Collect PPG data from Muse S

[ErikBjare]

• [x] Basic support
  • Basic support implemented in https://github.com/ErikBjare/thesis/commit/9a4b89ab7c6c07f0579c47ae89e458a068aac5cf
  • Might have some issues though, see https://github.com/alexandrebarachant/muse-lsl/issues/110
• [x] Test resilience during longer recordings
  • There seems to be some issues, see #11.
• [ ] Convert PPG data to actionable features (such as HR, HRV).
  • Unclear how to do this from the PPG1, PPG2, PPG3 columns in CSV.

[chatchavan]

TL;DR: I suggest using PPG2.

PPG data has three channels: ambient, infrared, red (source: muse-lsl code). Muse S has a green LED instead of the red one.

Although you have obtained data from all three PPG channels, your code currently uses PPG1.

I've logged Bluetooth packets that the iOS Muse app communicates with my Muse S while using the "heart meditation" (using XCode PacketLogger). The log shows data from only the PPG2 channel. From the sensor, only the green light appears. Therefore, I suspect that Muse S uses the green light and stream it through the PPG2 channel.

Maeda et al. (2008) correlates signals from infrared vs. green light with a reference signal from ECG. They found that green light has a better correlation than infrared. Therefore, it is possible that Muse S uses only the green light at the moment.

[ErikBjare]

Very useful info, thanks!

On Sat, Jul 24, 2021, 09:35 Chat Wacharamanotham @.***> wrote:

TL;DR: I suggest using PPG2.

PPG data has three channels: ambient, infrared, red (source: muse-lsl code https://github.com/alexandrebarachant/muse-lsl/blob/master/muselsl/constants.py). Muse S has a green LED instead of the red one.

Although you have obtained data https://github.com/ErikBjare/thesis/blob/master/data/eeg/muse/subject0000/session001/recording_2021-07-15-11.32.37.PPG.csv from all three PPG channels, your code https://github.com/ErikBjare/thesis/blob/master/notebooks/PPG.ipynb currently uses PPG1.

I've logged Bluetooth packets that the iOS Muse app communicates with my Muse S while using the "heart meditation" (using XCode PacketLogger https://www.bluetooth.com/blog/a-new-way-to-debug-iosbluetooth-applications/). The log shows data from only the PPG2 channel. From the sensor, only the green light appears. Therefore, I suspect that Muse S uses the green light and stream it through the PPG2 channel.

Maeda et al. (2008) https://doi.org/10.1109/IEMBS.2008.4649649 correlates signals from infrared vs. green light with a reference signal from ECG. They found that green light has a better correlation than infrared. Therefore, it is possible that Muse S uses only the green light at the moment.

— You are receiving this because you authored the thread. Reply to this email directly, view it on GitHub https://github.com/ErikBjare/thesis/issues/9#issuecomment-886014536, or unsubscribe https://github.com/notifications/unsubscribe-auth/AAKXDOTXOO7QMYIYACOIT33TZJUKJANCNFSM4S7AY7BQ .