Engineers at Columbia University in New York City have created a tiny ‘chip as a system’ device powered by ultrasound that can be injected into the body to monitor health.


We spoke to Ken Shepherd, Lau Professor of Electrical Engineering, to find out more.

What’s the background to this research?

Moore’s Law states that you can cram more and more transistors into a certain area on an integrated circuit chip. And that’s number’s been growing exponentially for the last 30 or 40 years. It’s primarily been used not to make the chips smaller, but to put more transistors on a chip that’s the same size. So, we’ve gone from chips with a thousand transistors to chips with tens of billions. But another thing you could do with that density is use it to make chips that are very, very small.

How small are we talking?

So this is the smallest autonomous single chip system that we know of that supports both power and bidirectional communication – it’s roughly 300 x 300 microns [one micron = 0.001mm].

What are the main challenges of producing a chip this tiny?

That a chip needs to be powered and you need to be able to communicate with it, otherwise it’s of no use. So, what we’ve been doing is an example of a device where the chip is the entire system. There’s nothing else; no external sensor array, no external antenna, no external battery, there’s no external anything. And for a chip to operate as an autonomous system, it needs to meet a few criteria.

All of the power and communication to the chip needs to be done wirelessly. So, all the antennae for that wireless powering and communication need to be integrated. And then in the case of these kind of implantables, the chip is also sensing something, so that sensing function also has to be integrated.

You’ll be very challenged to communicate with a device this small with electromagnetics, such as radio waves, because the wavelength is too large relative to the size of the device. Even at tens of gigahertz, you’re talking about wavelengths in the several millimetre range. This device is much less than a millimetre in size, so that’s why we use ultrasound. This device is powered and communicated with acoustics, not electromagnetics, which is useful because sound waves travel very well in the body.

More like this

Read more about health:

How do you power the chip with ultrasound?

We’re looking at using these devices to augment ultrasonography, to provide additional information that’s not intrinsically available. The way ultrasound works is that it sends a sound wave into your body. And when there’s an acoustic mismatch, a difference in the acoustic impedance [the amount of resistance an ultrasound beam encounters as it passes through tissue] due to different materials or interfaces in your body, that reflects some of that acoustic energy back to the imager.

And that’s what you see in an ultrasound image. But there are many things that aren’t available or known to you. For example, this particular chip measures temperature. There’s no way in intrinsic ultrasound imaging that I can know anything about local temperature. So, I put one of these devices in your body and when the ultrasound beam hits it, the energy turns the device on, which then measures the local temperature and modulates the reflected energy back to the ultrasound imager accordingly.

So what you see is this tiny chip in your ultrasound image flashing at you. And that flashing is sending information back to you that tells you what it measured locally.

It does what’s sometimes called energy harvesting – it harvests the energy from the ultrasound beam. And the way it does that is on the chip, we’ve integrated a piezoelectric material that converts sound to electricity. So, when you apply a pressure wave to this material - which is what sound is, a pressure wave, the material gets squeezed a little bit, which generates a voltage and that voltage is used to power the chip.

How deeply can you implant the chip?

We’re using about five-megahertz ultrasound for these devices. Most clinical ultrasound is going a little lower frequency, usually about a megahertz or so. As you go up in frequency, you can penetrate less deeply as the ultrasound is absorbed more in your tissue. But at one megahertz, the wavelengths are too large to communicate with this device. So, at five megahertz, we can go about maybe 6-7cm deep before the attenuation of the ultrasound becomes too great, which is pretty substantial.

How to you place them inside the body?

They’re small enough to fit in an 18-gauge hypodermic needle, so that’s how we put them in. They can also be removed in the same way.

How do they function once they’re in?

There are two ways the devices could be used. One is where it’s chronically implanted – you simply put it in and leave it alone. But much more testing has to be done to understand the long-term consequences of having something like that in your body.

The belief is that being so small will help it to be acceptable and so there’ll be less of a foreign body response. The other way would be that you simply remove it after a period of time. And you can do that using the same kind of hypodermic needle, but guided with ultrasound. You use the ultrasound imager to guide the needle, find the device and then suck it out.

Read more interviews:

What are the potential applications?

This particular design we used to provide additional information to an ultrasound imager. And that can be used in almost any context in which you’re doing ultrasound imaging. For example, there are many clinical applications in which clinicians apply heat. So, if you wanted to know how much heat you’re applying, you could use the chip that measures temperature.

There also might be specific biomarkers that you’re looking for, maybe you’re doing continuous ultrasound imaging over time to verify that a tumour hasn’t come back. But it might make sense to implant devices like this that measure biomarkers to indicate even earlier if there’s a concern. We’re also looking at trying to improve healing by monitoring various biomarkers within a wound.

What are the next steps?

Well, there are lots of other things you could do with these ‘chip as a system’ implants. There’s a lot of interest right now in interfaces to the central nervous system – brain/computer interfaces and devices that interface with the peripheral nervous system for things like pain management, and interactions with the autonomic nervous system to control things like blood pressure.


What these devices are delivering is what we call volumetric efficiency – a statement of how much function you’re able to get out of the implantable device for a given amount of displaced volume. These devices are the most volumetrically efficient devices you can imagine, because you can get the maximum amount of function out of these devices for a minimum amount of displaced volume. And that gives them a lot of advantages.


Jason Goodyer
Jason GoodyerCommissioning editor, BBC Science Focus

Jason is the commissioning editor for BBC Science Focus. He holds an MSc in physics and was named Section Editor of the Year by the British Society of Magazine Editors in 2019. He has been reporting on science and technology for more than a decade. During this time, he's walked the tunnels of the Large Hadron Collider, watched Stephen Hawking deliver his Reith Lecture on Black Holes and reported on everything from simulation universes to dancing cockatoos. He looks after the magazine’s and website’s news sections and makes regular appearances on the Instant Genius Podcast.