A local network of implants uses your body as the wiring

Most implants like pacemakers and insulin pumps work in isolation. To help them coordinate with each other, a team of Georgia Tech researchers built a networking system that sends signals through body tissue instead of antennas and radio waves.

Radio problems

Implants that communicate today mostly rely on radio protocols like Bluetooth Low Energy or near-field communication (NFC). Both are a poor fit for in-body data transfer, says Alex Abramson, a Georgia Tech engineer and co-author of the new study.

The first problem is power. “If you want an implant to remain in an active state such that it can respond within milliseconds, it’s very difficult to do that with the Bluetooth system,” Abramson said. According to the paper, Bluetooth components, when they’re activated, can cut an implant’s battery life by up to 90 percent.

The second problem is that radio waves don’t travel well through the body. “Bluetooth and near-field communication are attenuated quite a lot in the tissue,” Abramson said. He said that implant-to-implant radio communication systems run into attenuation issues if the signal has to travel more than one centimeter through the tissue. Then there’s size. Radio needs antennas, and commercial Bluetooth components require a device at least five millimeters wide. Implants thinner than three millimeters can be injected with a syringe at an outpatient clinic; bigger ones usually need surgery.

The fix Abramson’s team came up is called SWANS (Smart Wireless Autonomous Networking System) and was inspired by the way body’s own internal communication networks. “The nervous system can take a lot of inputs from all over the body, harvest all that data, and make a specific decision. And our system mimics that,” Abramson said. SWANS relies on ionic conduction just like neurons, which communicate by shuttling sodium and potassium ions through their membranes, creating voltage differences. “But instead of using nerves, we use normal body tissue to send those signals,” Abramson said.

Name calling

The idea isn’t entirely new. A Food and Drug Administration-cleared pill called Abilify MyCite uses ionic conduction to tell a skin patch it was swallowed. But such systems typically link just two devices, while Abramson’s team wanted SWANS to connect many.

The first SWANS component is a wearable hub. It’s a flexible circuit board that reads sensor data, runs decision-making algorithms, and emits voltage pulses of up to 12 volts. The second is a patch of stainless-steel microneedles that delivers those pulses into the body, bypassing the skin’s outermost, poorly conductive layer. The third is a network of syringe-injectable implants, each packing two receiving pads, a transistor switch, a battery, and either a sensor or an actuator such as a nerve stimulator.

“We created all of the smarts in the wearable hub,” Abramson said. The wearable has more room and more battery power, so the implants could be kept small and simple. When the hub fires a pulse, it creates a brief electric field that spreads through the tissue in all directions. Every implant within range picks it up, but only the right one(s) react, a bit like people in a crowded room who turn around only when they hear their own name.

The team achieved that by making each implant’s transistor switch on only when the incoming pulse crosses a specific threshold. Adding a resistor in front of the transistor raises the voltage needed to flip the switch. Adding a capacitor means the pulse must last long enough to charge it first. By mixing and matching these components, the team made implants that respond only to specific combinations of pulse strength and length. This, the authors admit in the paper, means that voltage thresholds need to be tuned for each body and implant placement.

Because SWANS implants are built from passive components, they draw almost no power while listening, extending battery life more than 15 times compared to Bluetooth and NFC. A complete implant with a battery measures 3 by 1.1 by 17 millimeters and fits through a 6-gauge needle.

To test the system, Abramson and his colleagues installed it in chicken breasts, skin-on, bone-in pork bellies, and living rats.

Twitching legs

“Pork belly is a very thick and heterogeneous tissue,” Abramson said. The signal had to get through fat, muscle, bone, and skin. A single 10-volt pulse produced a detectable voltage gradient more than 30 centimeters across the tissue and up to 14 centimeters deep—more than 10 times the coverage of Bluetooth or NFC.

In live rats, the signals reached implants under the skin, in the abdominal cavity, and even in the stomach. Placement didn’t matter much either. “We put our wearable on the stomach of the rat, and then we were able to get actuation in the back of the rat,” Abramson said.

The most advanced experiment used strain sensors placed on the rats’ front legs. When a sensor detected the left or right forelimb moving, the hub sent a limb-specific pulse through the body that reached an implant on the corresponding hind leg, which stimulated the sciatic nerve and made that leg twitch. The team also built an implant-to-implant relay where a device with a temperature sensor passed a signal to a second implant only when it registered a fever above 40° C (104° F).

Abramson notes that pacemakers and neurostimulators already use similar voltages and pulse lengths, so the team did not expect any major safety issues. In a two-month study in rats, scar tissue up to a millimeter thick grew around the implants, but by raising the voltage within safe limits, the team kept communication going throughout. The pulses didn’t stimulate nerves other than the targeted one, didn’t change the heart’s electrical activity, and caused no more cell death or oxidative stress than needles and implants that weren’t electrified.

But SWANS is not a do-it-all in-body communication system, and it likely never will be.

Implant multiplexing

The first limitation of SWANS is that it can’t send much data. “We don’t want to send large amounts of data using our devices because we want to create ultra-small, ultra-low-power systems,” Abramson said. SWANS, he argues, is meant to pass key information (like a temperature reading) between body parts, but not much else. The system also hasn’t been tested in large animals or humans yet, though Abramson says preliminary large-animal studies have looked promising. And he already thinks about potential applications.

“Right now, neurostimulation and drug delivery are completely separate,” Abramson said. Today, he argues, a person cannot have a drug delivery pump connected with their neurostimulator to allow for combined therapies, even though there are benefits to taking a drug at a specific time in conditions like epilepsy.

The strength of SWANS, Abramson argues, lies in how easily it can be integrated with existing implants. “This new communication protocol can be plugged into any of those previous systems,” he said. “Because it’s completely agnostic to the sensor or the actuator that we’re using, we’d be able to create this full network of in-body therapeutics.”

Science, 2026. DOI: 10.1126/science.adz5300

Jacek Krywko is a freelance science and technology writer who covers space exploration, artificial intelligence research, computer science, and all sorts of engineering wizardry.