Breathing Back Pressure Data Analysis
I've uploaded some datasets of the blower in different modes. I also measured the current (blue) and whether I am exhaling or inhaling (green). The breath signal is generated by me pressing a button. Visualizing the current data suggests that it looks very easy to detect the breath back-pressure in CPAP mode (top plot). The middle plot shows the regular shape of the current curve in each breath. The bottom plot shows the irregular shape that occurs when the breath is out of sync with the blower cycle. This suggests it would be possible to detect the patient's breath cycle, or adapt to minimize out of phase behavior. Finding a way to get phase alignment is harder. It would be great to get help with this since I don't have great ideas on how to do this at the moment.
See Colab notebook to easily visualize this data on your own.
- Arduino capable algorithms for doing phase/frequency alignment between the blower cycle and the breathing cycle.
- [Done]
Jupyter notebook integration with this Github project to enable fast experimentation.
These plots were generated with the python visualization code.
Current sensing uses this version of the schematic. It uses a 0.1 ohm 1 watt current sensing resistor on the low side of the BLDC power supply. As the current draw by the BLDC motor goes up, the voltage across the sense resistor increases. This voltage passes through a small RC filter to smooth out noise before being read by an analog input on the Arduino. The reference voltage is set to the 1.1V internal reference which provides ~1mV of resolution using the 10bit ADC which corresponds to ~10mA of current resolution.
