Until now, prosthetics have mostly been controlled by muscle signals. How does the new motor neuron-based approach differ from the traditional EMG-controlled prosthesis?

Current prosthetic systems are mainly controlled using electromyographic (EMG) signals, which measure the electrical activity generated by muscles. While this approach has enabled major advances in prosthetic control, these signals only reflect the summed activity of many muscle fibres. As a result, the underlying neural commands that actually drive movement can only be inferred indirectly and with limited precision.

The new motor neuron-based approach goes one step further by accessing the neural information closer to its biological source. Instead of relying solely on global muscle activity, it aims to decode the activity of individual motor neurons, which are the cells directly responsible for transmitting movement commands from the nervous system to the muscles. This provides a much more detailed and physiologically meaningful representation of the user’s movement intention.

In practice, this can enable prosthetic control that is more intuitive, precise, and responsive, because the system is based more directly on the natural neural strategies used by the human body to generate movement.

What are the benefits of detecting neural signals?

Motor neurons are nerve cells in the spinal cord that transmit signals from the brain to the muscles, triggering muscle contractions. Their activity, therefore, reflects the intention to move much more directly than the standard EMG signal and thus provides finer and more nuanced information about the intended movement. If we can therefore read the activity of the motor neurons directly, future prostheses will likely be controlled more directly and thus more precisely. From a scientific perspective, motor neuron signals in our trials with healthy individuals exhibit more distinct activation patterns. This makes it clear that there is a future opportunity to control prostheses more precisely and individually, and to make grasping and moving with prostheses more natural and intuitive. 

How can the activity of a motor neuron be measured?

To do this, we use high-density surface EMG (HD-sEMG), a commercially available method for the non-invasive measurement of muscle activity using high-resolution electrode arrays. The high spatial resolution enables a significantly more accurate recording of muscle activity. It is also necessary to decompose signals into the activity of individual motor neurons and to reveal the actual neural control information.

For us, this means in concrete terms:

•    better separation of simultaneously active muscle components, 

•    access to neural control strategies rather than just muscle output, 

•    and thus the basis for significantly more powerful and intuitive control algorithms. 

What are the next steps?

So far, we have conducted our research with healthy volunteers and only one subject with limb loss. We will now extend our research to a series of trials with people who rely on a prosthesis in their daily lives. We will also continue working to ensure the prosthesis can be controlled in real-time. 

Publication

Muscle and Motoneuron Synergies in Biomimetic and Non-Biomimetic 3-DoF Tasks; Johanna Happold, Laura Ferrante, Patricia Capsi-Morales, Deren Y. Barsakdioglu, Dario Farina, Cristina Piazza; IEEE Transactions on Neural Systems and Rehabilitation Engineering, 3-2026; https://ieeexplore.ieee.org/document/11457049

Text: Andreas Schmitz