HRMonitor

Heart Rate Monitor Post-Processing

Developed By: Harvey Shi (@rvshi)

Course: Duke BME 590-05 Medical Software Design (Spring 2018)

Overview

This project defines a Python 3 class, HRMonitor, for processing ECG data.

• In order to parse the data, a custom DataHandler class is used, to allow further extensibility in the future.
• After being parsed, the data is analyzed internally by the module to determine useful attributes such as heart rate or duration of signal.
• For more detailed information on the individual class methods, see the documentation site.

Usage

• First, you want to set up a Python virtual environment and install the packages in requirements.txt.
• Next, you can import the HRmonitor class from hrmonitor.py and start using this module in your code.
• Here is a basic example of usage, which you can run from the root directory of this repo:

from hrmonitor import HRMonitor
hr = HRMonitor('./test_data/test_data5.csv')
hr.plot_data()

Which produces the following files:

./test_data/test_data5.json:

{
    "peak_interval": 0.808,
    "mean_hr_bpm": [74.25743, 73.71007, 74.25743],
    "voltage_extremes": [-1.155, 1.72],
    "duration": 27.775,
    "num_beats": 36,
    "beats": [0.061, 0.886, 1.714, 2.511, 3.306, 4.081, 4.869, 5.736, 6.564, 7.372, 8.181, 8.986, 9.772, 10.556, 11.403, 12.25, 13.075, 13.886, 14.678, 15.489, 16.258, 17.072, 17.928, 18.742, 19.569, 20.378, 21.178, 21.958, 22.742, 23.111, 23.411, 23.875, 24.436, 25.253, 26.058, 26.867]
}

./test_data/test_data5.png:

Some important usage notes

• Input ECG data should be in .csv format.
• Only two float values should be present on each line of the .csv, in this order: time, voltage.
• Time units are assumed to be continuous, i.e. with roughly the same interval between every time entry.
• Example ECG files can be found in the test_data/ directory.
• The data units can be specified by setting the time_units and voltage_units arguments in the HRMonitor constructor function.
  • The default units are seconds and mV.
  • time_units is a float representing the new unit of time in terms of seconds.
  • voltage_units is a float representing the new unit of time in terms of seconds.
  • For example, if the units were milliseconds and volts, you would specify this as:

HRMonitor('./file.csv', time_units = 0.001, voltage_units = 1000)

Features

• Calculates several class attributes from the data:
  • time: numpy vector of the time data
  • voltage: numpy vector of the voltage data
  • peak_interval: interval between all peaks in the signal, as estimated via autocorrelation of the signal (in seconds)
  • mean_hr_bpm: numpy vector containing the average heart rate (in bpm) for contiguous bins with a size specified by the user
    • default chunk size is 10 seconds
    • calculated by obtaining peak_interval over chunks of the data
    • this can be recalculated by running the get_mean_hr(<window_size>) function, where window_size is the size of the chunks in seconds
  • voltage_extremes: tuple of the (min, max) of the voltage data
  • duration: total length of the signal in seconds
  • beats: numpy array with the approximate time locations of heart beats via peak detection following a bandpass filter
  • num_beats: estimated number of beats in the data, taken as the length of beats
• Ensures that the input data is consistent to the (time, voltage) format.
  • Makes sure that the data consists of pairs of floats.
  • Performs interpolation for pairs with missing or invalid values.
• Exports some calculated attributes as a JSON file with the same name as the input .csv.
  • time and voltage are not exported since they are already in archival format.
• Can be used to generate plots of the data using the plot_data() method.

Other notes

The current module has only been tested with Python 3.6.4 on MacOS 10.13