Ultrasound backscatter-based implant for detecting electrophysiological signals

Miniaturized ultrasound transducers in implantable devices enable efficient wireless sensing and transmission of electrophysiological signals, addressing scalability and resolution issues in existing technologies, facilitating long-term use and cost reduction.

JP7580134B2Active Publication Date: 2024-11-11RGT UNIV OF CALIFORNIA
View PDF 3 Cites 0 Cited by

Patent Information

Application Number
JP2022131246
Authority / Receiving Office
JP · JP
Patent Type
Patents
Current Assignee / Owner
Priority Date
2016-07-07
Filing Date
2022-08-19
Publication Date
2024-11-11
Estimated Expiration
2037-07-07

AI Technical Summary

Technical Problem

Current implantable devices for sensing electrophysiological signals face challenges in achieving high spatial resolution, scalability, and efficient wireless communication, particularly for small nerves and muscles, with existing electrode-based techniques lacking the ability to record from multiple sites and wireless systems being inefficient at sizes smaller than 5mm.

Method used

Implantable devices utilizing miniaturized ultrasound transducers and electrodes to sense electrophysiological signals, which encode and transmit information via ultrasound backscatter, allowing for wireless power and data transmission, and include integrated circuits for modulation and digitization.

Benefits of technology

Enables efficient, long-term sensing of electrophysiological signals in deep tissues with high spatial resolution and scalability, reducing tissue inflammation and manufacturing costs, while overcoming limitations of electromagnetic communication.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure 0007580134000014
    Figure 0007580134000014
  • Figure 0007580134000015
    Figure 0007580134000015
  • Figure 0007580134000016
    Figure 0007580134000016
Patent Text Reader

Abstract

An implantable device configured to sense electrophysiological signals is provided. [Solution] An implantable device includes first and second electrodes attached to tissue and configured to sense electrophysiological signals, an integrated circuit having a multi-transistor circuit and a modulation circuit configured to modulate a current based on the electrophysiological signal, and an ultrasound transducer configured to emit ultrasound backscatter from the tissue based on the modulated current, the ultrasound backscatter encoding the electrophysiological signal. Also described herein are a system including an implantable device and an interrogator having an ultrasound transducer configured to transmit ultrasound to the implantable device or receive ultrasound backscatter from the implantable device, as well as a computer system for operating the implantable device and a method for detecting electrophysiological signals in tissue.
Need to check novelty before this filing date? Find Prior Art

Description

Detailed Description of the Invention

[0001] CROSS-REFERENCE TO RELATED APPLICATIONS This application claims priority to U.S. Provisional Application No. 62 / 359,672, filed July 7, 2016, entitled "NEURAL DUST AND ULTRASONIC BACKSCATER IMPLANTS AND SYSTEMS, AND APPLICATIONS FOR SUCH SYSTEMS," which is incorporated herein by reference for all purposes.

[0002] [Statement Regarding Federally Funded Research] This invention was made with government support awarded by the Defense Advanced Research Projects Agency (DARPA) under Grant No. HR0011-15-2-0006, and with government support awarded by the National Institutes of Health (NIH) under Grant No. R21-E027570.

[0003] [Technical field] The present invention relates to an implantable device that senses electrophysiological voltage signals in tissue and reports the electrophysiological voltage signals using ultrasound backscatter.

[0004] 〔background〕 The emerging field of bioelectronic medicine seeks to decode and modulate electrophysiological activity in vivo to achieve therapeutic effects in target organs. Current approaches to interface with peripheral nerves, the central nervous system, and muscles rely heavily on wires, making them problematic for long-term use, while new wireless approaches lack the size scalability required to interrogate small diameter nerves. Furthermore, traditional electrode-based techniques lack the ability to record from nerves with high spatial resolution or independently from many individual sites in a nerve bundle.

[0005] Recent technological advances and radical discoveries have sparked renewed interest in implantable systems that interface with the peripheral nervous system. Early clinical successes with peripheral nerve stimulation devices, such as those used to treat sleep apnea or control bladder function in paraplegics, have led clinicians and researchers to propose new targets for diseases ranging from diabetes to rheumatoid arthritis.

[0006] There is a need for new electrode-based recording techniques that can detect abnormalities in physiological signals and can be used to update stimulation parameters in real time. Such techniques would preferably include high-density, stable recording of many channels in a nerve, wireless implantable modules that can functionally characterize specific neural and myoelectric signals, and scalable device platforms that can interface with small nerves, 100 mm or less in diameter, and specific muscle fibers. Current approaches to recording peripheral nerve activity fall short of the following goals: For example, cuff electrodes provide stable long-term performance, but are limited to recording composite activity from the entire nerve; single-lead intra-vascular electrodes can record from multiple sites within a bundle, but cannot record densely from individual sites in multiple bundles; similarly, surface EMG arrays can record very densely, but do not capture fine details of deep or small muscles. Recently, wireless devices capable of untethered recording in rodents and non-human primates, as well as mm-scale integrated circuits for neural sensing applications, have been developed. For example, "Biederman et al., A 4.78 mm 2fully-integrated neuromodulation SoC combining 64 acquisition channels with digital compression and simultaneous dual stimulation, IEEE J. Solid State Circuits, vol. 5, pp. 1038-1047 (2015), Denison et al., A 2μW 100 nV / rtHz chpper-stabilized instrumentation amplifier for chronic measurement of neural field potentials, IEEE J. Solid State Circuits, vol. 42, pp. 2934-2945 (2007)” and “Muller et al., A minimally invasive 64-channel wireless uECOoG implant, IEE J. Soid State Circuits, vol. 50, pp. 344-359 (2015).” However, most wireless systems use electromagnetic (EM) energy coupling communication, which is highly inefficient in systems smaller than 5 mm due to the inefficiency of coupling electromagnetic waves into tissue at sizes smaller than 5 mm. New approaches are needed to further miniaturize wireless electronic platforms that can efficiently interface with small diameter nerves.

[0007] Summary of the Invention Disclosed herein are implantable devices configured to sense electrophysiological signals in tissue (e.g., muscle tissue or nerve tissue, etc.) and methods of sensing electrophysiological signals. Also disclosed herein are systems having one or more implantable devices and interrogators for activating the one or more implantable devices.

[0008] In some embodiments, the implantable device includes a first electrode and a second electrode attached to tissue and configured to sense an electrophysiological signal, an integrated circuit having a multi-transistor circuit and a modulation circuit configured to modulate a current based on the electrophysiological signal, and an ultrasound transducer configured to emit ultrasound backscatter based on the modulated current that encodes the electrophysiological signal from the tissue.

[0009] In some embodiments, the multi-transistor circuit is a digital circuit. In some embodiments, the digital circuit is configured to operate the modulation circuit. In some embodiments, the digital circuit is configured to send a digitized signal to the modulation circuit, the digitized signal being based on the sensed electrophysiological signal. In some embodiments, the digitized signal includes a unique implantable device identifier or a unique electrode pair identifier. In some embodiments, the digitized signal is compressed by a factor of 5 or more compared to an analog signal.

[0010] In some embodiments, the implantable device comprises a body having an ultrasonic transducer and a modulation circuit, the body having a length of a maximum dimension of about 5 mm or less. In some embodiments, the first electrode and the second electrode are spaced apart by about 0.5 mm or more. In some embodiments, the implantable device comprises a body having an ultrasonic transducer and a modulation circuit, the body having a volume of about 5 mm or less. 3 The following is the result.

[0011] In some embodiments, the implantable device comprises a non-responsive reflector.

[0012] In some embodiments, the tissue is muscle tissue or nerve tissue. In some embodiments, the tissue is part of the peripheral nervous system or part of the central nervous system. In some embodiments, the tissue is brain tissue or a peripheral nerve. In some embodiments, the tissue is skeletal muscle, cardiac muscle, or smooth muscle.

[0013] In some embodiments, the implantable device comprises three or more electrodes. In some embodiments, the integrated circuit comprises a spike detector. In some embodiments, the integrated circuit comprises a power circuit. In some embodiments, the integrated circuit comprises an analog-to-digital converter (ADC). In some embodiments, the integrated circuit comprises an amplifier string.

[0014] In some embodiments, the ultrasonic transducer is configured to receive ultrasonic waves that power an implantable device. In some embodiments, the ultrasonic transducer is configured to receive ultrasonic waves from an interrogator having one or more ultrasonic transducers. In some embodiments, the ultrasonic transducer is a bulk piezoelectric transducer, a piezoelectric micromachined ultrasonic transducer (PMUT), or a capacitive micromachined ultrasonic transducer (CMUT).

[0015] In some embodiments, the implantable device is implanted in a subject. In some embodiments, the subject is a human. In some embodiments, the implantable device is implanted in, on, or near the tissue. In some embodiments, the first electrode and the second electrode are implanted in or on the tissue.

[0016] In some embodiments, the implantable device is at least partially encapsulated by a biocompatible material. In some embodiments, at least a portion of the first electrode and the second electrode are not encapsulated by the biocompatible material.

