Group Actions
Engagement
  • Author Author: fustini
  • Date Created: Date Created
  • Views 1356 views
  • Likes 3 likes
  • Comments 4 comments
Related
Recommended

Biosensor crunch time

fustini

Our biosensor array team worked late into the night on Saturday at Pumping Station: One.  With the hackerspace challenge build deadline on Tuesday night, it's definately crunch time!

 

image

Thankfully we've got a nice array of working sensors (in the order of signal speed):

 

  • EKG
  • Pulse oximeter
  • GSR (Galvanic Skin Response)
  • CO2
  • Temperature

image

 

Dan yet again played the role of test subject:

 

image

The QRD1114QRD1114 sensor from SparkFun contains an IR LED and a compatible phototransistor.  It dectects pulse via the change in the amount of infrared light reflected by finger as blood pumps through it.

image

We observed a great correlation between EKG and pulse oximetry signals:

image

 

Cheers,

Drew

  • Hi Drew,

      ECG and PPG (Photoplethysmograph) signals are very highly correlated. Consider your heart like an electric pump. ECG talks about the armature current and PPG talks about the pumping action. PPG is also correlated to the pressure at which the blood has been pushed through the blood vessels. Algorithms used to calculate the BP using cuffless monitors use this correlation.

    Ventricular Depolarization happens -> ventricles contracts -> when the pressure increases more than 120mmhg -> aortic valve opens -> bloods gets squeezed out from the chamber -> it travels through the vessels and reaches you finger tip where you are measuring the PPG.

    calculate - pulse transit time (PTT). This will be proportional to the pressure (or pulse wave velocity). But we are all different and we do have slightly different anatomy. Our blood vessel thickness and elasticity is different. This is a variable parameter in the BP calculation. So we have to perform calibrations with different readings and finally make a regressive model to calculate the BP. (Blood vessels looses its elastic property over time so calibrations is required periodically)

     

    Regards,

    Aswin

  • in reply to Former Member

    Hi Ron,

     

       I think what you are facing is an arrhythmic condition. Probably tachycardia. A pulse rate monitor doesn't give you insights to what kind of arrhythmia you are facing. It can be due to Ventricular issues or Atrial issue (It can also be atrial flutters). What you need is a holter monitor (Record full ECG for longer period of time). You can build a cheap ECG monitor using ADS1292 (Texas Instruments), MAX30001( Maxim Integrated) or AD8232 (Analog  Devices). Cuffless BP monitors are still under R&D phase. Cuff-based BP monitors will have low temporal resolution. But once in every minute will do the job. ( If you are interested to do some research you can read up on using PPG (The raw waveforms from Pulse Ox meters) and ECG to monitor BP. Calculate pulse transit time and pulse arrival time and using these you can estimate BP with calibration.

     

    MIT-BIH has a lot of annotated ECG database. You can train a system using that and perform inference on your ECG monitor to classify the Arrhythmia.

     

    If you are looking for cardiac damage look at ST Elevation/ Depression. From ECG you can derive this parameter. If you are facing cardiac tissue damage due to lack of oxygen supply to the heart muscles (Ischemia) it will reflect on your ST values.

     

    SpO2 (Relative oxygen saturation) or oxygen level as you mentioned is actually a measure of your lung function. SpO2 is measured at places where you have a lot of arterioles (like finger tips or earlobes). If blood carries enough oxygen the SpO2 value will be above 95. If its less than 90 something is going wrong with your lungs.

     

    I recently wrote an article on how to use MAX30100 PPG(photoplethysmograph) sensor with BeagleBone black. Oxygen Saturation - SpO2 Measurement

    It's not really well written but I explained the steps there.

     

    Note: I assume that you are already consulting a cardiologist on the problem you have explained. If not thats something you should do at the earliest.

     

    I will be happy help you. I am a biomedical engineer with research experience.

     

    Regards.

    Aswin

  • ron5329, you mention your BP drops and your heart rate is rocketing? Please look into Postural Orthostatic Tachycardia Syndrome, this may be what you are suffering from. I have been diagnosed with this and it is the reason I am interested in building this project. If you have a heart monitor you can test yourself, lie down for 10-20 minutes, completely flat, no moving or talking and relax. Work out your resting heart rate. Then immediately stand and stand as still as possible, no moving or talking or fidgeting. Monitor your heart rate for the next 30 minutes. If your heart rate increase more than 20 bpm (ie, goes from 60 lying down to 80+ standing) this is considered abnormal. If within 10 minutes of standing your heart rate increase more than 30 bpm, or at any point goes over 120bpm, then you meet the diagnostic criteria for POTS. I hope this is helpful, your heart should not be increasing that much without significant exercise. Please see a Doctor, preferably a cardiologist and mention POTS.

  • Very interesting, I want to build something like this to monitor BP,PR,temp, and oxygen level with the addition of a way to record all of these things when they are off normal. With A-fib sometimes my BP drops, and my PR goes skyhi and bounces all over the place from 98 to 170. Sometimes for only a couple minutes sometimes for hours. I can tell when it's out of whack but it would be nice to be able to monitor it. Thanks for your comments. ron5329