A 3D-Printed Silicone Implant Cools the Brain and Records It at the Same Time

Cooling a small patch of brain tissue has long been floated as a way to quiet seizures that drugs cannot control. Building a device that can do it, small enough to sit on the cortex and able to record what the brain is doing at the same time, has been the hard part. A team from the Leibniz-Institut für Polymerforschung Dresden (IPF), TU Dresden and the University of Sheffield now reports one answer, made largely by extruding silicone. In a paper in Advanced Science, Moore, King, Da Silva and colleagues describe a "printing-assisted" design framework. It combines extrusion-based 3D printing of silicones with off-the-shelf electronic components to make soft implants for the cortical surface of rodents. In a rat model of epilepsy, the implant's localized cooling suppressed seizure-like activity.

The work was covered on October 29 by VoxelMatters and also by 3D Printing Industry. It is an early animal study, not a clinical device. For anyone who follows functional printing, though, the interesting part is how the device was put together: one printed soft package holding thermal, fluidic and electrical parts.

Three Layers in a Fingernail-Sized Package

According to 3D Printing Industry's account, the finished implant measures 10 x 10 x 3 mm. It was made by direct ink writing (DIW) of Dow DOWSIL SE1700 silicone on a RegenHU 3D Discovery printer. DIW is a paste-extrusion process: a viscous ink is pushed through a nozzle and then cures in place, rather than being melted and solidified like FDM filament. That makes it a good fit for soft elastomers that have to conform to tissue.

The device is built from three stacked modules:

Sensing: four electrodes made from a graphite-silicone composite, which record electrocorticography (ECoG) signals from the brain's surface.
Cooling: a miniature thermoelectric cooler, a solid-state Peltier-type element that pumps heat away from the tissue it touches.
Heat removal: a water channel that carries away the heat the thermoelectric element moves, so the hot side doesn't simply warm the brain from behind.

The paper's abstract stresses the integration. Printed silicone forms the soft housing and the conductive composite, and discrete electronics are embedded in it. An "autonomous auxiliary driver unit" runs the system. The external driver, 3D Printing Industry reports, measures 17 x 8 x 5 cm, so the implant itself is tiny but it is still tethered to a box about the size of a thick paperback.

Bench Numbers Before Brains

Before any animal work, the team tested the implant in agar held at 37°C, a standard stand-in for tissue. In those tests the system held its target temperature within ±0.25°C. It brought the contact surface down to 23°C in about five seconds, and the cooling reached 1 to 3 mm into the material. At steady state it drew 2.6 W.

Thermal management works in both directions here. A thermoelectric cooler makes heat as well as moving it, and the researchers note that surrounding tissue should not warm by more than 1°C. The water channel exists to meet that limit. The team also points out a limitation of its own test setup: agar tends to overstate how far the cooling spreads. Living tissue, with blood flowing through it, carries heat back in more effectively, so the 1–3 mm figure should be read as an upper bound rather than a promise.

VoxelMatters reports that the design was also checked with finite element simulations alongside the agar models. The printed composite electrodes produced ECoG recordings of quality comparable to a commercial system. That matters because a cooling implant that also records lets researchers see whether the cooling is working, without needing a second device.

What Happened in the Rats

The in vivo work used nine anesthetized rats and the 4-aminopyridine model, a chemical method for inducing seizure-like activity in the cortex. The researchers delivered 60-second cooling events at target temperatures from 25°C down to 15°C and recorded through the implant's own electrodes as they did.

Cooling reduced signal energy, a measure of how intense the abnormal electrical activity was, at every temperature tested. One result runs against intuition: the coldest setting, 15°C, was slightly less effective than milder cooling. Colder is not automatically better, though the researchers say the finding needs further study because it differs from earlier studies and rests on few trials.

The abstract states the authors' ambition plainly. The approach "may advance the clinical translation of focal hypothermia" as an alternative therapy for drug-resistant focal epilepsies. This is a step along that road, not the end of it. Anesthetized rats, short cooling bouts and a chemically induced seizure model are a long way from a chronic implant in a human patient.

What It Means for Makers

Few readers will be printing neural implants. But the study is a clear example of a trend that has been building in multi-material printing: using additive processes to make finished functional assemblies, not just enclosures. Several lessons carry over to other work.

Paste extrusion is a serious tool. The whole build rests on DIW of a commercial silicone. That is the same family of process that hobbyists use with syringe-based paste extruders for silicone, clay and conductive inks. The RegenHU machine is a lab-grade printer, but the underlying idea, extruding a curable elastomer layer by layer, is not exotic.

Conductive composites can replace metal traces. Mixing graphite into silicone gave the team electrodes that bend with the package and match it mechanically. Anyone who has fought with rigid wires breaking out of a soft cast or print will see the appeal of making the conductor from the matrix itself.

Hybrid builds beat all-printed ones. The researchers did not try to print a thermoelectric cooler. They printed around an off-the-shelf one and added a printed fluid channel to deal with its waste heat. That pattern of buying the hard part and printing the soft part, the packaging and the plumbing, is the most practical form of "functional printing" today, whether the product is a wearable sensor or a liquid-cooled electronics enclosure.

Validate thermals honestly. The team's own note that agar overstates heat spread is a useful reminder. Bench proxies flatter designs, and anyone designing printed heat-management parts should assume the real environment will be less forgiving than the test rig.

Bottom Line

The work was funded by an ERC Starting Grant called IntegraBrain, according to VoxelMatters. It shows that a soft, printed package can combine sensing, active cooling and liquid heat removal in a 10 mm square and still do something useful in living tissue. The clinical road is long and the external driver is still bulky. As a demonstration of what extruded silicone and embedded components can do together, it is a convincing functional-printing result.