[0017] Also disclosed herein is a system including one or more implantable devices and an interrogator having one or more ultrasound transducers configured to transmit ultrasound to or receive ultrasound backscatter from the one or more implantable devices. In some embodiments, the system includes a plurality of implantable devices. In some embodiments, the interrogator is configured to beam steer transmitted ultrasound to alternately focus the transmitted ultrasound on a first portion of the plurality of implantable devices or a second portion of the plurality of implantable devices. In some embodiments, the interrogator is configured to simultaneously receive ultrasound backscatter from at least two implantable devices. In some embodiments, the interrogator is configured to transmit ultrasound to or receive ultrasound backscatter from the plurality of implantable devices using time division multiplexing, spatial multiplexing, or frequency multiplexing. In some embodiments, the interrogator is configured to be wearable by a subject.

[0018] Further disclosed herein is a computer system comprising an interrogator having one or more ultrasound transducers, one or more processors, and a computer-readable non-transitory storage medium storing one or more programs configured to be executed by the one or more processors, the one or more programs including instructions for detecting an electrophysiological signal based on ultrasound backscatter encoding an electrophysiological pulse emitted from an implanted device, and instructions for determining a position of the implanted device relative to the one or more ultrasound transducers in the interrogator or detecting a movement of the implanted device based on ultrasound backscatter not responsive to the electrophysiological pulse emitted from the implanted device. In some embodiments, the one or more programs include instructions for determining a position or movement of the implanted device relative to the one or more ultrasound transducers in the interrogator. In some embodiments, the ultrasound backscatter responsive to the electrophysiological signal includes a digitized signal encoding the electrophysiological signal. In some embodiments, the one or more programs include instructions for attributing the sensed electrophysiological signal to the implantable device of a plurality of implantable devices that emitted ultrasound backscatter encoding the electrophysiological signal. In some embodiments, the sensed electrophysiological signal is attributable to the implantable device using time division multiplexing, spatial multiplexing, or frequency multiplexing. In some embodiments, the sensed electrophysiological signal is attributable to the implantable device using a unique identifier encoded in the ultrasound backscatter encoding the electrophysiological pulse.

[0019] Also disclosed herein is a method of detecting electrophysiological signals in tissue, the method including receiving ultrasound waves at one or more implantable devices, converting energy from the ultrasound waves into an electrical current flowing in a modulation circuit, detecting the electrophysiological signals, digitizing the electrophysiological signals, modulating the electrical current based on the digitized electrophysiological signals, converting the modulated electrical current into ultrasonic backscatter that encodes information about the electrophysiological signals, and emitting the ultrasonic backscatter to an interrogator having one or more transducers configured to receive the ultrasonic backscatter.

[0020] In some embodiments, a method of detecting electrophysiological signals in tissue includes transmitting ultrasound waves from an interrogator having one or more ultrasound transducers to one or more implanted devices, and receiving ultrasound backscatter from the one or more implanted devices, the ultrasound backscatter encoding digitized information about the electrophysiological signals.

[0021] In some embodiments, the method includes receiving the ultrasound backscatter with the interrogator, hi some embodiments, the ultrasound powers the one or more implantable devices.

[0022] In some embodiments of the methods disclosed herein, the tissue is muscle tissue or nerve tissue. In some embodiments, the tissue is part of the peripheral nervous system or part of the central nervous system. In some embodiments, the tissue is brain tissue or peripheral nerve. In some embodiments, the tissue is skeletal muscle, cardiac muscle, or smooth muscle.

[0023] In some embodiments, the method comprises reconstructing an electromyogram, an electroneurogram, an electrocardiogram, a compound action potential, multi-unit activity of multiple neurons, a local field potential, or an action potential.

[0024] In some embodiments, the method comprises implanting the one or more implantable devices. In some embodiments, the subject is a human.

[0025] In some embodiments, the method includes measuring the position or movement of the one or more implantable devices.

[0026] In some embodiments, in the methods disclosed herein, the digitized signal includes a unique implantable device identifier or a unique electrode pair identifier.

[0027] In some embodiments, in the methods disclosed herein, the one or more implantable devices comprises a plurality of implantable devices.

[0028] BRIEF DESCRIPTION OF THE DRAWINGS FIG. 1 is a schematic diagram of the Neural Dust system, including an external transceiver, a subdural interrogator, and a Neural Dust mote, as disclosed in Seo et al., Neural dust: an ultrasonic, low power solution for chronic brain-machine interfaces, arXiv: 1307.2196v1 (July 8, 2013).

[0029] FIG. 2A is a block diagram of an exemplary interrogator in the system disclosed herein. The illustrated interrogator includes an ultrasonic transducer array including multiple ultrasonic transducers. Each ultrasonic transducer in the ultrasonic transducer array is activated by a channel. The channel includes a switch that alternately configures the transducer to receive or transmit ultrasonic waves. FIG. 2B is a schematic diagram of another exemplary interrogator in the system disclosed herein. The illustrated interrogator includes two ultrasonic transducer arrays, each of which includes multiple ultrasonic transducers. The interrogator also includes an integrated circuit. The integrated circuit may include digital circuitry, and the digital circuitry may include a processor. The integrated circuit is connected to a user interface (which may include a display, a keyboard, buttons, etc.), a storage medium (i.e., a non-transitory storage device), an input / output device (which may be wireless (e.g., Bluetooth®)), and a power supply device (e.g., a battery).

[0030] FIG. 3A shows a block diagram of an exemplary interrogator that is wearable by a subject. The interrogator includes a wireless communication system (in the figure, a Bluetooth® radio). The wireless communication system may be used for communication with a computer system. FIG. 3B shows an exploded view of the wearable interrogator. The interrogator includes a battery, a wireless communication system, and a transducer array. FIG. 3C shows the wearable interrogator shown in FIG. 3B fully assembled with a harness for attachment to a subject. FIG. 3D shows the wearable interrogator attached to a subject, i.e., a rodent (but may be any animal, such as a human, dog, cat, horse, cow, pig, sheep, goat, chicken, monkey, rat, mouse, etc.). The interrogator includes a transducer array. The transducer array is secured to the subject's body with an adhesive. FIG. 3E shows a cross-sectional view of the transducer array of the interrogator shown in FIGS. 3A-3D.

[0031] 4 is a schematic diagram illustrating communication between a transducer in an interrogator and an implantable device having a miniaturized ultrasonic transducer. The interrogator transmits ultrasonic waves to the implantable device, the miniaturized ultrasonic transducer emits ultrasonic backscatter that is modulated by a sensor, and the interrogator receives the backscatter.

[0032] FIG. 5A shows a series of cycles of ultrasonic pulses emitted from the interrogator. When the transceiver board of the interrogator receives a trigger from the interrogator (e.g., FPGA), it generates a series of transmit pulses. At the end of the transmit cycle, the switch on the ASIC disconnects the transmit module and connects to the receive module. Each cycle is 100 microseconds. FIG. 5B shows an enlarged view of the transmit pulse sequence (i.e., one cycle) shown in FIG. 5A. In the cycle, there are six ultrasonic pulses of 1.85 MHz, and the ultrasonic pulses are repeated every 540 nanoseconds. FIG. 5C shows ultrasonic backscatter emitted from the implantable device. The ultrasonic backscatter arrives at the interrogator transducer at about 2t Rayleigh. FIG. 5D shows an enlarged view of the ultrasonic backscatter, which can be analyzed. The analysis of the ultrasonic backscatter includes the steps of filtering, rectifying, and integrating the ultrasonic backscatter wave. FIG. 5E shows an enlarged view of the filtered ultrasonic backscatter wave. The backscattered waves include a responsive region that responds to changes in impedance to the miniaturized ultrasonic transducer and a non-responsive region that does not respond to changes in impedance to the miniaturized ultrasonic transducer.

[0033] FIG. 6 illustrates one embodiment of an implantable device having a miniaturized transducer and a modulation circuit configured to modulate the current flowing in the miniaturized ultrasound transducer based on an electrophysiological signal sensed by an electrode pair.

[0034] FIG. 7A illustrates one embodiment of a miniaturized ultrasound transducer (referred to as a "piezoelectric element") connected to an ASIC. The ASIC has a power supply circuit and a modulation circuit (or "backscatter circuit"). The power supply circuit has an energy storage capacitor ("cap"). The electrodes may be embedded in tissue.

[0035] FIG. 7B illustrates an implantable device having a miniaturized ultrasound transducer, an integrated circuit, and an electrode pair. The integrated circuit has a modulation circuit, an AC composite amplifier string, and a power supply circuit. The power supply circuit has a full-wave rectifier, a voltage doubler, a reference, and a regulator. FIG. 7C illustrates an exemplary rectifier that can be used in the integrated circuit shown in FIG. 7B. FIG. 7D illustrates an exemplary amplifier string that can be used in the integrated circuit shown in FIG. 7B.

[0036] 8A shows a schematic diagram of an exemplary implantable device including a miniaturized ultrasound transducer and an ASIC on a printed circuit board (PCB). FIG. 8B shows a schematic diagram of another exemplary implantable device including a miniaturized ultrasound transducer and an ASIC on a printed circuit board (PCB).

[0037] FIG. 9 illustrates a method of manufacturing the implantable devices described herein.

[0038] FIG. 10 is a flow chart of a method for encapsulating an implantable device comprising amorphous silicon carbide.

[0039] Figure 11A shows different shaped vias used to connect components of an implantable device, and Figure 11B shows a serpentine trace arrangement for a deformable interconnect.

