Description

This algorithm is derived from the Brainflow SpO2 algorithm and uses PPG-R and PPG-IR data from the EmotiBit to calculate SPO2 level. The repository has been setup to enable using the spO2 source code in both Arduino and a standalone python script. We did this to enable testing source in pyhton, to fast track testing and enable us to improve algorithm performance with a shorter loop.

• Arduino: To use this library in Arduino, just clone the git repo or download it using the Arduino Library manager into your Arduino/libraries directory.
• Python:
  • We use pybind11 to create bindings for the c++ source. The resulting pyd file is then used by the Python script.
  • To learn more about using pybind and python, check out the pybind/README.md.

Scripts

run.py <path_to_data_folder>

Runs the Brainflow algorithm on the provided data folder (which needs to contain a separate .csv file for PPG-R and PPG-IR from the EmotiBit ending in PR.csv and PI.csv) and saves the calculated SpO2 with timestamps to calculated_spo2.csv. Also shows a plot of the PPG data overlayed with the calculated SpO2.

Performance

Plots were created using scorer.py from the EmotiBit Biometric Lib repo

sit-stand-sit_v0.0.0

Device	Custom Dataset?	Resample Test	Scatterplot	Mean-Difference Plot
IP900AP	No

simulated-unobstructed-airway_v0.0.0

Device	Custom Dataset?	Resample Test	Scatterplot	Mean-Difference Plot
IP900AP	No