Georgia Tech Team Builds Implants That Network Through Tissue

A Georgia Tech engineering team has built a system that lets medical implants exchange data through the body's own tissue rather than radio waves, according to a study described by researchers at the university and reported by Ars Technica. The system, called SWANS (Smart Wireless Autonomous Networking System), is designed to let multiple small implants — sensors, nerve stimulators, drug-delivery devices — coordinate with each other the way neurons do, using electrical signals carried through conductive tissue instead of Bluetooth or near-field communication.

What problem is SWANS trying to fix?

Most implants today, such as pacemakers and insulin pumps, work alone. When they do communicate, they typically rely on Bluetooth Low Energy or near-field communication. Alex Abramson, a Georgia Tech engineer and co-author of the study, told Ars Technica that both protocols create three practical failures inside the body. First, power: keeping a Bluetooth radio active enough to respond within milliseconds can cut an implant's battery life by up to 90 percent, according to the paper. Second, range: radio signals attenuate sharply in tissue, and implant-to-implant radio links run into trouble once a signal has to travel more than one centimeter through the body. Third, size: commercial Bluetooth components need an antenna at least five millimeters wide, which pushes a device past the threshold for outpatient injection. Implants under three millimeters can be placed with a syringe in a clinic; anything larger typically requires surgery.

How does SWANS send signals through the body?

SWANS swaps radio for ionic conduction — the same basic mechanism neurons use to pass electrical signals by shuttling sodium and potassium ions across their membranes. "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, according to Ars Technica. "But instead of using nerves, we use normal body tissue to send those signals."

The architecture has three parts. A wearable hub — a flexible circuit board worn outside the body — reads sensor data, runs the decision-making software, and emits voltage pulses of up to 12 volts. A patch of stainless-steel microneedles delivers those pulses into the body, piercing the skin's outer layer, which otherwise conducts electricity poorly. From there, a network of syringe-injectable implants receives the signals through tissue; each implant carries 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, noting the external device has more room and more battery capacity than anything implanted.

Has anyone tried sending signals through tissue before?

The underlying idea is not new. Abilify MyCite, a Food and Drug Administration-cleared pill, already uses ionic conduction to signal a skin patch when it has been swallowed. What Abramson's team says is different is scale: Abilify MyCite links exactly two devices, while SWANS is built to connect many implants into a single coordinated network, with the wearable hub acting as a central processor for all of them.

Why build a hub-and-spoke model instead of direct implant-to-implant links?

Because implants have the least room and the least battery capacity of any component in the system, the Georgia Tech design concentrates computing power and power draw in the external wearable rather than the implants themselves. That division of labor is also what lets individual implants stay small enough for syringe injection rather than surgical placement, addressing the size problem Abramson identified with Bluetooth-based designs.

SWANS by the numbers

• Bluetooth activation can cut implant battery life by up to 90 percent, per the study.
• Implant-to-implant radio links degrade once signals travel more than 1 centimeter through tissue.
• The SWANS wearable hub emits voltage pulses of up to 12 volts through microneedle patches.
• Implants under 3 millimeters can be syringe-injected in a clinic; larger devices generally need surgery.

"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." — Alex Abramson, Georgia Tech

What to watch

• Whether SWANS moves from bench testing toward trials in living subjects, and what safety data that would require.
• How regulators, including the FDA, might evaluate a multi-device ionic network versus the two-device precedent set by Abilify MyCite.
• Whether the 12-volt pulse levels and microneedle delivery method hold up for long-term, chronic implant use outside a lab setting.
• Whether device makers building pacemakers, nerve stimulators, or drug pumps show interest in licensing a tissue-based signaling approach over existing Bluetooth designs.

The full study details, including any planned follow-on testing, are described in the original Ars Technica report.