[0040] Figure 12 shows the time vs. temperature relationship for curing silver epoxy, which is one example of a material for attaching wire bonds in the manufacture of implantable devices.

[0041] Figure 13 shows electroneurograms (ENG) recorded using the implanted device. The dotted line shows the signal recorded by the ground truth electrodes. From the obtained data, general features including compound action potentials were reconstructed. The general features match the ground truth features.

[0042] FIG. 14 shows a schematic for encapsulating an implantable device in silicon carbide.

[0043] Figure 15 shows the prototype assembly schematic and PCB.

[0044] Figures 16A-E show the process steps to ensure a miniaturized ultrasonic transducer (PZT) of the desired size is assembled onto a PCB. In Figure 16A, epoxy solder paste is placed onto the substrate. In Figure 16B, the piezoelectric material is attached to the PCB. In Figure 16C, the piezoelectric material is diced to form a bulk piezoelectric ultrasonic transducer of the desired size. In Figure 16D, the ultrasonic transducer is wire bonded to the PCB. In Figure 16E, the PCB and ultrasonic transducer are encapsulated in PDMS.

[0045] FIG. 17 shows a schematic diagram of electrical impedance measurement using a vector network analyzer (VNA).

[0046] Figure 18A shows that the measured power transfer efficiency of bulk piezoelectric ultrasonic transducers of different sizes matches the simulated behavior. Figure 18B shows that the measured impedance spectroscopy of the PZT crystal matches the simulation. Figure 18C shows that the frequency response of the miniaturized ultrasonic transducer to the resulting power is about 6.1 MHz.

[0047] FIG. 19 is a schematic diagram of an exemplary ultrasonic transducer that can be used as part of an interrogator.

[0048] 20 is a schematic diagram of an acoustic characterization setup using a calibrated ultrasonic transducer for verification of power delivery. The ultrasonic receiver is separate from the ultrasonic transmitter.

[0049] Figure 21A shows the output of a 5 MHz transducer as the hydrophone is moved away from the transducer surface, and Figure 21B shows that the peak of the ablated output is shifted to the left compared to the peak in water.

[0050] Figure 22A shows an XZ cross section of the transducer output illustrating the Rayleigh distance and the sharp transition from near to far propagation, while Figure 22B shows an XY beam cross section showing the 6 dB bandwidth of a 2.2 mm beam.

[0051] Figure 23A shows a focused 2D beam pattern from a transducer array in the XY plane. The measured beam is close to the simulated beam in both the X and Y directions. Figure 23B shows the delay time applied to each transducer element in the ultrasound transducer array. Figure 23C shows the simulated 2D XZ cross-sectional beam pattern.

[0052] Figure 24A shows the beam steering of an ultrasound beam transmitted from a transducer array. Under each beam pattern is the delay required for each transducer in the transducer array to obtain the measured beam pattern, as shown in Figure 24B. Figure 24C shows the one-dimensional beam pattern in the X-axis for each beam pattern shown in Figure 24A. The measured beam patterns are very close to the simulated beam patterns.

[0053] FIG. 25 shows the coupling efficiency and received power of a simulated miniaturized ultrasound transducer in tissue at a transmission distance of 5 mm.

[0054] 26A-D show an overview of an exemplary system with an implantable device. FIG. 26A shows an external transducer that powers and communicates with an implantable device placed locally in the body. Driven by a custom transceiver board, the transducer alternates between sending a series of pulses that power the device and receiving reflected pulses that are modulated by the electrophysiological signal. FIG. 26B shows the implantable device fixed to the sciatic nerve of an anesthetized rat. The insert in FIG. 26B shows the implantable device with optional test leads. FIG. 26C shows components of an exemplary implantable device. The implantable device is assembled on a flexible PCB and includes a piezoelectric crystal, one custom transistor, and a recording electrode pair. FIG. 26D shows a close-up of the implantable device with calibration leads for measuring the electrophysiological signal (ground truth) and voltage obtained on the piezoelectric crystal on a flexible PCB. The calibration leads were removed during in vivo experiments.

[0055] FIG. 27 illustrates the communication between an exemplary interrogator and an implantable device. The top of FIG. 27 is a schematic of the information flow. The bottom of FIG. 27 shows the time course of signals at each step represented by the diagram shown at the top of the figure. In FIG. 27A, the FPGA from the interrogator generates a trigger signal to start recording. FIG. 27B shows the extracellular electrophysiological potential sent to the recording electrodes on the implantable device. FIG. 27C shows the transceiver board generates a series of transmit pulses upon receiving the trigger from the FPGA. At the end of the transmit cycle, a switch on the interrogator ASIC disconnects from the transmit module and connects with the receive module. FIG. 27D shows a close-up of the successive transmit pulses, showing 6 pulses at 1.85 MHz. FIG. 27E shows the backscatter from the implantable device. The backscatter reaches the transducer at about 2t Rayleigh distance. FIG. 27F shows a close-up of the backscatter waveform. The backscatter waveform contains a large saturating signal, whose overlap with the transmit pulse feeds through electrically and is ignored. As the backscatter pulse returns, it can be seen following the transmit window. FIG. 27G shows the filtered and rectified backscatter waveform and the area under the curve calculated to reconstruct the waveform. FIG. 27H shows the reconstructed waveform sampled at 10 kHz. Each portion of the reconstructed waveform is calculated by calculating the area under the curve of the properly reflected pulse received every 100 μs.

[0056] FIG. 28A shows the reduced normalized peak pressure as a function of distance from the surface of an exemplary transducer (interrogator), with a reduced focus at about 8.9 mm at 1.85 MHz. FIG. 28B shows the beam patterns in the XY cross section at near, Rayleigh, and far distances, and the corresponding 1-D voltage plot at y=0. These patterns show that the beam is focused at the Rayleigh distance. FIG. 28C shows that the transducer output pressure is a linear function of the input voltage (peak-to-peak voltage up to 32V). FIG. 28D (a reproduction of FIG. 5E shown in a different context) is an exemplary backscatter waveform shown in different regions. It can be seen that the backscatter waveform is flanked (in time) by regions corresponding to reflections originating from non-responsive regions; these regions correspond to pulses reflected from other device components. Measurements in non-responsive regions may be taken as a reference, where the non-responsive regions are not reading biological data. This distinction allows us to subtract out the behavior of all the configurations of the external transducers in the experiment. Figure 28E shows the calibration curve obtained from the custom water tank fixture, with a noise floor of 0.18 mV. rms FIG. 28F shows the effect of the noise floor as a function of lateral mispointing, followed by a falloff in beam pattern power. FIG. 28G shows a one-dimensional plot of the transducer off-axis voltage and power drop at Rayleigh distance, y=0. FIG. 28H shows a plot of the drop versus angular mispointing. Angular mispointing distorts the beam pattern, i.e., the beam pattern becomes elliptical instead of circular. This lengthens the radius of the focal spot (and distributes the energy over a wider area). In other words, the distortion of the focal spot relaxes the constraint on mispointing.

[0057] FIG. 29A shows an in vivo experimental setup for recording EMG from the gastrocnemius muscle of a rat. The implantable device was placed on the exposed muscle surface and the wound was closed with surgical sutures. An external transducer couples ultrasound to the implantable device and wireless data is recorded and displayed on a computer system (e.g., a laptop computer). FIG. 29B shows a comparison of ground truth measurements and reconstructed EMG signals over multiple trials. 20 ms samples were recorded with 6 s between stimuli. FIG. 29C shows the power spectral density (PSD) of the recorded EMG signal. The PSD was 4.29e4μV2 / Hz and 3.11e4μV2 / Hz at 107 Hz for the ground truth and reconstructed dust data, respectively, with some harmonics seen at the edges of the waveform. FIG. 29D shows that the wireless backscatter data recorded at t=0 min and t=30 min matched with R=0.901.

[0058] FIG. 30A shows EMG signals of various intensities recorded in vivo with electrodes on a PCB at different stimulation intensities. FIG. 30B shows that EMG responses of similar slopes were recorded wirelessly with an implantable device. FIG. 30C shows that EMG signals reconstructed from ground truth and wireless backscatter data agreed with R=0.795 for response-saturating stimulation amplitude (100%) (R=0.60, 0.64, 0.67, 0.92 for 54%, 69%, 77%, 89%, respectively). In FIG. 30D, quantitative comparison showed agreement with salient features of less than 0.4 mV. In FIG. 30E, EMG peak-to-peak voltage showed the expected sigmoidal correlation with stimulation intensity.

[0059] FIG. 31A shows ENG signals of various intensities recorded in vivo with electrodes on a PCB at different stimulation intensities. FIG. 31B shows ENG responses of similar slopes recorded wirelessly with a mote. FIG. 31C shows that ENG signals reconstructed from ground truth and wireless backscatter data agreed with R=0.886 for response-saturating stimulation amplitude (100%) (R=0.822, 0.821, 0.69, 0.918, 0.87 for 44%, 61%, 72%, 83%, 89%, respectively). In FIG. 31D, quantitative comparison showed less than 0.2 mV agreement with salient features. In FIG. 31E, ENG peak-to-peak voltage showed the expected sigmoidal correlation with stimulation intensity.

[0060] Figure 32A shows the ENG response recorded over time at different electrode spacings, and Figure 32B shows the peak-to-peak ENG at various electrode spacings.

