Heart rate and other sensors
Heart rate and other sensors
I'm using Mind Monitor to stream OSC data into Pure Data for further processing and enjoyment, and for that it works wonderfully. I note that the Muse 2 has additional sensors such as heart rate and breathing; is it possible to stream this data as well?
Re: Heart rate and other sensors
Not currently, but I'm working on it. I have access to the PPG sensor data in the new version of Mind Monitor that I'm working on, but getting the heart rate from the raw data is proving difficult as I'm only self taught with DSP (digital signal processing).
Note that there is no "breath" sessor. The PPG data will only provide heart rate, and blood oxygen. Blood Oxygen level might be what Interaxon are refering to as "breath".
If there are any experts in DSP who'd like to help me figure out the PPG data, let me know!
Note that there is no "breath" sessor. The PPG data will only provide heart rate, and blood oxygen. Blood Oxygen level might be what Interaxon are refering to as "breath".
If there are any experts in DSP who'd like to help me figure out the PPG data, let me know!
Re: Heart rate and other sensors
What's the raw output? Light level returned from each wavelength?
Re: Heart rate and other sensors
It's light level, but it jumps like crazy when you move your head even slightly, so it's needs some serious data smoothing in order to do peak-to-peak calcs for heart rate.
Re: Heart rate and other sensors
It will be very useful to stream the other sensor data.
My guess is that we can first apply a low-pass filter to remove some high-frequency noise. Then apply some kind of Fourier transform to convert the data from time-domain to frequency-domain and read out the frequency of the pulse. It could be tricky, though. I compared the pulse rates reported by my smart watch, fingertip pulse oximetry, and treadmill, only to find most of the time they disagree, sometimes by quite a lot, especially when the heart rate is changing. I guess some may respond faster than others, depending on how they set the parameters such as the filtering constants (of course the sensors are also different). All I am trying to say is that it is not very straightforward to generate the numbers or interpret the numbers.
My guess is that we can first apply a low-pass filter to remove some high-frequency noise. Then apply some kind of Fourier transform to convert the data from time-domain to frequency-domain and read out the frequency of the pulse. It could be tricky, though. I compared the pulse rates reported by my smart watch, fingertip pulse oximetry, and treadmill, only to find most of the time they disagree, sometimes by quite a lot, especially when the heart rate is changing. I guess some may respond faster than others, depending on how they set the parameters such as the filtering constants (of course the sensors are also different). All I am trying to say is that it is not very straightforward to generate the numbers or interpret the numbers.
Re: Heart rate and other sensors
If anyone would like some 52hz raw ppg data to play with algorithms, email me and I'll give you some test recordings