Fill out the survey to improve HRW prototype in the future releases

HRW alpha version

Comming on 2015

Heart Risk Warner alpha application. Bug fixes and simplified heart risk assessment using Bluetooth.

- improved PPG feature extraction
- various fixes

setup 1.0.7.msi

heart risk warner

hrw setup (alpha).msi

heart risk warner

HRW alpha version

2014 September 17

2014 September 7

Heart Risk Warner application. Older OS like WinXP requires:
- Microsoft Windows Installer 3.1, dowload from official Microsoft page
- Microsoft .NET Framework 3.5, dowload from official Microsoft page
Win7 or newer OS versions already have installed .NET 3.5 framework. Installer package is known to run on WinXP SP2/Win7/Win8

Main features:
- Database allowing store physiological information for further analysis
- C++ ECG feature extraction library
- PVC heart beat detection using artificial neural networks (not veriefied)
- Resting heart rate measurement with PPG or/and ECG sensors
- Heart rate variability
- Vascular age prediction
- Pulse transit time estimation
- Stiffness index estimation
- Blood pressure estimation (not reliable)
- 32feet.NET libraries for Bluetooth conectivity
- USB HID for device configuration
- DSP libraries that include various filters
- Supported MIT-BIH Arrhythmia Database files with '212' formats
- GUI for signal visualisation

setup (demo).msi

software (demo)

2014 April 7


Version 1.5.4
- Auto turn off to save power then PPG and ECG are not present
- Turns on only then PPG cable is present
- Auto restore after error


flowchart 1.5.4.pdf

2014 August 3

Description: Upon power up initialization starts with single beep signal. After successful initialization arrives second beep signal and the middle yellow led start blinking in a short intervals. Data collected at the following rates: ECG @ 500 Hz, PPG @ 100 Hz, accelerometer @ 120 Hz, temperature readings every 1 second. At the same time data can be sent over Bluetooth or/and stored to SD card. Prototype is not optimized for power saving, thus working duration is approx. 8 hours and standby duration is approx. 36 hours.
How turn it OFF: Remove PPG cable and put it still. If it is powered from USB then reset it and remove USB and PPG cables. Then put it still. How turn it ON: plug PPG cable in, then should appear beep signal
Problems: Data corruption (best performance over Bluetooth)
New features: Works with or without SD card. Data format (not tested)


System initialization (flowchart)

2014 April 7

First release that may requires some fixes.
- send data over Bluetooth
- send data over USB
- store data into microSD
Note: Micro SD card must be always inserted, otherwise you get an error.




System start (flowchart)

2014 Mar 25

2014 Mar 24

2014 Feb 22


Record samples recorded in a real time using Bluetooth.




PhysioBank it is a large archive of well-characterized digital recordings of physiological signals. Some records can be read and anlyzed by HRW software.

MIT-BIH Arrhythmia Database

The MGH/MF Waveform Database

Record sample include ECG, PPG, AXL and TMP signals. Signal sampling rates (SR): ECG (500Hz), PPG (100Hz), AXL (120Hz), TMP (1Hz). Average heart rate estimated by ECG/PPG 73 BPM, record duration ~4 min. Record was created by prototype and recorded into micro SD card.


Artificial neural networks (ANN)

ANN is the important element of algorithm it allows to classify every heart beat. Here are different ANN training sets which may have different results in evaluating arrhythmias. Each ANN was trained using only single heart beat type therefore currently lacking performance to classify heart beats correctly. Combining heart beats of ill and healthy individuals it would be possible to create classifier's accuracy close to 99%. Try download new ANN and change it with existing. Rename it with the same name as it was before.












Software Tools

Parses *.ecg binary data files into readable form of text files. Code sample is a Visual Studio 2012 solution.

2015 Heart Risk Warner