[0061] Figure 33A shows ultrasound backscatter from an implantable device, with the implantable device embedded in ultrasound coupling gel used to mimic tissue. The ultrasound scattering includes transmit penetration and termination centered at 26 microseconds, and miniaturized ultrasound transducer backscatter centered around 47 microseconds. Figure 33B is a close-up of the miniaturized ultrasound transducer backscatter region (response region). This shows that the signal input to the implantable device results in amplitude modulation.

[0062] FIG. 34 shows digital data corresponding to the ASCII characters "hello world" wirelessly obtained from an implantable device by unipolar encoding and modulation of the backscattered amplitude pulse.

[0063] Detailed Description of the Invention The implantable device described herein includes a miniaturized ultrasound transducer (such as a miniaturized piezoelectric transducer) and two or more electrodes configured to detect electrophysiological voltage signals (such as action potentials). The electrophysiological voltage signals are sensed, for example, in nerve tissue or muscle tissue. The miniaturized ultrasound transducer receives ultrasonic energy from an interrogator and provides power to the implantable device. The interrogator may be an external interrogator or an implanted interrogator. The interrogator includes a transmitter and a receiver, which may be on the same component or on different components. Here, the transmitter and receiver may be integrated as a combined transceiver. The mechanical energy of the ultrasound transmitted from the interrogator vibrates the miniaturized ultrasound transducer on the implantable device, which generates a current. The current flow in the miniaturized ultrasound transducer is modulated by an electrical circuit in the implantable device based on the detected electrophysiological voltage signals, and the modulated current passes through the miniaturized ultrasound transducer. The miniaturized ultrasound transducer emits ultrasonic backscatter that communicates information representative of the sensed electrophysiological voltage signal. The emitted ultrasonic backscatter is detected by a receiver component of the interrogator.

[0064] An important advantage of the implantable device is that it can detect electrophysiological voltage signals in deep tissues while being wirelessly powered and wirelessly transmit information about the detected electrophysiological signals to an interrogator, which can be an external interrogator or an interrogator that relays information to an external component. Thus, the implantable device can remain in the subject for long periods of time without the need to charge batteries or retrieve information stored in the device. The above advantages, in turn, allow for smaller devices and reduced manufacturing costs. Another advantage is that the use of ultrasound allows for the correlation of relative time and distance of data communication, which helps determine the real-time location or movement of the implantable device.

[0065] Electromagnetic (EM) power transmission is not practical for powering small implantable devices because of the attenuation of power in tissue and the relatively large openings (e.g., antennas or coils) required to harvest them. See, e.g., Seo et al., Neural dust: an ultrasonic, low power solution for chronic brain-machine interfaces, arXiv: 1307.2196v1 (July 8, 2013). To use EM to adequately power implanted devices, the implant would need to be shallowly implanted and the EM waves would need to heat the tissue extra much to penetrate the tissue and reach the implantable device. In contrast to EM, ultrasonic transmission has low attenuation of power in tissue. This is due to the relatively low absorption of ultrasonic energy by tissue and the short wavelength of ultrasound (compared to electromagnetic waves). Furthermore, the shorter wavelength delivered by ultrasound offers better spatial resolution at lower frequencies than radio waves.

[0066] Ultrasound transducers have found applications in a variety of fields, including imaging, high intensity focused ultrasound (HIFU), non-destructive testing of materials, communication and power delivery through steel walls, underwater communication, transcutaneous power delivery, and energy harvesting. For example, Ishida et al., Insole Pedometer with Piezoelectric Energy Harvester and 2 V Organic Circuits, IEEE J. Solid-State Circuits, vol. 48, no. 1, pp. 255-264 (2013), Wong et al., Advantages of Capacitive Micromachined Ultrasonics Transducers (CMUTs) for High Intensity Focused Ultrasound (HIFU), IEEE Ultrasonics Symposium, pp. 1313-1316 (2007), Ozeri et al., Ultrasonic Transcutaneous Energy Transfer for Powering Implanted Devices, Ultrasonics, vol. 50, no. 6, pp. 556-566 (2010), and Richards et al., Efficiency of Energy Conversion for Devices Containing a Piezoelectric Component, J. Micromech. Microeng., See, e.g., vol. 14, pp. 717-721 (2004). Unlike electromagnetics, the use of ultrasound as an energy transfer method has not been included in widespread civilian applications and has often been overlooked because electromagnetics are more efficient over short distances and large apertures. However, the low speed of sound allows operation at dramatically lower frequencies within tissue at the size of the implantable devices discussed herein, and acoustic losses in tissue are generally substantially less than the attenuation of electromagnetics in tissue.

[0067] Ultrasound has a substantially shorter wavelength compared to EM due to its relatively slow sound speed. Thus, for the same transmission distance, ultrasonic systems can operate at longer distances and therefore can cover a larger space than EM transmitters. Furthermore, acoustic losses in tissue are generally less than electromagnetic attenuation in tissue. This is because acoustic wave transmission relies on tissue compression and rarefaction rather than time-dependent electric and / or magnetic fields that generate displacement currents at the tissue surface.

[0068] In some embodiments, the "Neural Dust" system includes very small, implantable devices called Neural Dust, or "motes," an implantable ultrasound transceiver that communicates with each of the motes using ultrasound transmissions and backscattered transmissions reflected from the motes, and an external transceiver that wirelessly communicates with the ultrasound transceiver. See Seo et al., Neural dust: an ultrasonic, low power solution for chronic brain-machine interfaces, arXiv: 1307.2196v1 (July 8, 2013) (“Seo et al., 2013”), Seo et al., Model validation of untethered, ultrasonic neural dust motes for cortical recording, J. Neuroscience Methods, vol. 244, pp. 114-122 (2014) (“Seo et al., 2014”), and Bertrand et al., Beamforming approaches for untethered, ultrasonic neural dust motes for cortical recording: a simulation study, IEE EMBC, vol. 2014, pp. 2625-2628 (2014). The neural dust systems described in these papers are used for cortical recordings (i.e., recording of electrical signals from the brain). In the application described in these papers, a mote is implanted in brain tissue (cortex), an ultrasound transceiver is implanted below the dura, above the cortex, and an external transceiver is placed on the patient's head in close proximity to where the subdural ultrasound transceiver is implanted, as shown in Figure 1 of Seo et al., 2013.

[0069] It was shown in Seo et al., 2013 and Seo et al., 2014 that, in theory, the neural dust system can be used to develop small implants (millimeter scale or less) for wirelessly recording neural data. Accurate detection of electrophysiological signals is enhanced by accurately measuring the position or movement of the implanted device. This ensures accurate attribution of detected signals to the tissue that generated them and filtering reflected signals that may result from movement. As described herein, the position and movement of the implanted device can be accurately measured by analyzing non-responsive ultrasound backscatter. It has further been found that the implanted device can transmit a digitized signal encoded in the ultrasound backscatter. The digitized signal improves the reliability of electrophysiological signal detection (e.g., by filtering false positive signals) and data compression (which can be particularly advantageous, for example, when the implanted device includes multiple electrodes). The digitized signal can also include a unique identification signal in the ultrasound backscatter when multiple implanted devices are used or when the implanted device includes multiple electrodes.

[0070] It has further been found that the implantable devices described herein can be used to sense electrophysiological signals in vivo, including electrophysiological signals occurring in peripheral nerve or muscle tissue. In some embodiments, the implantable device is attached to nerve tissue (such as tissue in the central or peripheral nervous system) or muscle tissue (such as smooth muscle, skeletal muscle, or cardiac muscle) and can be used to report electroneurograms, electrocardiograms, or electromyograms. In some embodiments, attachment to the nerve or muscle tissue is such that the implantable device does not completely surround the nerve or muscle tissue. In some embodiments, the implantable device is on, embedded in, or adjacent to the nerve or muscle tissue. In some embodiments, an electrode of the implantable device is attached to the nerve or muscle tissue. For example, the electrode may be on, embedded in, or adjacent to the nerve or muscle tissue. In some embodiments ... embedded in the nerve or muscle tissue. In some embodiments, the electrode may be In some embodiments, the implantable device is positioned near tissue and the electrodes may extend from the implantable device to reach the tissue.

[0071] The nervous tissue may be part of the central nervous system (such as the brain (e.g., cortex) or spinal cord) or may be part of the peripheral nervous system (such as nerves that may be somatic or autonomic nerves). Exemplary nerves include the sciatic nerve, vagus nerve, vagus nerve branches, tibial nerve, splenic nerve, splanchnic nerve, pudendal nerve, sacral nerve, supraorbital nerve, and occipital nerve. The muscle tissue may be, for example, skeletal muscle, smooth muscle, or cardiac muscle. Exemplary muscles include the gastrocnemius muscle, pelvic muscle, gastric smooth muscle, and cardiac muscle.

[0072] The implantable devices described herein may be implanted in or used in a subject (i.e., an animal). In some embodiments, the subject is a mammal. For example, the subject may be a rodent (such as a mouse, rat, or guinea pig), cat, dog, chicken, pig, cow, horse, sheep, rabbit, chicken, bat, monkey, etc. In some embodiments, the subject is a human.

[0073] <Definition> As used herein, the singular forms (a, an, and the) include the plural forms unless the context clearly indicates otherwise.

