You are on page 1of 3

Want to help us reduce stress in the world?

Introduction
● We are working to develop a mobile application that aims to help individuals learn to
reduce stress through meditation proved by biometrics such as Cardiac Coherence,
HRV (heart rate variability) and RR (respiratory rate).

What is our Goal?


● To extract a valid IBI signal and derive cardiac coherence AND/OR Respiratory Rate
directly from the camera of the user's smartphone.

How to Calculate IBI?


● IBI can be plotted in a graph over time - see Section 5 Figure 6 of this paper:
http://www.es.mdh.se/pdf_publications/1929.pdf - so though it hasn't been explicitly said,
IBI is calculated through an array of timestamps.
● For example, every time an R peak is detected it is logged (RP = R peak): `[{RP1:
12:32:10.4},{RP2: 12:32:11.1},{RP3: 12:32:12.3}]`
● The IBI value for RP2 is the difference of the two: IBI for RP2 = RP2-RP1
● In this case, the IBI for RP2 = 0.7

How to Calculate Coherence?


For cardiac coherence to be accurate, millisecond precision between RR intervals is
important.

Physiological coherence is reflected in more ordered and sine wave-like HRV patterns at a
frequency of around 0.1 Hz (10 seconds rhythm). A coherent rhythm can be defined as “a
relatively harmonic (sine wave-like) signal with a very narrow, high-amplitude peak in the LF
region of the HRV power spectrum with no major peaks in the VLF or HF regions. Coherence is
assessed by identifying the maximum peak in the 0.04–0.26 Hz range of the HRV power
spectrum, calculating the integral in a window 0.03 Hz wide centered on the highest peak in that
region, and then calculating the total power of the entire spectrum. The coherence ratio is
formulated as: [Peak Power/(Total Power - Peak Power)]” (McCraty and Childre, 2010; p.14).
https://www.ncbi.nlm.nih.gov/pmc/articles/PMC4179616/

Which Computer Vision Technologies to Use?


Technologies that we are currently exploring are listed below but if you have another effective
solution we are happy to consider it:

1. Eulerian Video Magnification (EVM) as described here:


a. https://people.csail.mit.edu/mrub/vidmag/
b. https://www.youtube.com/watch?v=e9ASH8IBJ2U
2. Remote-PPG (rPPG) as demonstrated here:
a. https://sites.google.com/site/rppgwenjin/
b. https://www.youtube.com/watch?v=Q9MK-vtWzUM
c. https://ieeexplore.ieee.org/document/6894148/?tp=&arnumber=6894148&queryT
ext%3Dmotion%20robust%20rPPG
3. ICA:
a. https://ieeexplore.ieee.org/document/7882667/
4. Open-Source Github Libraries:
a. There are some Open-Source Github libraries we can start with, just Google
“eulerian video magnification github”, “rppg github”, “hrv github”, “respiratory rate
github”
b. https://github.com/topics/eulerian-video-magnification
c. https://github.com/topics/heart-rate-variability
d. https://github.com/topics/rppg

Other References
There is already an app on the market called Cardiio that can read the pulse from using the
front-facing camera (https://www.cardiio.com/)

Measurement of heart rate variability using off-the-shelf smart phones:


https://www.ncbi.nlm.nih.gov/pmc/articles/PMC4731953/

Remote heart rate variability for emotional state monitoring: https://hal-univ-bourgogne.archives-


ouvertes.fr/hal-01678244/document

DistancePPG: Robust non-contact vital signs monitoring using a camera:


https://arxiv.org/pdf/1502.08040.pdf
We also want to be able to calculate Breathing Rate (Respiratory Rate):
https://www.youtube.com/watch?v=GSdGZ-E_bvs

What Next?
Apply if you have any expertise to help with this project or know any other technology we can
use for extracting stress-related user biometrics using no additional hardware but the typical
smartphone.

Current Requirements:

- Previous programming experience working with computer vision and/or image


processing
- Optional: technical knowledge of sensor inputs of modern smartphones
- Fluent spoken and written English
- Join our 9am PST discussion calls

Looking forward to having you on our team!

~Conscious

You might also like