A Python Interface for ProtoCentral HealthyPi v3

HealthyPi is a fully open-source vital signal monitoring system with Raspberry Pi. The HealthyPi add-on HAT would work independently to collect vital signals via package streaming.

This repository provides a Python interface for data collection and simple monitoring window that works on any platform that can run with Matplotlib. This repository also demonstrates the streaming package decoding method as an example for interface development in other languages.

Use python main.py -h to print out the detail user manual.

Example usage:

1. Connect the HealthyPi v3 HAT to your computer via USB interface
2. Check the communication port. In Windows you may check with device manager:
3. Run the Python interface with command:

   python main.py -p COM3 -g -v

   -g refers to the Matplotlib graphical interface for logging; and -v refers to verbose output from decoded data package
4. Press CTRL + C to quit (This is a proper way to close communication, you may need to manually close thread if you don't use this way to close.)
5. If you want to output collected data in CSV format, you can add the command option -c. The collected data will be stored in ./data/healthypiv3_yyyymmdd-HHMMSS_Numerics.csv and ./data/healthypiv3_yyyymmdd-HHMMSS_Signals.csv. The data formation in CSV file follows the pattern of BIDMC PPG and Respiration Dataset.

The HealthyPi sends data out on the serial port in the following packet format.

• Matplotlib rendering is slow (~10FPS) when comparing with data collection speed (~120Hz). The synchronized data collection currently not able to provide smooth plotted curves.
• The PPG probe provides both infrared and visible light signals. By default the we display in Matploblib with IR signals but you may change the line

   y_ppg = np.append(y_ppg,collector.data[-1]["ppg_ir"])

  into

   y_ppg = np.append(y_ppg,collector.data[-1]["ppg_red"])

• ProtoCentral HealthyPi v3 Official Site
• ProtoCentral HealthyPi v3 Github Repository
• BIDMC PPG and Respiration Dataset