[0074] The term "about" with respect to values ​​and parameters herein includes (and refers to) the variation that is directed to the value or parameter itself. For example, the term "about X" includes the term "X."

[0075] The term "miniaturized" refers to any material or component that has a length in its greatest dimension that is about 5 millimeters or less (e.g., about 4 mm or less, about 3 mm or less, about 2 mm or less, about 1 mm or less, or about 0.5 mm or less). In certain embodiments, the length in its greatest dimension of a "miniaturized" material or component is about 0.1 mm to about 5 mm (e.g., about 0.2 mm to about 5 mm, about 0.5 mm to about 5 mm, about 1 mm to about 5 mm, about 2 mm to about 5 mm, about 3 mm to about 5 mm, or about 4 mm to about 5 mm). "Miniaturized" also refers to any material or component that has a length in its greatest dimension that is about 5 millimeters or less in volume. 3 Less than (for example, about 4 mm 3 Below, 3mm 3 Below, 2mm 3 Less than or equal to 1mm 3 In certain embodiments, the volume of a "miniature" material or component is less than about 0.5 mm 3 ~about 5mm 3 , about 1mm 3 ~about 5mm 3 , about 2 mm 3 ~about 5mm 3 , about 3mm 3 ~about 5mm 3, or about 4 mm 3 ~about 5mm 3 It is.

[0076] A "piezoelectric transducer" is a type of ultrasound transceiver that includes a piezoelectric material, which may be quartz, ceramic, polymer, or any other natural or synthetic piezoelectric material.

[0077] The term "subject" refers to an animal.

[0078] A "non-responsive" ultrasound wave is one that has a reflectivity that is independent of the detected signal. A "non-responsive reflector" is a component of an implantable device that reflects ultrasound waves such that the reflected waveform is independent of the detected signal.

[0079] Each embodiment and each variant of the invention described herein is to be understood as including "consisting of" and / or "cosisting essentially of" each embodiment and variant.

[0080] When a range of values ​​is given, it is understood that each intervening value between the upper and lower limits of that range, as well as any other stated or intervening value within the stated range, is encompassed within the scope of the disclosure. Where a stated range includes an upper or lower limit, ranges excluding those limits are also included in the disclosure.

[0081] One or more or all of the features of the various embodiments described herein may be combined to form other embodiments of the present invention. The headings used herein are organizational and should not be construed as limiting the subject matter of the described invention.

[0082] The features and preferred configurations described above with respect to the "embodiments" are separate preferred configurations and are not limited to the particular embodiment. The features and preferred configurations may be freely combined with the features of other embodiments, if technically possible, to form preferred combinations of the respective features.

[0083] This specification is presented to enable one skilled in the art to make and use the invention and is provided in the context of a patent application and its respective requirements. Various modifications to the described embodiments will be readily apparent to those skilled in the art, and the general principles herein may be applied to other embodiments. Thus, the present invention is not intended to be limited to the embodiments shown, but is to be accorded the widest scope consistent with the principles and features described herein. Additionally, the headings are organizational and should not be considered limiting. Finally, the disclosures of all patents and publications cited in this application are incorporated herein by reference for all purposes.

[0084] <Caller> The interrogator can wirelessly communicate with one or more implanted devices using ultrasound. The ultrasound is used to power and / or operate the implanted devices. The interrogator can also receive ultrasound backscatter from the implanted devices. The ultrasound backscatter encodes information representative of the sensed electrophysiological voltage signal. The interrogator includes one or more ultrasound transducers. The ultrasound transducers can operate as ultrasound transmitters and / or ultrasound receivers (or transceivers that can be configured to alternately transmit or receive ultrasound). The one or more transducers may be arranged as a transducer array, and the interrogator may optionally include one or more transducer arrays. In some embodiments, the ultrasound transmission function is separate from the ultrasound reception function on another device. That is, optionally, the interrogator includes a first device that transmits ultrasound to the implanted devices and a second device that receives ultrasound backscatter from the implanted devices. In some embodiments, the transducers in the array may be regularly spaced, irregularly spaced, or sparsely spaced. In some embodiments the array is flexible, in some embodiments the array is planar, and in some embodiments the array is non-planar.

[0085] An exemplary interrogator is shown in Figure 2A. The illustrated interrogator shows a transducer array having multiple ultrasonic transducers. In some embodiments, the transducer array includes 1 or more, 2 or more, 3 or more, 5 or more, 7 or more, 10 or more, 15 or more, 20 or more, 25 or more, 50 or more, 100 or more, 250 or more, 500 or more, 1000 or more, 2500 or more, 5000 or more, 10,000 or more transducers. In some embodiments, the transducer array includes 100,000 or less, 50,000 or less, 25,000 or less, 10,000 or less, 5000 or less, 2500 or less, 1000 or less, 500 or less, 200 or less, 150 or less, 100 or less, 90 or less, 80 or less, 70 or less, 60 or less, 50 or less, 40 or less, 30 or less, 25 or less, 20 or less, 15 or less, 10 or less, 7 or less, or 5 or less transducers. The transducer array may be, for example, a chip that includes 50 or more ultrasound transducer pixels. The interrogator shown in FIG. 2A illustrates one transducer array. However, the interrogator may include one or more, two or more, or three or more separate arrays. In some embodiments, the interrogator includes 10 or fewer transducer arrays (e.g., 9, 8, 7, 6, 5, 4, 3, 2, or 1 transducer array). Separate transducer arrays can be located, for example, on different parts of the subject and can communicate with the same implantable device or different implantable devices. In some embodiments, the transducer arrays are located on both sides of the implantable device. The interrogator may include an ASIC. The ASIC includes a channel for each transducer in the transducer array. In some embodiments, the channel includes a switch (shown as "T / Rx" in FIG. 2A). The switch can alternately configure the transducer connected to the channel to transmit ultrasound or receive ultrasound. The switch can isolate the ultrasound receive circuitry from the higher voltage ultrasound transmit circuitry.In some embodiments, the transducer connected to the channel is configured to only receive ultrasound or only transmit ultrasound, and the switch is optionally omitted from the channel. The channel may include a delay control. The delay control operates to control the transmitted ultrasound. The delay control may control, for example, phase shift, time delay, pulse frequency, and / or waveform (including amplitude and wavelength). The delay control may be connected to a level shifter. The level shifter shifts the input pulse from the delay control to a higher voltage than the transducer uses to transmit ultrasound. In some embodiments, data indicating the waveform and frequency for each channel may be stored in a "wave table." This allows the transmit waveform for each channel to be different. The delay control and level shifter may then "stream" this data into the actual transmit signal sent to the transducer array. In some embodiments, the transmit waveform for each channel may be generated directly from a high speed serial output from a microcontroller or other digital system and sent to the transducer elements via a level shifter or high voltage amplifier. In some embodiments, the ASIC includes a charge pump (shown in FIG. 2A) that converts a first voltage supplied to the ASIC to a second voltage of higher voltage. The second voltage is applied to the channel. The channel can be controlled by a controller (e.g., a digital controller) that activates a delay control. In the ultrasound receiving circuit, the received ultrasound is converted to a current by a transducer (set in a receive mode), and the current is transmitted to a data acquisition circuit. In some embodiments, the receiving circuit includes an amplifier (which can compensate for losses in tissue), an analog-to-digital converter (ADC), a variable-gain-amplifier or a time-gain-controlled variable-gain-amplifier, and / or a bandpass filter. The ASIC can draw power from a power source.An example of a power source is a battery, which is preferred in embodiments where the interrogator is wearable. In the embodiment shown in FIG. 2A, the voltage to the ASIC is provided from a 1.8V power supply and increased to 32V by a charge pump. However, any suitable voltage can be used. In some embodiments, the interrogator includes a processor and / or a non-transitory computer readable memory. In some embodiments, the above-mentioned channels do not include a T / Rx switch, but instead include an Rx (receive) with a separate Tx (transmit) and a high voltage Rx (receive circuit) in the form of a low noise amplifier with good saturation recovery. In some embodiments, the T / Rx circuit includes a circulator. In some embodiments, the transducer array includes more transducer elements than there are processing channels in the interrogator transmit / receive circuits, and a multiplexer selects a different set of transmit elements for each pulse. For example, 64 transmit and receive channels are connected to 192 physical transducer elements via a 3:1 multiplexer, such that only 64 transducer elements are active for any given pulse.

[0086] FIG. 2B illustrates another embodiment of an interrogator. As shown in FIG. 2B, the interrogator includes one or more transducers 202. Each transducer 202 is connected to a transmit / receive switch 204. The transmit / receive switch 204 can be configured to alternately configure the transducer to transmit ultrasound or receive ultrasound. The transmit / receive switch is connected to a processor 206 (e.g., a central processing unit (CPU), a custom dedicated processing processor ASIC, a field programmable gate array (FPGA), a microcontroller unit (MCU), or a graphics processing unit (GPU)). In some embodiments, the interrogator further includes an analog-to-digital converter (ADC) or a digital-to-analog converter (DAC). The interrogator may also include a user interface (e.g., a display and one or more buttons for controlling the interrogator), a storage device, a power source (e.g., a battery), and / or an input / output port (which may be wired or wireless).

