Neural implant merges photovoltaics with custom analog, PPM encoding

It seems as if every technical advance these days is either directly related to AI software and data centers, or at least tries to establish such a connection, even if that connection is somewhat of a tenuous “stretch.” Despite this, there are a lot of innovative and interesting projects underway that are very analog-centric, with little or no AI association. These advances show what “small” analog can do, where small refers to both physical size and focused functionality.

Consider a neural implant dubbed the Microscale Optoelectronic Tetherless Electrode, or MOTE, developed at Cornell University (Figure 1). Measuring about 300 microns long and 70 microns wide (yes, that’s microns), researchers maintain it’s the smallest neural implant capable of wirelessly transmitting brain activity data.

Figure 1 This brain-implantable MOTE measures just 300 microns long and 70 microns wide and requires no tether or wireless RF link for power or data. Source: Cornell University

It’s connected via red and infrared laser beams that pass harmlessly through brain tissue. The MOTE transmits data back using tiny pulses of infrared light, which encode the brain’s electrical signals. An aluminum gallium arsenide (AlGaAs) semiconductor diode both captures light energy to power the circuit and emits light to communicate the data.

The device also includes a low-noise amplifier and optical encoder, all built using the standard CMOS process technology. The optical link uses pulse position modulation (PPM) for its data encoding as that format is very power efficient, especially in this situation (Figure 2).

Figure 2 System overview shows a MOTE implanted in an awake mouse brain to chronically record neural activity in vivo—incoming light powers the MOTE, and the MOTE, in turn, emits the PPM pulses communicating the recorded data (a). Optical microscopy image compares a MOTE with a strand of human hair (b). MOTE is powered and is communicating optically; it’s continuously powered at a shorter wavelength and communicates at a longer wavelength, making the powering system easier to implement and avoiding power–communication crosstalk (c). Source: Cornell University

The dual-use diode, dubbed a photovoltaic light-emitting diode (PVLED), provides space-saving benefits, functioning as both an LED and a data-link transmitter. An external 623-nm LED source provides power to the PVLED, while MOTE emits 825-nm PPM pulses that encode electrophysiological signals.

The diode is used as a photovoltaic for 93.4% of the time and as an LED for 0.06% of the time, with the remainder of the time spent on transitions. By concentrating the transmitted power into short, bright pulses and encoding information in the timing of those pulses, PPM is much more resistant to noise than amplitude modulation and is very power efficient.

Atomic layer deposition (ALD) of SiO₂, Si₃N₄ and Al₂O₃ encapsulates MOTE against corrosive biological media without substantially increasing its volume (total encapsulation thickness is under 1.5 µm). High-pressure platinum (Pt) sputtering then provides not only favorable electrode impedance but also an effective and conformal light shield to prevent incident light from generating unwanted photocurrents in the electronics. Critically, each fabrication step is done in parallel, simultaneously fabricating close to 100 MOTEs per chip—and scalable to thousands of MOTEs per square centimeter of silicon (Figure 3).

Figure 3 Bulk fabrication of MOTEs (left) integrating two disparate technologies—CMOS (silicon based) and PVLED (AlGaAs based)—and a cross-sectional view (right) of a fully fabricated MOTE illustrating how the ALD dielectrics and sputter Pt together constitute a shield against biological media and unwanted photocurrents. Source: Cornell University

The underlying CMOS circuits provide low-noise amplification, stable biasing and PPM encoding, and drive the PVLED as an LED (Figure 4). Overall power budget is miserly: nominal power consumption is just one microwatt, divided among the amplifier (50.0%), encoder (10.5%), LED driver (26.2%), and support circuits (13.3%).

Figure 4 Systemic description of a MOTE and its external counterpart for communication—MOTE’s output PPM pulses are detected by an external photodiode before being passed through a decoder (a). Schematics of the front-end amplifier based on pseudo-resistors (left) and the charge pump for optical pulse generation shown on the right (b). Power and area distributions of a MOTE in which the amplifier and filter take most of the power for low-noise amplification (left), and the frame and integration overhead for protection against unwanted light and photocarriers take most of the area, as shown on right (c). Source: Cornell University

How well did they do?

By design, incident LED irradiance is limited to less than 70 mW/mm², well below the allowed threshold of 250 mW/ mm², which may inflict heat damage in the brain. The team first performed Petri-dish “static” tests before moving on to live rats. The heads of the implanted live mice were “restrained” while computer-controlled motor moved a rod to stimulate a whisker of an awake, head-fixed mouse.

The implant successfully recorded spikes of electrical activity from neurons as well as broader patterns of synaptic activity—all while the mice remained healthy and active. In two of the six implanted mice, they placed MOTEs on the brain surface, from which they were able to measure the electrocorticographic (ECoG) signals; in the other four mice, they inserted MOTEs into the barrel cortex.

As expected, MOTEs captured the neural responses to whisker stimulations and transmitted the neural signal spike. MOTES were left in the test “subjects” for up to 300 days and continued to function, although there was some degradation in performance, which the Cornell researchers attribute to deterioration of the platinum electrodes.

Why even bother with such a project, rather than using conventional “stick-in” electrodes? In addition to the obvious limitation imposed by the associated wired tether or even a wireless interface attached to the rat, one of the motivations is that traditional electrodes can irritate the brain as the tissue moves around the implant and thus can trigger an immune response. Their goal was to make the device small enough to minimize that disruption while still capturing brain activity faster than imaging systems, and without the need to genetically modify the neurons for imaging.

In you want to know more about the project, its circuitry, and the test results on the rats (I didn’t feel the need to go into detail on that!), check out their detailed and highly readable paper “A subnanolitre tetherless optoelectronic microsystem for chronic neural recording in awake mice” published in Nature Electronics.

Whether it’s rat implants or something non-biologic, these projects—with their tight focus, custom die, minimized number of functional blocks, and no frills or features beyond what is absolutely needed—show what analog designs can do in micropower and microsize designs, and that innovative analog design has not reached a terminal point. As the late, great analog designer Bob Pease liked to remind us, “one good op amp can do more than a thousand logic gates.”

Bill Schweber is a degreed senior EE who has written three textbooks, hundreds of technical articles, opinion columns, and product features. Prior to becoming an author and editor, he spent his entire hands-on career on the analog side by working on power supplies, sensors, signal conditioning, and wired and wireless communication links. His work experience includes many years at Analog Devices in applications and marketing.

Source link