High Density Electromyography (HDEMG) Sensor

This project designs and builds a 64-electrode surface array with 2mm spacing to capture motor neuron activity from muscle tissue, enabling high-resolution sEMG data collection for applications in prosthetics and neural interface research.I designed the electrode array layout and built the accompanying headstage, integrating bio-signal amplifiers to condition and read out muscle activity signals from the array.

The project is part of a larger effort to develop and validate a system for rapidly and comfortably diagnosing disorders of motor unit function in facial muscles:

  1. Identify nature of disorder (degeneration, hyperactivity)
  2. Localize anatomical site of disorder
  3. Monitor changes after treatment

My work revolved around designing and testing the hardware for the same. I developed the Sensor array and myo headstage in the system diagram below.

System Diagram

I designed an 8x8 electrode array with 2mm spacing with holes in between to allow for simulatenous use of under the skin electrodes.

Left depicts Kicad design for array and right depicts manufactured product

I also designed a 64 pin headstage that would interface with the HDEMG tile to allow for data readout.

Left depicts Kicad design for headstage and right depicts manufactured product

Then we proceeded with testing. The mean channel RMS was 6.36 uV with an standard deviation of around 1.11 uV across the RMS channel values. The lowest SNR was around 20db, prooving high signal quality.

Signal to Noise Ratio histogram across all channels

I was the first subject on the device and it showed smooth rainbow waterfall trends when I was smiling - successfully picking ip motor neuron activity.

Left depicts waterfall graph for open smile and Right depicts subject 0 smiling (me)

Further more the device allows for discrimination of motor units (MU) on the face with clear signal propogation accross channels below on a 20 uV scale

Left MU 1,Right MU3

Orderly recruitment of distinct motor units prooved to be observable as well.

Spike train graph across 80 seconds and 5 motor units (for two smile iterations)

Innervation sites identifiable using channels with greatest peak-2-peak amplitudes with motor unit action potentials.

Innervation sites heatplot motor units

However Spread of MU innervation sites should become more distinct as electrode spacing increases. Current work includes building a 4x4 mm spacing tile for the same.