The story of brain–machine interfaces begins with a bold idea: if scientists could listen to individual neurons, they might one day help people move again, communicate, or understand how thoughts arise. That idea has powered neuroscience for decades and has already made remarkable things possible, from paralyzed patients controlling robotic arms to detailed maps of how neurons fire in real time.
Yet their achievements carried a heavy compromise. Once inside the soft, shifting tissue of the brain, their rigid structures often stirred up inflammation, pushed neurons away, and slowly lost their ability to record.
This is the long-standing impasse that Dr. Fei He's team at the Shanghai Institute of Optics and Fine Mechanics, Chinese Academy of Sciences, aims to overcome. In International Journal of Extreme Manufacturing , they propose that, instead of working around the limitations of rigid electrodes, the entire design and manufacturing process must be pushed to its limits to produce implants that are fundamentally different from anything used today.
The authors begin by stepping back and asking what an ideal intracortical electrode should be able to do. It must record from many neurons at once without damaging them. It must be small enough to blend into the tissue, yet strong enough to be inserted without buckling. It must maintain a stable electrical connection for years, and ideally work hand-in-hand with other sensing methods, such as light-based or chemical probes.
When the team mapped out these requirements, the picture that emerged was neither a needle nor a chip, but something much closer to a soft thread with the thickness of a tiny blood vessel and the flexibility to move with the brain's own rhythms.
But making something so small and soft creates new difficulties. A probe that bends easily once inside the brain is too floppy to insert. A device the width of a capillary is fragile to handle. And shrinking electrodes means less room for the metal wires, sensors and electronics needed to carry signals out of the brain. Improvements in one area often create problems in another.
Dr. He and colleagues tackle these trade-offs by laying out a clear "design map" for intracortical electrodes during his tenure as a research fellow at UT Austin and Rice University (Drs. Chong Xie and Lan Luan's Group) starting from 2016. They show how mechanical flexibility, electrical stability, spatial resolution and compatibility with optical or chemical tools all intersect. Within this map, the most brain-friendly devices are those with micrometer-scale dimensions and very low stiffness, but these require clever solutions for insertion and robust materials to survive fabrication.
The review then walks readers through the manufacturing approaches that are beginning to meet these demands. Advanced silicon micromachining can carve ultra-sharp, slender tips. Ultrathin polymer and metal films can form soft, flexible bodies. Thermal fiber drawing can combine electrodes, optical waveguides and fluidic channels into a single multifunctional thread. Emerging 3D microprinting and laser-based methods can sculpt complex shapes with remarkable precision. Each method sets limits on what is possible: which materials can be used, how small features can be made, and how many functions can be packed into one device.
Researchers are also exploring clever ways to help implants change from stiff to soft. Some devices are delivered on temporary shuttles that dissolve once they reach their target. Others use coatings that are firm at room temperature but soften in the warmth of the body. Materials that change shape with heat and fluid-guided insertion tools are also being tested. The authors argue that controlling these transitions, rather than simply shrinking devices, is the heart of extreme manufacturing.
The review also reminds us that electrodes cannot be separated from the electronics that power them. Modern brain implants depend on amplifiers, connectors and protective packaging that must handle thousands of channels with low noise. Designing these parts in isolation is no longer workable. They must be engineered together as one system, from fabrication to surgical implantation.
Still, major hurdles remain. Manufacturing sub-10-micrometre structures over entire wafers is technically demanding. Testing methods that accurately mimic the softness, chemistry and motion of real brain tissue are not yet standardized. And the field lacks shared design rules that link the shape and materials of implants to the brain's long-term biological responses.
The authors believe the next chapter will depend on forging closer ties among fields that rarely work in lockstep. They imagine shared platforms where engineers test new fabrication ideas, neuroscientists evaluate them in vivo, surgeons refine deployment methods and regulatory experts help chart a safe path to clinical use. Together, these cycles of design and feedback could speed the arrival of implants that are not only technically impressive but also ready for real-world challenges.
"As extreme manufacturing concepts mature, we expect intracortical electrodes to evolve from rigid, planar arrays into three-dimensional, tissue-conformal neural fabrics that can be delivered through minimally invasive procedures," says Prof. He.
Such implants could provide stable, high-density access to the brain for therapies, cognitive prostheses and large-scale neural mapping. More importantly, they could narrow a gap that has persisted for decades, a gap between what engineers can build and what the brain can safely accept.
International Journal of Extreme Manufacturing (IJEM, IF: 21.3) is dedicated to publishing the best research related to the science and technology of manufacturing functional devices and systems with extreme dimensions (extremely large or small) and/or extreme functionalities
- Maintained #1 in Engineering, Manufacturing for consecutive years
- Average time to First Decision after Peer Review: 34 days
- Open Access Publishing after Double-Anonymous Peer Review
Visit our webpage , like us on Facebook , and follow us on Twitter and LinkedIn .