[0087] In some embodiments, the interrogator is implanted. An implanted interrogator may be preferred if the implanted device is implanted in an area blocked by obstacles that do not easily transmit ultrasound. For example, the interrogator may be implanted subcranially (subdurally or epidurally). A subcranial interrogator can communicate with an implanted device implanted in the brain. Because ultrasound is blocked by the skull, a subcranially implanted interrogator can communicate with an implanted device implanted in the brain. In another example, the implanted interrogator can be implanted as part of, behind, or within another implanted device or prosthesis. In some embodiments, the implanted interrogator can communicate with and / or be powered by an external device, for example, by EM or RF signals.

[0088] In some embodiments, the interrogator may be an external (i.e., non-implanted) interrogator. As an example, an external interrogator may be wearable and secured to the body with a strap or adhesive. In another example, an external interrogator may be rod-like and held by a user (such as a medical professional). In some embodiments, the interrogator may be held to the body via sutures, simple surface tension, a retainer in clothing such as a furoshiki, a sleeve, a rubber band, or subcutaneous fixation. The interrogator transducer or transducer array may be located separate from the rest of the transducers. For example, the transducer array may be secured to the subject's skin at a first location (such as a location adjacent to one or more implanted devices), and the rest of the interrogator may be located at a second location, with the transducer or transducer array connected to the rest of the interrogator by wires. FIGS. 3A-3E show examples of wearable external interrogators. FIG. 3A shows a block diagram of an interrogator having a transducer array including multiple transducers, an ASIC including a channel for each transducer in the transducer array, a battery (in the illustrated embodiment, a lithium polymer battery (LiPo)), and a wireless communication system (such as a Bluetooth® system). FIG. 3B shows an exploded view of a wearable interrogator having a printed circuit board (PCB) 302 (including the ASIC), a wireless communication system 304, a battery 306, an ultrasonic transducer array 308, and wires 310 (connecting the ultrasonic transducer array 308 to the ASIC). FIG. 3C shows the wearable interrogator 312 shown in FIG. 3B with a harness 314 attached, which can be used to attach the interrogator to a target. FIG. 3D shows the interrogator 316 assembled and attached to a target, with the transducer array 308 attached in a first position and the remainder of the interrogator attached in a second position.3E shows a cross-sectional schematic diagram of an exemplary ultrasonic transducer array 308 including a circuit board 318, vias 320 attaching each transducer 322 to the circuit board 318, a metallized polyester film 324, and an absorptive support layer 326. The metallized polyester film 324 provides a common ground and acoustic matching for each transducer, while the absorptive support layer 326 (such as polyurethane infused with tungsten powder) can reduce resonance of the individual transducers.

[0089] The particular design of the transducer array depends on the desired penetration depth, the size of the aperture, and the size of the transducers in the transducer array. The Rayleigh distance "R" of the transducer array is calculated as follows:

[0090]

number

[0091] where D is the size of the aperture and λ is the wavelength of ultrasound in the propagation medium (i.e., tissue). As understood in the art, the Rayleigh distance is the distance at which the beam emitted by the transducer array is fully formed. That is, to maximize the power received, the applied pressure focuses at the natural focus of the Rayleigh distance. Therefore, in some embodiments, the distance of the implantable device from the transducer is approximately the same as the Rayleigh distance.

[0092] Individual transducers in a transducer array can be modulated to control the Rayleigh distance and position of the beam of ultrasound emitted by the transducer array through a process of beamforming or beamsteering. Techniques such as Linearly Constrained Minimum Variance (LCMV) beamforming can be used to allow multiple implanted devices to communicate with an external ultrasound transceiver. See, for example, Bertrand et al., Beamforming Approaches for Untethered, Ultrasonic Neural Dust Motes for Cortical Recording: a Simulation Study, IEEE EMBC (Aug. 2014). In some embodiments, beamsteering is performed by adjusting the power or phase of the ultrasound emitted by the transducers in the transducer array.

[0093] In some embodiments, the interrogator has one or more of instructions for beam steering ultrasound using one or more transducers, instructions for determining a relative position of one or more implantable devices, instructions for monitoring relative movement of one or more implantable devices, instructions for recording relative movement of one or more implantable devices, and instructions for deconvoluting backscatter from multiple implantable devices.

[0094] <Communication between embedded devices and interrogators> The implantable device and the interrogator communicate with each other wirelessly using ultrasound. The implantable device receives ultrasound from the interrogator via a miniaturized ultrasound transducer on the implantable device. Vibration of the miniaturized ultrasound transducer on the implantable device generates a voltage across the electrical terminals of the transducer, causing a current to flow in the device (including the sensor and / or ASIC, if present). Electrophysiological pulses in the tissue are sensed by electrodes on the implantable device, and the current flowing in the transducer (which is converted to ultrasound backscatter and emitted from the implantable device) is modulated based on the sensed electrophysiological signal. In some embodiments, the modulation of the current may be an analog signal, which may be modulated directly by, for example, the sensed electrophysiological signal. In some embodiments, the modulation of the current encodes a digitized signal, which may be controlled by digital circuitry in an integrated circuit. The backscatter is received by an external ultrasound transceiver (which may be the same or different from the external ultrasound transceiver that transmitted the initial ultrasound). In this way, information from electrophysiological signals can be encoded by changes in the amplitude, frequency, or phase of the backscattered ultrasound waves.

[0095] FIG. 4 shows an interrogator communicating with an implantable device. An external ultrasound transceiver emits ultrasound waves ("carrier waves") that can pass through tissue. The carrier waves generate mechanical vibrations on a miniaturized ultrasound transducer (e.g., a miniaturized bulk piezoelectric transducer, PMUT, or CMUT). A voltage is generated in the miniaturized ultrasound transducer, which causes a current to flow in an integrated circuit on the implantable device. In some embodiments, the implantable device includes an ASIC, and a current flows through the miniaturized ultrasound transducer and the ASIC. The integrated circuit is configured to detect electrophysiological pulses, and the integrated circuit can modulate the current flowing through the miniaturized ultrasound transducer based on the detected electrophysiological pulses. The current flowing through the miniaturized ultrasound transducer causes backscattered ultrasound waves to be emitted from the transducer on the implantable device. The detected electrophysiological signal directly or indirectly (such as via the integrated circuit) modulates the current flowing through the miniaturized ultrasound transducer, so that the backscattered waves encode information about the detected electrophysiological pulses. The backscattered waves can be detected by the interrogator and analyzed to distinguish electrophysiological signals detected from the implanted device.

[0096] Communication between the interrogator and the implantable device can use a pulse-echo technique of transmitting and receiving ultrasound. In the pulse-echo technique, the interrogator transmits a series of interrogation pulses at a predetermined frequency and then receives backscattered echoes from the implantable device. In some embodiments, the pulses are about 200 nanoseconds (ns) to about 1000 ns long (e.g., about 300 ns to about 800 ns, about 400 ns to about 600 ns, or about 540 ns long). In some embodiments, the pulses are about 100 ns long or longer (e.g., about 150 ns long or longer, 200 ns long or longer, 300 ns long or longer, 400 ns long or longer, 500 ns long or longer, 540 ns long or longer, 600 ns long or longer, 700 ns long or longer, 800 ns long or longer, 900 ns long or longer, 1000 ns long or longer, 1200 ns long or longer, or 1500 ns long or longer). In some embodiments, the pulses are about 2000 ns or less in length (e.g., 1500 ns or less, 1200 ns or less, 1000 ns or less, 900 ns or less, 800 ns or less, 700 ns or less, 600 ns or less, 500 ns or less, 400 ns or less, 300 ns or less, 200 ns or less, or 150 ns or less). In some embodiments, the pulses are separated by a dwell time. In some embodiments, the dwell time is about 100 ns or more in length (e.g., about 150 ns or more, 200 ns or more, 300 ns or more, 400 ns or more, 500 ns or more, 540 ns or more, 600 ns or more, 700 ns or more, 800 ns or more, 900 ns or more, 1000 ns or more, 1200 ns or more, or 1500 ns or more). In some embodiments, the dwell time length is about 2000 ns or less (e.g., about 1500 ns or less, 1200 ns or less, 1000 ns or less, 900 ns or less, 800 ns or less, 700 ns or less, 600 ns or less, 500 ns or less, 400 ns or less, 300 ns or less, 200 ns or less, or 150 ns or less). In some embodiments, the pulse is square, rectangular, triangular, sawtooth, or sinusoidal. In some embodiments, the pulse output is two-stage (GND and POS), three-stage (GND, NEG, POS), five-stage, or any other multi-stage (e.g., when using a 24-bit DAC). In some embodiments, the pulse is transmitted continuously from the interrogator during operation.In some embodiments, when pulses are transmitted continuously from the interrogator, some of the transducers on the interrogator are configured to receive ultrasound and some of the transducers on the interrogator are configured to transmit ultrasound. The transducers configured to receive ultrasound and the transducers configured to transmit ultrasound may be on the same transducer array in the interrogator or on different transducer arrays. In some embodiments, a transducer on the interrogator may be configured to alternately transmit and receive ultrasound. For example, a transducer may cycle through periods of transmitting one or more pulses and periods of pausing. The transducer may be configured to transmit ultrasound when transmitting one or more pulses, and then switch to a receive mode during the periods of pausing. In some embodiments, the one or more pulses in the cycle include about 1 to about 10 (e.g., about 2 to about 8, or about 4 to about 7, or about 6) ultrasound pulses in any cycle. In some embodiments, the one or more pulses in the cycle include about 1 or more, 2 or more, 4 or more, 6 or more, 8 or more, or 10 or more ultrasonic pulses in any cycle. In some embodiments, the one or more pulses in the cycle include about 20 or less, about 15 or less, about 10 or less, about 8 or less, or about 6 or less in the cycle. The pulse cycle may be repeated regularly, for example, about every 50 microseconds (μs) to about 300 μs (e.g., about every 75 μs to about 200 μs, or about every 100 μs) during operation. In some embodiments, the cycle is repeated every 50 μs or more, every 100 μs or more, every 150 μs or more, every 200 μs or more, every 250 μs or more, or every 300 μs or more. In some embodiments, the cycle is repeated every 300 μs or less, every 250 μs or less, every 200 μs or less, every 150 μs or less, or every 100 μs or less. The cycle frequency may be set based on, for example, the distance between the interrogator and the implanted device and / or how quickly the transducer can switch between transmit and receive modes.

