At least the application I showed seems to be quite common. At least in my country and at a certain age.
Very interesting, informative and potentially useful video, thanks !
You are welcome!
The Arduino sketch may not work with the max 30102. But a more powerful arduino type device may work using the sketch. A reflexive rather than translucent hemoglobin device is useful very useful in determining blood perfusion in less than translucentParts of the body.
About the bluetooth interface... perhaps you could simply use it to connect to a Pi 4, and poll data continuously? There _has_ to be some way to do that. Also, if this was to be used over a long time, is there no way to put a cutoff switch for the Oled display, and possibly run an external power source to the oximeter?
Maybe you are right. I think, spendig 10 dollars more and save on batteries is also a deal... I would not overcomplicate things with external batteries, but of course, it can be done.
I would love to see a video detailing how you took apart the CPAP and read its function! Thanks for the video!
No need to take the CPAP apart. It has an SD card where it stores its information
@@AndreasSpiess my cpap had a SD card too but i cannot read data. the pc tell that SD card is empty, but maintenance people can read it!
can you tell us why, maybe in a video?
Hoping all the cpaps brands uses the same code
tnx in advantage
I do not think there is a standard. So you have to google. Maybe you find information about your particular device. I selected mine because it is compatible with this OSCAR software. Maybe you place an issue there and ask for support? For this channel, this is too special.
@@AndreasSpiess Yes, each manufacturer probably uses a different standard, or personalizes the data so that only those in charge of the work: maintenance technicians or doctors, can read the data. Ultimately there may be some fool who wants to tamper their data for some sort of convenience.
p.s. my CPAP is a Respironics: a medical brand of Philips
Hi
You are amazing, exploring in detail.
Thank you.
Excellent review, kudos for your thorough review, you Dear Sir have a wonderful mind. All the best!
Thank you!
A breath of fresh air....great work.
Thank you. Maybe we have now more time for walks if fresh air. Everything shutdown here...
I’d like to see a more in-depth look at decoding the data. I tried to make one once but trying to process the data to only read IR or red light was difficult. Afterwards I realised that having a background readout between the IR and red light pulses was important for self-calibration purposes, and that there were too many degrees of freedom otherwise. But I think using IR and red LEDs as photodiodes for seperate sensing would allow for better differential measurements in order to cut down on one of those degrees of freedom, I’d be able to have both LEDs on at once. Maybe with a third colour sensor I’d be able to ditch the switching signal altogether.
The problem is that you get 100Hz or 120Hz noise from overhead lighting, so filtering that out becomes vital. I used a 2nd order sallen-key filter but I think a notch filter would have worked better, especially for strobing the red and IR LEDs at a similar frequency like 200Hz.
As I mentioned (And you experienced): It is not easy to extract those values. But this topic is very special and would not be interesting for most of my viewrs.
Watched your video breathless.
:-))
A brilliant response to the current situation 🏴😁 thanks Andreas 👍
You are welcome! I hope you are well and not affected too much.
Very good topic - got those meters at home :) - BTW @Andreas your english gets better and better with every video :)
Thank you for your compliment!
The performance of D1 mini with 30102 is good. You may give it a try.
Thank you for the info.
I got a good one($40) a couple of months ago, and I guess I'm too healthy because it doesn't go below 95% even when training. I got a cheap ebay blood pressure tester and it goes from medium BP to low after 20 mins of weight training. I go to free UNI lectures and training to fatigue is the only thing that partially helps with dementia.
Sounds you are healthy. Very good for the comming times ;-)
The graph at 13:50 actually shows SpO2 levels rising during the time breathing stops. I assume this is because Oximeter doesn't keep track of time and software calculates time from the starting time that you enter manually and sample number, so this SpO2 plot is actually off by about 40 seconds.
You have to synchronize the two manually and it is hard to align the start on hte second. The Oximeter has no internal time.
