callibri_ecg 1.0.0

Flutter wrapper for CallibriECG library

Mathematical library for working with ECG data from the Callibri sensor.

The main functionality is the calculation of cardio-interval lengths, heart rate and Stress Index (SI).

During the first 6 seconds the algorithm is learning, if no 5 RR-intervals are found in the signal 5 RR-intervals are not found, the training is repeated. Further work with the library is iterative (adding new data, calculating indicators).

Initialization

Determine the basic parameters

1. Raw signal sampling frequency. Integer type. The allowed values are 250 or 1000.
2. Data processing window size. Integer type. Valid values of sampling_rate / 4 or sampling_rate / 2.
3. Number of windows to calculate SI. Integer type. Allowable values [20...50].
4. The averaging parameter of the IN calculation. Default value is 6.

Creating a library instance

Firstly you need to determine lybrary parameters and then put them to library. Tne next step is initialize the filters. In the current version the filters are built-in and clearly defined: Butterworth 2nd order BandPass 5_15 Hz.

You can initialize averaging for SI calculation. It is optional value.

// 1. Raw signal sampling frequency
int samplingRate = 250;
// 2. Data processing window size
int dataWindow = samplingRate / 2;
// 3. Number of windows to calculate SI
int nwinsForPressureIndex = 30;
final math = CallibriEcgMath(samplingRate, dataWindow, nWinsForPressureIndex);

// optional
// 4. The averaging parameter of the IN calculation. Default value is 6.
int pressureIndexAverage = 6;
math.setPressureAverage(pressureIndexAverage);

Initializing a data array for transfer to the library:

The size of the transmitted array has to be of a certain length:

• 25 values for a signal frequency of 250 Hz
• 100 values for a signal frequency of 1000 Hz

final rawData = List<double>.generate(25, (i) => i);
// or
final rawData = List<double>.generate(100, (i) => i);

Optional functions (not necessary for the library to work)

Check for initial signal corruption. This method should be used if you want to detect and notify of a distorted signal explicitly.

if (math.isInitialSignalCorrupted){
    // Signal corrupted!!!
}

Work with the library

1. Adding and process data:

math.push(samples)
math.process()

2. Getting the results:

if(math.rr_detected){
  // check for a new peak in the signal
  // RR-interval length
  double rr = math.rr
  // HR     
  double hr = math.hr
  // SI
  double pi = math.pressureIndex
  // Moda
  double moda = math.moda
  // Amplitude of mode
  double amplModa = math.amplitudeModa
  // Variation range
  double variationDist = math.variationDistance
  math.checkRR()
}

Finishing work with the library:

math.dispose();