[0097] FIG. 5 illustrates a cycle of pulse-echo ultrasound communication between an interrogator and an implantable device. FIG. 5A illustrates a series of pulse cycles at a frequency of 100 microseconds. During the transmission of the pulses, the transducers in the transducer array are configured to transmit ultrasound. After the pulses are transmitted, the transducers are configured to receive backscattered ultrasound. FIG. 5B illustrates a close-up of the cycle, showing six ultrasound pulses at a frequency of 540 nanoseconds. The backscattered ultrasound detected by the interrogator is shown in FIG. 5C. A close-up of one of the pulses is shown in FIG. 5D. As shown in FIG. 5D, the ultrasound backscatter received by the implantable device can be analyzed. The analysis includes filtering the backscattered waves (e.g., removing attenuation of the sound waves), rectifying the backscattered waves, and integrating the sound waves to determine data encoded by the sound waves. In some embodiments, the backscattered waves are analyzed using machine learning algorithms. FIG. 5E illustrates a close-up of the filtered backscattered waves. The backscattered wave shown in FIG. 5E includes four distinct regions corresponding to reflections occurring from mechanical interfaces: (1) reflections from the biocompatible material encapsulating the implantable device, (2) reflections from the top surface of the miniaturized ultrasound transducer, (3) reflections from the interface between the printed circuit board and the miniaturized ultrasound transducer, and (4) reflections from the back surface of the printed circuit board. The amplitude of the backscattered wave reflected from the surface of the miniaturized transducer changes as a function of the change in impedance of the current flowing through the miniaturized ultrasound transducer and can be referred to as "responsive backscatter" because this region of backscatter encodes information about the detected electrophysiological pulse. The other region of ultrasound backscatter can be referred to as "non-responsive backscatter" and is useful for determining the location of the implantable device, movement of the implantable device, and / or temperature changes near the implantable device, as described below. In some embodiments, the device further includes a non-responsive reflector. In some embodiments, the non-responsive reflector is a cube. In some embodiments, the non-responsive reflector includes silicon.In some embodiments, the non-responsive reflector is a surface of a rigid material that is attached to the implantable device but is electrically isolated and can reflect ultrasound waves that are not responsive to changes in impedance of electrical current due to, for example, a detected electrophysiological pulse.

[0098] The frequency of the ultrasound transmitted by the transducer may be set depending on the drive frequency or resonant frequency of the miniaturized ultrasound transducer on the implantable device. In some embodiments, the miniaturized ultrasound transducer is a broadband device. In some embodiments, the miniaturized ultrasound transducer is narrowband. For example, in some embodiments, the frequency of the pulse is within about 20%, within about 15%, within about 10%, within about 5% of the resonant frequency of the miniaturized ultrasound transducer. In some embodiments, the pulse is set to approximately the same frequency as the resonant frequency of the miniaturized ultrasound transducer. In some embodiments, the frequency of the ultrasound is about 100 kHz to about 100 MHz (e.g., about 100 kHz to about 200 kHz, about 200 kHz to about 500 kHz, about 500 kHz to about 1 MHz, about 1 MHz to about 5 MHz, about 5 MHz to about 10 MHz, about 10 MHz to about 25 MHz, about 25 MHz to about 50 MHz, or about 50 MHz to about 100 MHz). In some embodiments, the frequency of the ultrasound is about 100 kHz or more, about 200 kHz or more, about 500 kHz or more, about 1 MHz or more, about 5 MHz or more, about 10 MHz or more, about 25 MHz or more, or about 50 MHz or more. In some embodiments, the frequency of the ultrasound is about 100 MHz or less, about 50 MHz or less, about 25 MHz or less, about 10 MHz or less, about 5 MHz or less, about 1 MHz or less, about 500 kHz or less, or about 200 kHz or less. A higher frequency can allow for a smaller miniaturized ultrasound transducer on the implantable device. However, a higher frequency limits the depth of communication between the ultrasound transducer and the implantable device. In some embodiments, the implantable device and the ultrasound transducer are separated by about 0.1 cm to about 15 cm (e.g., about 0.5 cm to about 10 cm, or about 1 cm to about 5 cm). In some embodiments, the implantable device and the ultrasound transducer are separated by about 0.1 cm or more, about 0.2 cm or more, about 0.5 cm or more, about 1 cm or more, about 2.5 cm or more, about 5 cm or more, about 10 cm or more, or about 15 cm or more.In some embodiments, the implantable device and the ultrasound transducer are separated by about 20 cm or less, about 15 cm or less, about 10 cm or less, about 5 cm or less, about 2.5 cm or less, about 1 cm or less, or about 0.5 cm or less.

[0099] In some embodiments, the backscattered ultrasound is digitized by the implantable device. For example, the implantable device may include an oscilloscope or an analog-to-digital converter (ADC) and / or storage device, which may digitally encode information about the detected electrophysiological signal. The digitized current fluctuations reflect the electrophysiological signal sensed by the implantable device and are received by an ultrasound transducer. The ultrasound transducer then transmits the digitized sound waves. The digitized data may be compressed using, for example, singular value decomposition (SVD) and least squares-based compression of the analog data. In some embodiments, the compression is performed by a correlator or pattern detection algorithm. The backscattered signal may undergo a series of nonlinear transformations, such as 4th order Butterworth bandpass filter rectification integration of the backscattered region, to form a reconstructed data point at a single time. Such transformations may be performed in hardware (i.e., hard-coded) or software.

[0100] In some embodiments, the digitized data may include a unique identifier. The unique identifier is useful, for example, in a system including multiple implantable devices and / or implantable devices with multiple electrode pairs. For example, the unique identifier can identify the implantable device that is the source of the signal among the multiple implantable devices. In some embodiments, the implantable device includes multiple electrode pairs, which may simultaneously or alternately receive the electrophysiological signal sensed by one implantable device. Different electrode pairs may be configured to sense electrophysiological signals in, for example, different tissues (e.g., different nerves or different muscles) or different sites in the same tissue. The digitizing circuitry may encode the unique identifier to identify which electrode pair sensed the electrophysiological signal.

[0101] In some embodiments, the digitized signal compresses the size of the analog signal. Because the digitized signal has a reduced size, it can more efficiently report the detected electrophysiological signal encoded in the ultrasound backscatter. This can be useful, for example, when an implantable device includes multiple electrode pairs that simultaneously or nearly simultaneously detect electrophysiological signals. By compressing the size of the electrophysiological signal through digitization, potentially overlapping signals can be accurately transmitted.

[0102] In some embodiments, the interrogator communicates with multiple implanted devices. The communication can be performed, for example, using multiple-input multiple-output (MIMO) system theory. For example, the communication between the interrogator and the multiple implanted devices uses time division multiplexing, spatial multiplexing, or frequency multiplexing. In some embodiments, two or more (e.g., three, four, five, six, seven, eight, nine, ten or more, twelve or more, about fifteen or more, about twenty or more, about twenty or more, about twenty or more, about fifteen or more, about fifty or more, or about one hundred or more) implanted devices communicate with the interrogator. In some embodiments, about two hundred or less implanted devices (e.g., about fifty or less, about one hundred or less, about fifty or less, about twenty or less, about fifteen or less, about twelve or less, or about ten or less implanted devices) communicate with the interrogator. The interrogator can receive the combined backscatter from the multiple implanted devices. The combined backscatter is deconvolved, allowing information from each implantable device to be extracted. In some embodiments, the interrogator focuses the transmitted ultrasound from the transducer array to a particular implantable device via beam steering. The interrogator focuses the transmitted ultrasound to a first implantable device and receives backscatter from the first implantable device, and focuses the transmitted ultrasound to a second implantable device and receives backscatter from the second implantable device. In some embodiments, the interrogator transmits ultrasound to and receives ultrasound from multiple implantable devices.

[0103] In some embodiments, the interrogator is used to determine the location or velocity of the implanted device. For example, the velocity can be determined by measuring the location or movement of the device over a period of time. The location of the implanted device can be a relative location, for example, relative to a transducer on the interrogator. Knowing the location or movement of the implanted device can provide a precise location of the detected electrophysiological signal in the tissue. By measuring the location of the implanted device and correlating the detected electrophysiological signal with the location, the tissue can be characterized or monitored at a more localized location. Multiple transducers can collect backscatter from one implanted device. The multiple transducers on the interrogator can be located on the same transducer array or on two or more different transducer arrays. The location of the implanted device can be determined based on the difference between the waveform of the backscattered waves generated from the same implanted device and the previously known location of each transducer. This can be done, for example, by triangulation or clustering and maximum likelihood. The difference in backscatter can be based on responsive backscattered waves, non-responsive backscattered waves, or a combination thereof.

