HealthyPi Python Interface

A Python Interface for ProtoCentral HealthyPi v3

Introduction

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.

Usage

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.

Development Note

Streaming Packet Format

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

Offset	Byte Value	Description
0	0x0A	Start of frame
1	0xFA	Start of frame
2	Payload Size LSB
3	Payload Size MSB
4	Protocol version	(currently 0x02)
5-6	ECG Value	Signed int 16,LSB first
7-8	Respiration Value	Signed int 16,LSB first
9-12	PPG IR Value	Signed int 32, LSB first
13-16	PPG Red Value	Signed int 32, LSB first
17-18	Temperature	Signed int 16, LSB first
19	Respiration Rate	Unsigned int 8
20	SpO2	Unsigned int 8
21	Heart Rate	Unsigned int 8
21-23	BP (not implemented)	Not implemented yet
24	SpO2 and ECG lead status
25	0x00	Footer
26	0x0B	End of Frame

Precautions

• 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"])

Reference

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