The first in vivo tests demonstrate ultrasound can be used to wirelessly power and communicate with millimetre-scale devices surgically placed in muscles and nerves.
Therapeutic modulation of the activity of the body’s peripheral nervous system (PNS) holds a world of potential for mitigating and treating disease and other health conditions—if researchers can figure out a feasible long-term mechanism for communicating with the nerves and pathways that make up the body’s information superhighway between the spinal cord and other organs.
What does feasible look like? Small is the best start—small enough to someday perhaps be injected or ingested—but also precise, wireless, stable, and comfortable for the user. Modern electrode-based recording technologies feature some, but not all of these qualities. Hardwired solutions present challenges for chronic use, while existing wireless solutions cannot be adequately scaled down to the sizes needed to record activity from small-diameter nerves and record independently from many discrete sites within a nerve bundle.
DARPA’s Electrical Prescriptions (ElectRx) program is focused in part on overcoming these constraints and delivering interface technologies that are suitable for chronic use for biosensing and neuromodulation of peripheral nerve targets.
Now, a DARPA-funded research team led by the University of California, Berkeley’s Department of Electrical Engineering and Computer Sciences has developed a safe, millimetre-scale wireless device small enough to be implanted in individual nerves, capable of detecting electrical activity of nerves and muscles deep within the body, and that uses ultrasound for power coupling and communication.
They call these devices ‘neural dust’. The team completed the first in vivo tests of this technology in rodents. Neural dust represents a radical departure from the traditional approach of using radio waves for wireless communication with implanted devices, said Doug Weber, the DARPA program manager for ElectRx.
The soft tissues of our body consist mostly of saltwater. Sound waves pass freely through these tissues and can be focused with pinpoint accuracy at nerve targets deep inside our body, while radio waves cannot. Indeed, this is why sonar is used to image objects in the ocean, while radar is used to detect objects in the air.
By using ultrasound to communicate with the neural dust, the sensors can be made smaller and placed deeper inside the body, by needle injection or other non-surgical approaches.
The prototype neural dust ‘motes’ currently measure 0.8mm x 3mm x 1mm as assembled with commercially available components. The researchers estimate that by using custom parts and processes, they could manufacture individual motes of 1 cubic millimetre or less in size—possibly as small as 100 microns per side.
The small size means multiple sensors could be placed near each other to make more precise recordings of nerve activity from many sites within a nerve or group of nerves.
Though their miniscule size is an achievement in itself, the dust motes are as impressive for the elegant simplicity of their engineering.
Each sensor consists of only three main parts: a pair of electrodes to measure nerve signals, a custom transistor to amplify the signal, and a piezoelectric crystal that serves the dual purpose of converting the mechanical power of externally generated ultrasound waves into electrical power and communicating the recorded nerve activity.
The neural dust system also includes an external transceiver board that uses ultrasound to power and communicate with the motes by emitting pulses of ultrasonic energy and listening for reflected pulses.
During testing, the transceiver board was positioned approximately 9mm away from the implant.
The piezoelectric crystal is key to the design of neural dust. Pulses of ultrasonic energy emitted by the external board affect the crystal. While some of the pulses are reflected back to the board, others cause the crystal to vibrate.
This vibration converts the mechanical power of the ultrasound wave into electrical power, which is supplied to the dust mote’s transistor.
Meanwhile, any extracellular voltage change across the mote’s two recording electrodes—generated by nerve activity modulates the transistor’s gate, which changes the current flowing between the terminals of the crystal.
These changes in current alter the vibration of the crystal and the intensity of its reflected ultrasonic energy. In this way, the shape of the reflected ultrasonic pulses encodes the electrophysiological voltage signal recorded by the implanted electrodes.
This signal can be reconstructed externally by electronics attached to the transceiver board to interpret nerve activity.
One of the most appealing features of the neural dust sensors is that they are completely passive. Because there are no batteries to be changed, there is no need for further surgeries after the initial implant, Weber said.
Another benefit of the system is that ultrasound is safe in the human body; ultrasound technologies have long been used for diagnostic and therapeutic purposes.
Most existing wireless PNS sensors use electromagnetic energy in the form of radio waves for coupling and communication, but these systems become inefficient for sensors smaller than 5mm.
To work at smaller scales, these systems must increase their energy output, and much of that energy gets absorbed by surrounding tissue. Ultrasound has the advantage of penetrating deeper into tissue at lower power levels, reducing the risk of adverse effects while yielding excellent spatial resolution.
This proof of concept was developed under the first phase of the ElectRx program.
The research team will continue to work on further miniaturising the sensors, ensuring biocompatibility, increasing the portability of the transceiver board, and achieving clarity in signals processing when multiple sensors are placed near each other.