Giacomo Valle, Assistant Professor at Chalmers University of Technology, has received a prestigious starting grant from the European Research Council, ERC. In the Bionic-Dexterity project, he aims to give people with paralysis the ability not only to control a bionic arm and hand with their mind – but also to feel where it is and how it is moving.
Human manual dexterity is exemplified by virtuoso pianism or the delicate assembly of a clockwork mechanism. It is the product of thousands of years of evolution and one of the defining features of human capability. It relies on fine motor control, high-resolution touch, and our often-neglected sixth sense, proprioception, which allows us to sense where our body is in space and the force we exert.
Imagine picking up an egg without looking at your hand. To avoid dropping or crushing it, your brain continuously receives information about where your arm and fingers are, how they are moving, what they are touching and how much force they are applying. Much of this happens without us being aware of it. One of the senses that makes it possible is proprioception.
For people with arm and hand paralysis following a spinal cord injury, recent advances in implantable brain-computer interfaces, iBCIs, have opened new possibilities for restoring lost functions. Brain signals can be used to control a bionic limb, while electrical microstimulation of specific areas in the brain can restore some sensations of touch.
But one crucial piece is still missing: the sense of proprioception.
“Today’s brain-controlled bionic hands can reproduce some movements, but they are still missing a critical component: a natural sense of the hand. A healthy hand continuously tells the brain where the fingers are, how they are moving, how much force is being applied and what is being touched,” says Giacomo Valle, Assistant Professor at the Department of Electrical Engineering at Chalmers Technical University.
Giving the brain a sense of the bionic hand
In the Bionic-Dexterity project, Giacomo Valle and his research team will develop a new approach aimed at restoring dexterous arm and hand control in people with spinal cord injury.
The central idea is to create a two-way communication between the brain and a bionic limb. Signals from the brain will be decoded to control the bionic arm and hand. At the same time, electrical microstimulation of specific areas of the somatosensory cortex will provide rich sensory information back to the brain.
A particular focus will be on restoring proprioception – something current neuroprosthetic technologies have not yet succeeded in doing.
“When we use our natural hand, the brain constantly combines information about touch, movement and force to guide every action. My central hypothesis is that restoring this sensory information, particularly proprioception, will fundamentally improve how naturally and precisely a person can control a bionic limb.”
To achieve this, the researchers will first develop computational models of how the nervous system processes touch and proprioception. These models will then be used to design stimulation strategies that recreate sensory feedback in the brain. The ultimate goal is to integrate sensory feedback, brain signals and robotic control into a fully biomimetic, implantable, bidirectional brain-computer interface in people with paralysis.
Picking up a glass of water – without looking
Why is sensory feedback so important? Giacomo Valle uses a simple example: imagine trying to pick up a glass of water without being able to see your hand – and without being able to feel your fingers because of a cold day or know where your hand is.
You might still be able to perform the movement, but it would become considerably harder to control. You would have to rely almost entirely on vision to judge where your fingers are, whether you have made contact, and how hard you are gripping.
“We normally use sensory feedback continuously to adjust our movements in real time. Without the right feedback, movements become slower and less accurate, and it becomes difficult to regulate grip force. You might grip too weakly and drop the glass, or too strongly and crush or spill what you are holding.”
If Bionic-Dexterity succeeds, the difference could become apparent in something as seemingly simple as picking up an egg.
“A future user could potentially feel where their fingers are, sense when they make contact with the glass, and adjust their grip naturally as the glass becomes heavier or lighter. More broadly, the goal is to enable complex everyday activities without having to constantly watch the bionic hand.”
From models of the brain to a functioning neural prosthesis
One of the project's major challenges will be translating the complex sensory information of the human body into electrical microstimulation patterns that the brain can interpret naturally and reliably.
This requires bringing together several research fields, from neuroscience and computational modelling to implantable neural interfaces and robotics.
“We have made substantial progress in decoding movement from the brain and in providing artificial touch, but we still do not fully understand how to recreate the rich, multimodal sensory information that makes natural dexterity possible.”
The researchers will need to bridge several levels of research: from computational models of sensory processing and implantation in specific areas of the brain, to neural stimulation and decoding and, ultimately, real-time robotic control.
For Giacomo Valle, this is also what makes the project particularly exciting.
“Success would represent not just an improvement in a bionic device, but a step towards understanding and recreating the neural principles that make human dexterity possible.”
ERC grant makes an ambitious research program possible
The European Research Council's grants are among Europe's most competitive research grants and provide researchers with the opportunity to pursue ambitious research at the scientific frontier.
For Giacomo Valle, the grant means that the different pieces needed to address the problem can be brought together within one long-term research program.
“Receiving an ERC grant is a tremendous recognition of the scientific vision behind Bionic-Dexterity. It gives me the freedom to pursue a high-risk, ambitious research program over the long term.”
The project will combine neuroscience, computational modelling, neural interfaces and robotics, covering the entire chain from understanding how the brain encodes touch and proprioception to translating that knowledge into neural stimulation and, ultimately, a functional bidirectional neuroprosthesis in patients.
“Most importantly, it gives us the opportunity to tackle the particularly challenging problem of restoring proprioception rather than focusing only on movement for brain-computer interfaces.
The long-term ambition is both scientific and clinical: to gain a deeper understanding of the mechanisms behind human dexterity while opening new possibilities for restoring arm and hand function in people with spinal cord injury.”
About Bionic-Dexterity
Bionic symbiosis: restoring manual dexterity after paralysis via a multimodal cortical neuroprosthesis) aims to develop a bidirectional implantable brain-computer interface (iBCI) to restore naturalistic sensorimotor function of the arm and hand in people with paralysis following spinal cord injury.
The project combines neuroscience, computational modelling, neural interfaces, and robotics to establish a direct link between the human brain and a bionic arm and hand. A central objective is to restore naturalistic somatosensory feedback, including touch and proprioception (the sense of limb position, movement, and interaction forces) through electrical microstimulation of the somatosensory cortex. By integrating this artificial sensory feedback with high-dimensional brain control of the robotic limb, BIONIC DEXTERITY aims to enable intuitive and dexterous manipulation, ultimately allowing users to perform everyday activities.
Principal investigator: Giacomo Valle, Assistant Professor, Biomedical Signals and Systems, Department of Electrical Engineering, Chalmers University of Technology
Amount: 2.5 million Euro
ERC press release
For more information, contact:
- Assistant Professor, Signal Processing and Biomedical Engineering, Electrical Engineering