@@AndreasSpiess The F and I models have timekeeping - unfortunately the CPAP's clock isn't as good. Mine needs to be reset at least monthly.
an oximeter is VERY expensive up here in scandinavia, about 50euro, i got myself one anyway, but i dont undestand the difference between the two different waveforms shown at the bottom line, Levi
I do not know which lines you refer.
@@AndreasSpiess i have two different "waveform" lines BELOW the digital numbers, i now know that one is the STRENGHT of the heartbeat, but still dont know what the other "waveform" shows, maybe we have different machines
@@AndreasSpiess (actually the line shows in your videoclip)
I assumed it shows the "amount" of blood or so. But I did not care too much.
Great Video. Did you tried to get data in real time from CMS50D+? in advertising of this product claim it is possible.
Nice review. Just after I watched your video yesterday I ordered a CONTEC CMS50D+ from AliExpress. Simply because I like gadgets and number-crunching in MS Excel ;) But after some googling i wonder if I shoud have ordered the CONTEC CMS50H instead. Just because of the additional Perfusion Index (PI) data. I have not figured out if this data is exported to the computer though, as I have not found the manual online for the CMS50H. Any thoughts on this?
The OSCAR software mentioned here accepts data from the CMS50 D+, F, and I models for sure. I don't think anybody has tried an H model, but believe the software should handle the Perfusion.Index. It is a work in progress.
Thank you for your excellent video. I learned a lot. Q: Given what you said about inaccuracy of mobile modules (like the MAX30102) , how do watches like the FitBit Sense achieve such accuracy? Have you considered doing a teardown?
Did you test if it provides accurate measurements? I did not find a comparison between the Fitbit and an accurate device.
My Polar watch also (sometimes) measure heartrate quite accurately even if the glossy leaflet says it works.
...i was just thinking about this lastnight...along with body temp..
unfortunately i dont have the sensors on hand...
thanks Andreas, some timely info!
I had them because I was fascinated about the simplicity of the concept. So I thought, the topic fits right now...
Now let's get something like this working on the bike. I'm quite curious what the readings would look like during a solo ride versus group ride versus criterium race. I don't race any more but was always amazed by the computer metrics after a crit. I usually had 10-15% higher average heartrate/speed/cadence than anything I could do by myself. I wonder what the blood oxygen would look like.
I will take it on one of my next rides to check it out. Just got a new bicycle.
Wow thank you so much!.. subscribed and liked.
If I want to optimize the home-made one with arduino, what would your advices be ?
I do not know.
Great technical review !!!!! You've just convinced not to waste money on those smartbands and pay a bit more on a real finger oximeter !!!!! Thnks !!!!
+1
Thank you for this vid. I'm interested in measuring this but I always thought such medical sensors would cost an arm and a leg, or even more. Do you think it would be possible to receive the data live by BLE, all through the night, using an ESP32-based project? Such project could pass the data to MQTT/NodeRed/Influx/Grafana and it could administer a small electric shock at an APNEU event, or, start beeping when the sensor falls off from your finger during sleep.
Another way to detect breathing would be to put a loop of flexible resistive wire around your chest and measure the resistance of that continuously. That would be a good alternative for people that would pull off the finger sensor in their sleep.
You should buy some AAA-sized 3.2V LiFePO4 cells for your oximeter. The ones I bought come with a short circuit dummy, for use in 3 Volt devices that hold two batteries.
Maybe somebody hacks the BLE with an ESP32. So far I have no time for such a project. Thank you for the LiFePo4 battery tip.
yes this is very possible. You need to be careful with power requirements for discharge you would want to deliver. Just esp32 ble with i2c Max chip is very very simple but can be very noisy as well.
Helpful information
Thank you!
Smarter by the minute. 17min to be exact 😁 Thank you Mr. Wizard ❤
You are welcome!
Andreas, at 9:40 you show that the polarity of the diode gets reversed to alternative between iR and Red light from the LED. Does this mean that they're using a module with two LEDs in it; one forward biased and the other, reverse biased?
