PsyLink An open source neural interface for non-invasive EMG

We don’t see a lot of EMG (electromyography) projects, despite the freshness of the apps. This could be due to technical difficulty seeing the tiny electrical muscle signals among the noise, it could be difficulty interpreting any signal you find. All the same, [hut] strove to move forward with a flood of prototypes, resulting in the aptly named “Prototype 8”

The current prototype uses a main power board host a Arduino Nano 33 BLE Sense, as well as a boost converter to pump the AAA battery to supply 5 volts to the Arduino and a selection of connected EMG amplification units. the EMG sensor revolves around the INA128 instrumentation amplifier, in a fairly simple configuration. EMG samples along with IMU data on the Nano 33 BLE Sense are transmitted to a PC connected via Bluetooth, running the PsyLink software stack. This is based on Python, using the BLE-GATT library for LV communications, PynPut the reset of the PC input devices (to emit keyboard and mouse events) and tensorflow for the machine learning side of things. The idea is to use machine learning from EMG data to associate with a specific UI event (like a key press) and with a little practice, to be able to play games. games on the PC with just hand / arm gestures. IMU data is used to augment this, but in this demo it’s not entirely clear.

An earlier prototype of the PsyLink.

All hardware and software can be found on the codeberg project page, which made us think twice about why GnuRadio was used, but come to think of it, it’s really good for signal processing and visualization. What a good idea!

Obviously, there are plenty of other use cases for such an EMG controlled input device, but who doesn’t want to play Mario Kart, you know, for science?

Check out the demo video (embedded below) and you will be able to see it for yourself, just know that it is streamed from peertube, so the video may be a bit choppy depending on your local peers. Finally Yes Mastodon is your cup of tea, here is the link for that. Previous projects have tried to tap into EMG before, like this Bioamp card from Upside Down Labs. We also dug a previous tutorial on the topic on our own. [Bil Herd.]