Oppgaven er ikke lenger tilgjengelig

Heart Rate Variability system

Heart rate variability (HRV) is a long-term analysis of human heart beat frequently explored for analysis of human state-of-mind. Presumably HRV may also be used to assess human vigilance or for early warning of critical health conditions like heart attack. Long-term monitoring of human heart rate normally requires some sensor to be work on the human body like a strap around the chest or possibly a vital sign sensing arm watch. Heart beat sensing without body contact is hard to do, especially when we are moving around. Even when we are not moving, the minimal vibrations of our body surface due to heart beats are hard to sense.

Recently the Norwegian company, Novelda/XeThru released anew short-range radar system known as X4M200 (https://www.xethru.com/x4m200-respiration-sensor.html) with embedded software for breathing detection. The sensitivity of this pulsed radar system is sufficient for picking up body vibrations due to heart rate at close range. Limiting heart-rate sensing to humans in stationary positions (i.e. sitting in a chair or lying in bed) we might be able to detect heart rate and collect data for heart rate variability analysis.

A typical stationary position for students at IFI is sitting in front of a screen working. The XeThru module is powered by USB and could easily be mounted facing the person sitting in front of the screen.

The project is to use available software from XeThru for collecting data and develop code for HRV measurements of computer operators. Some proposed milestones:

  1. Reliably measure heart rate (HR) of person in from of the computer using adequate signal processing (filtering).

  2. HRV analysis for mental state assessment. A number of methods are proposed and some software is also available (MATLAB). See http://www.sciencedirect.com/science/article/pii/S0169260713002599 for more info and adopt and implement code for the following mental states:

    1. Drowsiness or sleepiness

    2. Mental stress

    3. Fatigue or exhaustion

  3. Develop a suitable classification system possibly by using machine learning indicating if these mental states are observed for the individual in front of the screen.

This project will require skills in low-level programming, instrumentation, signal processing and simple classification/adaptive software/machine-learning.

Publisert 26. sep. 2017 10:48 - Sist endret 10. okt. 2018 19:36

Veileder(e)

Omfang (studiepoeng)

60