I think so.
I bought the CMS50DA+ from Walmart. I use it with Eneloop NI-MH batteries, it's good for one night, probably not two, on each charge. When the batteries are low it can give nonsensical results rather than shutting down.
I tried reading the data via USB with SleepyHead software with no luck. It does work with the supplied software running on Windows XP (i have no plans to ever get a newer windows machine).
That was also my experience with IKEA batteries. SleepyHead seems to be the (abandoned) predecessor of the one I used.
@@AndreasSpiess - Thanks, I didn't know SleepyHead was resurrected. That helps.
Question. That red led in MAX30102 is part where it should contact with our skin?
If i need to make casing box out of it.. will the function still work properly if that LED closed by kind of transparent plastic?
There is one sensor which has to contact the skin. I do not know if it works with something in between
The 2nd fastest way to detect bad sleep (snoring, stopped breathing) is to use an audio recorder like audacity, set the sample rate to 8000 and a single channel (mono) to not waste storage space, and let it record the whole night. In the morning, look at the waveform where it was loud. If there is no snoring or sudden gasping for air, there is probably no sleep apnea problem. ;)
(The fastest way is getting told so by your significant other)
You are right! I was in military service in large rooms an I know what you are talking about. But there is a second indication which is quiet: When the brain stops to create signals.
The brain is quite robust. If there is something wrong with breathing, the CO2 levels rise and then it initiates some emergency functions. Rising heart rate, increased breathing reflex and partial wakeup. The amount of signals created by the brain will increase to stop the dangerous situation. The biggest problem with this is, it is a lot of stress for the body instead of regenerative sleep. Other health problems, like cardiovascular diseases might follow and can really cause the brain to stop creating signals...
A CPAP machine can help with sleep apnea, so can a change in sleeping position or weight-loss.
My Grandmother didn't want to hear anything about sleep apnea or snoring. The moment when the partial wake-up with forceful inhalation happened was extremely loud. My bedroom was on another floor and I still woke up from it. I wish I had known about CPAP devices when she was still relatively healthy.
How do you see before buying it if it is a reflective model or not?
We can find some oxymeter on Aliexpress below 5€, without BLE which is not so useful, what do you think about that?
Thanks
The ones looking like the one I used are most probable not reflective and maybe you see the emitting diode on a picture. I do not have another answer.
OUTSTANDING - thank you
You are welcome!
The Max3102 is pretty accurate, you need the correct algorithm, a proper encloser. It's unrealistic to expect the raw values to be reflective and that too without an enclosure with stray light entering the sensor.
I never saw an accurate measurement on TH-cam with this chip :-(
@@AndreasSpiess I guess I will have to make a video :-) the Max30102 is made by Maxim integrated, this chip is used in 1000s of professional medical devices. The issues you are seeing are software related.
I never saw a reflective Oximeter so far.
Hi Andreas, just a quick heads up , is it possible to monitor multiple bluetooth oximeters with a single ESP32 ? And another really silly query, is it possible to reduce the SP02 for a few second. I tried holding my breath till I could but could not see much change :) , just to check if the oximeter is really working
I assume you can connect more than one, but never tried.
And I tried, too to hold the breath. It is not easy to reduce the SPO2 if you are healthy.
@@AndreasSpiess Thanks much Andreas, looking forward to see another new video soon.
Wonderful explanation.
Thanks!
You are welcome!
6:40 To measure our dog’s SPO2 the vet clipped the sensor to his tongue, poor guy couldn’t close his mouth but didn’t mind at the time. He had nearly “dry drowned” from drinking too much water whilst swimming and was unconscious. He recovered really well.
Interesting place. Probably more dangerous if he would have been awake ;-)
Great video!
Thank you :)
Glad you liked it!
