Human vital signs simulator for radar
Supervisors: Kristian Kjelgård, Mats Høvin, Dag T. Wisland, Tor Sverre Lande
Remote, non-contact monitoring of human vital signs like breathing and heart rate may be useful in several critical situations. Individuals like patients, drivers, pilots or operators may benefit from vital signa monitoring. Reports indicate breathing and heart rate frequency patterns are good indicators of state of mind (drowsiness, fatigue, stress...). Perhaps most interesting is fatigue which can be detected by heart rate variability (HRV) .
The most promising non-contact technology for detecting both breathing and heartrate remotely is short-range radar. Electromagnetic waves penetrate cloths easily and dielectric properties of the human body (high permittivity) result in good reflections from the body surface (air/body). Several studies have shown promising results using radar for heart and breathing . Especially, the pulsed radar developed by Novelda/XeThru has shown good results as presented in . Such systems are measuring breathing and heart rate by sensing chest surface movements. Movement due to breathing is significant and may be detected quite reliable even when the person is moving around. But chest vibrations due to heart rate is very small (<0.5mm) and is ”mixed” with much larger breathing vibrations making heart-rate detection difficult. Detecting heart rate when a person is moving around is even more challenging.
In this project we propose to design and build a mechanical simulator for validation of vital sign radar sensing. The complex, uncorrelated mix of breathing and heart beat chest movements results in higher order, overlapping harmonics (non-sinusoidal movements) and must be replicated. The dynamics are also large with a mix of small heart-beat chest movement in the millimetre range superimposed on a much more larger breathing movement. The proposed simulator is of interest not only to the scientific community but also for the industry including alternative sensing technologies like ultrasound.
Find a breathing and heart movement pattern space based on a literature study and using X4 radar data. Care should be taken to preserve the dynamics including non-sinusoidal behaviour. Design a mechanical simulator with the following properties:
A single moving radar reflector with radar cross section like a human
Programable breathing and heart frequency and amplitude
Breathing movement in cm range and heart movement in 100 um range
Realistic dynamic movement curve
Validate some of the most promising signal processing algorithms with a X4 radar module.
TEK4030 – Styring av manipulatorer og mobile roboter
TEK5150 – Radar - systemer og signalbehandling
IN5450 – Array Signal Processing