Funding programme/Agency: ERC

Acronym: Natural BionicS

Institution: Imperial College London

Project type: ERC-SyG – Synergy grant

Implementation period: 01.06.2019 – 30.11.2025

Budget: € 9,984,021.25

Project website: https://cordis.europa.eu/project/id/810346

PROJECT SUMMARY:

The loss of a limb results in profound impairments in movement and interaction with the environment, severely diminishing an individual’s quality of life. Beyond the physical limitations, amputation leads to functional “neural deafness” in substantial areas of the sensory-motor cortex. Robotic limbs offer a promising substitute for missing or damaged limbs, especially when they are connected to the human body through advanced neural interfacing technologies.

The NaturalBionicS project is focused on developing biological connectors, integrated into a compact bio-hub, to access the spinal cord circuitry in amputee patients. This enables the control and sensing of robotic limbs in a way that closely mimics natural limb function. The core strategy revolves around harnessing biological pathways that naturally encode and decode neural signals, leveraging residual biological structures present post-amputation. These existing structures provide a more effective interface with the patient’s nervous system.

Substantial progress has been made toward the project’s ambitious goals, with significant advances across neurosurgery, neural interfacing, and robotics, as well as in the integration of these disciplines. Preclinical surgical research has successfully demonstrated that multiple nerves, once responsible for controlling the missing limb, can be transferred to muscle tissue located above the site of amputation. This muscle, once reinnervated, becomes electrically active and effectively serves as a biological “screen” that reflects the neural commands once directed to the lost limb. Unlike previous methods that distribute nerve transfers across multiple muscles, this approach has shown that a single muscle can be hyper-reinnervated to create a compact neural representation of the missing limb’s motor functions.

These biological structures have been interfaced with novel implantable electrode arrays, establishing a bidirectional communication channel with the nervous system. Using these interfaces, it has been possible to record neural activity from the reinnervated muscles and decode it to accurately infer the patient’s motor intentions. Imperial College has been responsible for the development of the decoding algorithms used in this process, which rely on machine learning and artificial intelligence techniques. These algorithms have made it possible to disentangle the activity of individual nerves from the signals recorded in the hyper-reinnervated muscle, successfully validating one of the most high-risk scientific hypotheses of the project.