I liked this video. I do not understand how the processor measures absorption. Does the detector output an analog voltage. I say this because when the red light is on the detector output dips slightly not a 1 or 0. I will have to read up more on this. Thank you.
I showed the sensor which measures the remeining light emitted by the LEDs on the other side of the finger. It uses an Analog to Digital converter for that reason.
thanks Andreas
You are welcome!
About a month ago I bough one of those things from the German company Beurer for about 50EUR. I have most of the time about 90%-92% oxygen level. That is far away from optimal but not as bad as it sounds. I also visited a doctor and he confirmed those values. I write this here so people don't panic if they see low values. My girlfriend has constantly 99% measured with the same device.
Please be aware that the impact of skin color, segmentation of the skin etc etc are impacting the measurement. Basically in med applications the trend is of essence, so if you have a drop in saturation compared to your 'standard' value there is an issue, otherwise you should not become concerned!
@@juergwyser6520 In my case the values from my measurement and the values from the hospital match. I have "white" skin. My girlfriend has brown skin and her measurement is constantly 99%.
Well explained 😊 Thanks ✨🎉
You are welcome!
I'm sitting in the first row, I guess?
I think so ;-) You were fast.
@@AndreasSpiess Great video every week! I even do my own projects thanks to your tutorials. :D
Please respect the distancing rules of your country ;D
Interesting video. Are you doing a ECG amplifier next ;)
I do not think so. There are already quite some videos about this ECG chip around.
Another very interesting video from the old guy with the Swiss accent who is not a doctor, but plays one on TV, or at least on TH-cam.
I'm surprised that they don't use the blood oxygen meter as an additional test for detecting people with corona virus. Maybe it could detect people with 'shortness of breath'?
Final question: Do you really have to pay in Switzerland to see a doctor? Of course we pay quite a lot here in the US, but we are always told that people in really nice countries like Switzerland have free access to unlimited healthcare. Of course we also have to buy our own guns here and only the wealthy can afford their own fallout shelters...
I have to disapoint you: I am a doctor, just not a medical doctor ;-)
Love is like oxygen
.
You get too much you get too high
.
Not enough and you're gonna die.
Love gets you high.
Sweet !
hello, dont you try to use esp8266 or esp32? maybe it will be better a little bit. And I broke glass filter maybe it helps a little.
As I said: I gave up. Too difficult and a good alternative at hand...
Great video as always
Thanks for sharing :-)
You are welcome!
Great video as always! I purchased the CMS50D+ and cannot get OSCAR to show anything. I can connect directly via Oximetry wizard->start->choose CMSF37 (strange, CMS50D asks for upload file option in menu which is not there) Then communication starts (also visible in the log) Then nothing happens! Also by importing the . SpO2 file from SpO2 assistant nothing happens. I probably missed something. Any suggestions?
I would log an issue with Oscar. And I use the application which comes with the oximeter, store the results, and upload the file to Oscar. Much more reliable.
@@AndreasSpiess How do you upload? Also with the wizard?
Yes. But it had its problems (see my issues on the Oscar repo). This is why I use the app of the oximeter. One minute more work, but so far it always worked.
Very useful video! Thanks
You are welcome!
Excellent video as usual, I have a withings sleep monitor that gives me the oxygen level as well, I always wanted to know how it works. Thank you.
You are welcome!
Interessting video Andi, but want to buy a CMS50D+ but i only find a XMS50D. Is this the same? Could you send me a link? thkx
I thought I have a link in the video description? But I will check again
Wow this video came at just the right moment. Over the last year I have had a few heart and lung problems and this covid-19 pandemic puts me in the "at risk" group.I struggle to reach 94% oxygen during the day and drop below 50%, according to the hospital monitor, at night = sleep apnoea. The hospital monitor was great but I was going to see if I could get an arduino to monitor my oxygen levels and BPM ( it's what we nerds do during our last days). But now I shall be buying something a bit more professional off banggood or aliexpress. Hopefully it will arrive within a reasonable timeframe
BTW I'm English and we spit in the face of death.
