Combined methods could improve neuroprosthetics

A severe spinal cord injury or a stroke can cause a loss of sensation in the arms and legs. Over the past decades, advances in neuroprosthetics are enabling a return of sensation as well as movement to affected limbs. These systems work by using artificial sensors to that capture information from the environment that are then encoded to signals that can stimulate the nervous system via neural interfaces. However, as the complexity of such sensory neuroprosthetics increases, finding the best stimulation parameters to provide natural sensation becomes more complex.

Leong et al. present a perspective for optimizing three frameworks for sensory encoding in neuroprosthetics with a focus on the somatosensory system. The three frameworks — explicit, physiological, and self-optimized — are described independently and each method’s strengths and limitations are presented. While explicit frameworks, which rely on psychophysical tests or questionnaires, have been the gold standard, they are much slower than the physiological approach, which uses computer algorithms to refine parameters..

“Sensory feedback is a complex task because it is highly subjective,” said author Solaiman Shokur. “We think people might look at our review and recognize the approach that they’re already using and then be inspired by some other alternative ones.”

Additionally, combined approaches are discussed. Integrating methods could improve robustness of optimization, Shokur said. For example, the physiological framework could be used to establish a baseline parameter space, and an explicit or self-optimizing framework could be subsequently used to refine the parameters. This would allow for greater precision and fine tuning than could be done with physiological approaches alone.

“Hopefully, this idea of combining several of these techniques could even open up new avenues for restoration of sensory feedback,” Shokur said.

Source: “Optimization Frameworks for Bespoke Sensory Encoding in Neuroprosthetics,” by Franklin Leong, Silvestro Micera and Solaiman Shokur, APL Bioengineering (2025). The article can be accessed at https://doi.org/10.1063/5.0249434 .

This paper is part of the Bioengineering of the Brain Collection, learn more here .