Stanford creates tiny, high speed, ultrasound powered brain implants: At any given signal power the only theoretical limit to the amount of information that can be communicated is the noise level. For wireless transmissions to and from biomedical implants the challenges, or rather the problems we face, are essentially power problems. Engineers have pushed RF powering of implants to their practical limits and there’s still left much to be desired. Stanford researchers now suggest that the way to move beyond the legal, physiological, and physical constraints on the energy that can be beamed to an implant may be to move away from electromagnetic waves, and switch to ultrasound instead. Not content just to wax philosophical about these issues, they have now demonstrated proof of principle for ultrasound power transmission into the brain — or at least into a three-centimeter-thick slab of chicken meat. There is not a lot of room in the brain for on-board energy storage in the form of batteries, or through complex bioharvesting provisions. In that case powering an implant is much like sending energy to a beam-powered spacecraft from Earth. If we assume that the base station on Earth can be a giant laser that can focus as much energy onto the spacecraft as it can handle, the main limit becomes the size of the energy absorber. For the spacecraft it is the sail; for RF it is the antenna; and for ultrasound, the vibrational energy absorber (in this case a piezo receiver). For the deceleration phase (and the optional return trip), there will be even less energy available because the laser light will need to somehow be reflected and refocused back to the spacecraft. It is much the same for powering an implant with the exception that the implant needn’t return power, only information. What really puts the kibosh on the above mentioned theoretical limits to power are the down-to-earth limits of the human bodies. The FCC loosely, and we do mean loosely, embodies these flesh and blood (and societal as the case may be) frailties in a few mandatory guidelines: namely that RF intensity be limited to between 1 and 10 mW/cm2, depending on the application. While the effects of RF on living cells is still being explored and debated, nobody wants an FCC snoop truck to classify their brain as a violation. For ultrasound, it is the FDA who steps in and provides for up to 720 mW/cm2 of power. If you will notice, as the Stanford researchers are quick to point out, that’s about two orders of magnitude difference in power available in favor of ultrasound. But what what really kills it for RF is the high attenuation in tissue, and more importantly, the huge mismatch between what we will tentatively call the “aperture,” and the wavelength for millimeter-sized antennas. Using 65nm GP CMOS technology, the Stanford researchers managed to compress all the necessary hardware into a footprint just a couple of millimeters in dimension, with the antenna being the largest component. Using the chicken breast brain proxy, their implant was able to support a DC load of 100 µW. They were also able uplink back out using good old RF at megabits-per-second speeds, which at higher frequencies (here 4GHz) compared to ultrasound has some advantages for data transmission. Not to confuse, but we should mention that their implant could also receive RF data. This is important because the data available from the ultrasound power downlink was only in the kilobits-per-second range. In a real-world scenario, the downlink is likely to be lower bandwidth control signals while the uplink would have more intensive cell activity or image data. While the current prototype may be the size of the head of a ballpoint pen (pictured top), the next-generation implant is expected to be one-tenth that size. At this scale the now seemingly incredible concept of ultrasonically-interrogated neural smart dust is placed more in the realm of the possible. For brain-wide implantation, entirely new physical constraints come into play but other researchers have already laid much of the theoretical groundwork for what we might be able to handle. Here the primary limit becomes the ability of the brain to dissipate heat from the implant, with things like mechano-vibratory disruption and possible mutational mechanisms being slightly lesser concerns. As we have seen RF is still a powerful medium and its use in implants will likely remain with us for the long haul. Advanced antenna-matching techniques and mid-field focusing technology will undoubtedly extend their range of applicability. To that point, Ada Poon and others at Stanford have recently received research grants to produce implants that will address maladies like pain and depression. However as ultrasound hardware components increase in sophistication, and shrink, it will likely become impossible to ignore their advantages for raw power.
Posted on: Fri, 24 Oct 2014 09:53:58 +0000
Trending Topics
Recently Viewed Topics
© 2015