Maybe somebody hacks the BLE protocol. Or maybe they use a standard service. Then it is easier.
Great 👍 video to learn. Thank you 🙏 Andre!
You are welcome!
Sorry - bus was late so I'm a fair way to the back today!
It seems that the meters are OK in the UK but we only have 5000 ventilators in total - hence the governments panic this week to try and get some built (despite having a long warning).
Tempted to try this on my feet - I have a circulation problem so any rough guide would help me monitor the effect of the tablets rather than a 'gut feel'.
Thanks for the video - maybe a project for the lock down...
The ventilators you refer are in conjunction with intubation. Which needs a lot of other stuff...
do you think that if you use a STM32 or similar MCU can extract out data better from the MAX30102, than an uno???what's your opinion on the max30102??
I think I stated my opinion. If you follow the link of hte project in the video description you see that this guy used FFT to extract the pulse. You can definitively do more with a better CPU.
@@AndreasSpiess thanks a lot sir
should be nice to review body thermometers from aliexpress also. Big price differences and forehead thermometers are really popular now. I have now a forehead contact thermoemeter from Lidl and it works really perfect. But you need to read clearly the manual how to use and know the difference between contactless and contacted forehead thermometers. ;) But with Corona now all the thermometers are sold out and the price is very high.
Maybe I will have a look at the technological aspect. But, as you wrote, maybe later...
The coronavirus came here. So I guess I sit in the first row of the balcony just in case.
Anyway I like taking medical machines apart.
To be honest, CPAP machines are very interestingly engineerd. I repaired quite a few of those
I wondered what sensors they use to get such precise airflow and pressure values... I am sitting in the basement. But for me, this is not a big change ;-)
@@AndreasSpiess Some machines have a custom built sensor array. Others use a special version of a off the shelf sensor. It goes between the pump and the output. But the machine only measures airflow and pressure. The rest of the parameters are calculated. It is also common to see the machine monitor the pump itself.
Hello Sir ,
Will you please suggest the sensor name to use on wrist because i want to use reflectance method to measure the oxygen level and want to implement on wrist . because now days to many smart watches are coming with this feature and I want to test it and
if you have idea on reflectace method and if you know the sensor name which can be used on wrist or forehead please suggest .
You find the links to the products I used in the video description. I also made a second video about the topic.
@@AndreasSpiess
Yes sir i know that you have mentioned the sensor name and i am already aware of that sensor max30102 and but
my point is that can i use it on wrist and will it give accurate reading .
because to use on wrist i have to use reflectance oxymeter method .
I said in the video that i do not like it because it does not give exact values.
Thanks I found it interesting.
You are welcome!
Gread idea, and thank you, very useful for us..👍
You are welcome!
Great video! , i was wondering if i can read values of a normal oximeter using an arduino or esp32? , i do not wanna use a sensor but a real oximeter that i can buy , with a modification or idk! would i be able to do it?
I do not know :-(
@@AndreasSpiess thanks for answering
The esp32 has the capacity to auto connect or reconnect Bluetooth with the oximeter??
This depends on your sketch.
Did you try a restrictive bandage to reduce fresh blood flow to your fingers?
Depending on your cardio-vascular systerm, holding your arms up above your head for a couple of minutes ought to reduce the flow. There's not much hydrostatic head reserve in the human body.
Good idea! No. You see, I have no clue about medicine and the inner of the body...
having (a lot of) COPD, i use an Innovo INV-4301/PE . . . it seems stable, has long battery life and usually gives the same readings as my doctor's office machine.
just tested it putting my finger in sideways . . . it worked . . . might be a work-around for nail polish enthusiasts?
Thank you for sharing your experience! Iusually do not use nail-polish, but maybe you are right ;-)
inspiring, can we use similar devices to measure oxygen in water, this would be useful for aquaponic application
I do not think so. There is no haemoglobin
Thanks werry informatible video :)
You are welcome!
