Brain blood flow biofeedback meets affordable IoT technology

Dec 12, 2019

Project update 7 of 10

Firmware Fixed, Data Output UI, and First Kits Assembled

Helloooooooo everyone!

Time for another long overdue update. We’ve been banging our head on a table for a couple weeks now trying to figure out what we messed up in our own firmware, but we finally fixed it. Now let’s see some actual output from this thing!

Raw sensor output via serial. The Ratio = Red / IR.

What you are seeing above is two LED readings (Red in dark blue, IR in light blue) that are shown at different intensities based on the resistors we are using. This produces a ratio, seen in orange (and a filtered ratio which minimizes variance for a more solid trendline), and this is our primary metric for whether our brain blood oxygen is increasing or decreasing. More red over infrared = more O2, a simple approximation! This ratio is essentially a DC + AC blood volume reading which tells you the general O2 perfusion levels at a given site. Our ratio, in an averagely lit room, only has about a variance of about 0.02, pretty good! When training you can almost double the returned ratio just by increasing blood flow, so this data gives us a nice and tight error margin to go by (since your effort can create wide swings in the ratio), and 1 to 6 seconds of a moving average (or other filtering methods) quiets any noise. We use some smoothing in our software to improve user experience and visualize your physical effort in a more palpable way.

HEGstudio, controlling brightness with a simple moving average of a roughly calculated oxygenated blood flow ratio.

We have more software rolling out soon; one of our partners is working on a ThreeJS plugin that grabs the HEG interface via an iFrame to do its own rendering with the data and functions we provided in our API. This has opened up a lot of ideas for us considering just how flexible this framework is. We’ve also improved BLE thanks to Diego’s help from the Statechanger team and we’ve wrapped up most of our bugs on the current featureset.

This is representative of what you'll be getting (minus some safety tape).

We’ve started assembling our first bulk load of kits today while all our parts trickle in this week. We’re getting about 200 together to fill pre-orders and keep some stock, with plenty of parts to spare after thanks to all of your generous funding. We found out our costs actually decreased over time by a lot, so we splurged and got really nice USB cords that will take a long time to degrade, as well as got extra stock for future orders. We are so happy with how everything turned out, we cannot thank you all enough for helping us make a dream come true and we are looking forward to your impressions and training data!

If you are on Hackaday, we just posted our project and entered in the Adafruit Feather competition!

And if you missed it, I posted interview with Bob Marsh a couple weeks back, one of the creators of nIR HEG and an all around great guy.

Sign up to receive future updates for HEGduino.

Subscribe to the Crowd Supply newsletter, highlighting the latest creators and projects