Well done 👍
Thank you!
Andreas, thanks for another excellent video, I was wondering whether this sensor could be associated with a LoRa device in order to update patients data to a LoRa gateway, I am asking it because in Emilia Romagna (Italy) there are gateways already in place in several City Halls, including my town, these collect all kind of data, from private and public sources, the access is free and the entire network is covered by a local public company which only reserves to mamage and store the big data.
Given the critical time, this infrastructure could serve the purpose of gathering medical data from patients within the Region.
Thanks Claudio
I assume you could connect it to a lora node. But the question is more: What problem do you want to solve which cannot be solved by a Smartphone. A smartphone is simple to use and already in all households.
I need a cheaper(low cost) sulphur dioxide sensor.... which one do you suggest??? A cheap one is more expensive than you think...
SO2 and SPO2 are not exactly the same ;-) As I say in the video, I am no chemist, so I do not know the purpose of such sensors...
@@AndreasSpiess I only asked for sensor... not any allergic chemistry question!!!😂
@@AndreasSpiess purpose may be of any type... but my section of question that is Which Sensor do U recommend to trace Sulphur Dioxide is in electronics section... which one is low-cost or how can I detect... and my use is in Exhaust Gas Analyzer that is used by Mechanical Engineers to find the usefulness of Engine Exhaust Gasses to use them or not decision like in Power Plants like in economizer, turbo engines etc to use exhaust gases etc.... my and your concern is only from sensor... remaining concern is by the user who want to analyze them..
Brilliant
Thank you!
Hi Andreas, I'm using the "Sleep As Android" app to track my sleep for multiple years now. While it's not opensource, the data it records is not closed down at all. I always saw the "Pulse Oximeter" option in the Wearable section of the app, but I thought that it only works with expensive devices. Looks like it expects to connect to any random Bluetooth device, and can record the data from them.
If you have energy for some further investigation, you can check out the app and see if either the "CMS50D+" or the others can supply data to the app. I would be very interested to know, as I would buy one of them if I could add another data point to my sleep tracking (which works either with the phone's accelerometers, with a "sonar like" motion tracking, or with wearables like the Mi band, Garmin, etc.)
Whether you look into this or not, thanks for the informative video about the subject! Here's the program's documentation page about the subject, if you want to check it out: docs.sleep.urbandroid.org/devices/oximeter.html
Thank you for the information. I was more interested in the technology of he sensors. So I think I will not cover the topic in the future. This is more a "maker" channel...
@@AndreasSpiess Sure, it just seemed you might be also interested in sleep tracking, so I wanted to point out this app in case you want to gather more data easily with your existing SpO2 sensors.
You said it is hard to manipulate oxygen levels. I was testing myself a few days ago and got 94-93%, just like other family members. "It is crap" that's what I told to myself. But then we went to the shop for a little apocalypse hoarding and when I got back it was 97%. Maybe just going to fresh air can increase it by a few percent?
During my test I also saw some changes depending on the movements I did. But I did no stress test.
m328p definitely has more than 800 bytes of ram. It has 2k and those 2 buffers fit just fine in them.
I did not check it...
"the Arduino doesn't have enough SRAM to store 100 samples of IR and red data in 32-bit format"
This only shows that most Arduino-libraries are very poor at resource management. 200 values in 32-bit format only need 800 bytes and the ATmega328 has 2 kByte of SRAM...
I did not investigate furter When I saw the values jumping I started to doubt that you get reliable measurements even with more "power". But I might be wrong, of course.
@@AndreasSpiess I would just assume a very badly written library...
It is for sure possible to implement the complete detection reliably in an ATmega328 including the control of the two LEDs.
Yes we are talking data analysis here, but we are looking at very slow signals, data rates of 200 samples per second and in the case of the Uno about 16 MIPS