[0104] In some embodiments, the interrogator is used to track the movement of the implantable device. The movement of the implantable device that can be tracked by the interrogator includes lateral and angular movement. Such movement can occur due to a change in the position of one or more organs such as the liver, stomach, small or large intestine, kidneys, pancreas, gallbladder, bladder, ovaries, uterus, or spleen, bone, or cartilage (which can be the result of breathing or movement of the subject, for example), or a change in blood flow (e.g., due to a pulse). The movement of the implantable device can be tracked, for example, by monitoring changes in non-responsive ultrasound. In some embodiments, the movement of the implantable device is determined by comparing the relative position of the implantable device at a first time point to the relative position of the implantable device at a second time point. For example, as described above, the position of the implantable device can be determined using multiple transducers (which may be on one transducer array or on two or more transducer arrays) on the interrogator. A first position of the implantable device can be determined at a first time point, and a second position of the implantable device can be determined at a second time point. A movement vector can then be determined based on the first position at the first time and the second position at the second time.

[0105] <Embedded Devices> The implantable device includes a miniaturized ultrasound transducer (e.g., a miniaturized piezoelectric transducer, a capacitive micromachined ultrasound transducer (CMUT), or a piezoelectric micromachined ultrasound transducer (PMUT)) configured to emit ultrasound backscatter that encodes a detected electrophysiological signal, a backscatter circuit (also called a modulation circuit) configured to modulate a current through the miniaturized ultrasound transducer based on the detected electrophysiological signal, and first and second electrodes configured to sense the electrophysiological signal in tissue. In some embodiments, the implantable device includes an integrated circuit, which may include a modulation circuit, a digital circuit, and / or a power circuit.

[0106] The modulation circuit (or "backscatter circuit") includes a switch (e.g., an on / off switch or a field effect transistor (FET)). An example of a FET for use with some embodiments of the implantable device includes a metal oxide semiconductor field effect transistor (MOSFET). The modulation circuit can change the impedance of the current flowing through a miniaturized ultrasound transducer, and the change in the current flowing through the transducer encodes an electrophysiological signal.

[0107] FIG. 6 illustrates an exemplary implantable device for recording electrophysiological signals. The implantable device includes a miniaturized ultrasound transducer 602, a modulation circuit 604, a first electrode 606, and a second electrode 608. The first electrode 606 and the second electrode 608 are configured to be attached to tissue (e.g., nerve tissue or muscle tissue) to sense electrophysiological signals. The modulation circuit includes a transistor 610, which includes a drain 612, a source 614, and a gate 616. The gate 616 is connected to the first electrode 606. A resistive bridge 618 including a first resistor 620 and a second resistor 622 bridges the drain 612 and the source 614. The second electrode 608 is connected to the resistive bridge 618 between the first resistor 620 and the second resistor 622. The ultrasound transducer 602 includes a first transducer electrode 624 and a second transducer electrode 626. The ultrasonic transducer 602 can receive an ultrasonic carrier from the interrogator, and the ultrasonic transducer generates a current in a circuit. The impedance of the current through the modulation circuit is a function of the gate-to-source voltage, and the impedance is changed by the electrophysiological pulse. The modulated current causes ultrasonic backscatter to be emitted from the transducer 602. The ultrasonic backscatter encodes the electrophysiological pulse.

[0108] In some embodiments, the integrated circuit includes one or more analog circuits that can use power provided by the transducer to power one or more analog amplifiers that can increase the modulation width of the signal modulated on the backscattering impedance.

[0109] In some embodiments, the integrated circuit includes one or more digital circuits. The one or more digital circuits may include a memory device and one or more circuit blocks or systems for operating the implantable device. These systems may have, for example, an on-board microcontroller or processor, a finite state machine implementation, or a digital circuit that can execute one or more programs stored on the implant or provided via ultrasonic communication between the interrogator and the implant. In some embodiments, the digital circuit has an analog-to-digital converter (ADC) that can convert an analog signal to a digital signal. In some embodiments, the digital circuit has a digital-to-analog converter (DAC) that converts the digital signal to an analog signal before directing the signal to the modulator. In some embodiments, the digital circuit operates a modulation circuit (also referred to as a "backscatter circuit"). In some embodiments, the digital circuit transmits a signal to the modulation circuit that encodes the sensed electrophysiological signal. In some embodiments, the digital circuit can operate a modulation circuit (also referred to as a "backscatter circuit"). The digital circuit is connected to a miniaturized ultrasonic transducer. The digital circuit can also operate one or more amplifiers that amplify the current directed to the switch.

[0110] In some embodiments, the digital circuitry encodes the digitized signal, including the electrophysiological signal, with a unique identifier. The unique identifier can identify the implantable device that is the source of the ultrasound backscatter (e.g., in a system with multiple implantable devices) or it can identify which electrode of the implantable device sensed the electrophysiological signal.

[0111] In some embodiments, the digitized circuitry compresses the size of the analog signal. The reduced size of the digitized signal allows for more efficient reporting of detected electrophysiological signals encoded in ultrasound backscatter. This is useful, for example, when an implantable device includes multiple electrode pairs that sense electrophysiological signals simultaneously or near simultaneously. By compressing the size of the electrophysiological signals via digitization, potentially overlapping signals can be accurately transmitted.

[0112] In some embodiments, the integrated circuit filters false electrophysiological signals. In some embodiments, the filtering is performed by digital circuitry. Unfiltered voltage variations can cause changes in the modulated current. The modulated current can be encoded as ultrasound backscatter and recorded as a false positive. To limit false positives, the modulation of the current can be filtered. This filtering can be done, for example, by requiring that the electrophysiological signal must exceed a predefined threshold to cause a modulation of the current through the ultrasound transducer. In some embodiments, a spike detector is used to filter electrophysiological signals that are false positives.

[0113] In some embodiments, the implantable device includes one or more amplifiers, which can, for example, amplify the electrophysiological signal, which can occur, for example, before the signal is transmitted to a digital circuit.

[0114] In some embodiments, the integrated circuit includes a power supply circuit configured to provide power to components of the implantable device. The power supply circuit may include, for example, an identifier, a charge pump, and / or an energy storage capacitor. In some embodiments, the energy storage capacitor is included as a separate component. Ultrasound induces a voltage difference in the miniaturized ultrasound transducer to provide power to the implantable device. The ultrasound may be controlled by the power supply circuit.

[0115] In some embodiments, the implantable device includes multiple electrodes. In some embodiments, the electrodes are paired. An electrode pair can be formed from two electrodes. Thus, an implantable device with three electrodes can have three electrode pairs. An electrophysiological signal can be sensed between the electrodes in the electrode pair. In some embodiments, the implantable device includes one, two, three, four, five, six, seven, eight, nine, ten or more, or fifteen or more electrode pairs. In some embodiments, the implantable device includes two, three, five, six, seven, eight, nine, ten or more electrodes. In some embodiments, the implantable device includes an even number of electrodes, and in some embodiments, an odd number of electrodes. In some embodiments, the small implantable device includes a multiplexer. The multiplexer can select an electrode in the electrode pair for sensing an electrophysiological signal.

[0116] Two or more electrodes communicate with (or are attached to) tissue (e.g., nerve tissue or muscle tissue). The electrodes do not have to be arranged linearly along the tissue. For example, the electrodes may be attached to the nerve along a transverse axis to the nerve, and the electrodes can be used to sense electrophysiological signals flowing in the transverse axis. The two or more electrodes can be attached to the nerve along the transverse axis at any angle, such as diametrically opposed (i.e., 180°) or less than 180° (about 170° or less, about 160° or less, about 150° or less, about 140° or less, about 130° or less, about 120° or less, about 110° or less, about 100° or less, about 90° or less, about 80° or less, about 70° or less, about 60° or less, about 50° or less, about 40° or less, or about 30° or less, etc.).

[0117] In some embodiments, each electrode in the electrode pair is separated by about 5 mm or less (e.g., about 4 mm or less, about 3 mm or less, about 2 mm or less, about 1.5 mm or less, about 1 mm or less, or about 0.5 mm, etc.). In some embodiments, each electrode in the electrode pair is separated by about 0.5 mm or more (e.g., about 1 mm or more, about 1.5 mm or more, about 2 mm or more, about 3 mm or more, or about 4 mm or more, etc.). In some embodiments, each electrode is separated by about 0.5 mm to about 1 mm, about 1 mm to about 1.5 mm, about 1.5 mm to about 2 mm, about 2 mm to about 3 mm, about 3 mm to about 4 mm, or about 4 mm to about 5 mm.

[0118] FIG. 7A illustrates one embodiment of a miniaturized ultrasound transducer (shown as "piezoelectric") connected to an ASIC. The ASIC includes a power supply circuit, and a modulation circuit (or "backscatter circuit"). The power supply circuit includes an energy storage capacitor ("cap"). The electrodes may be implanted in tissue.

[0119] FIG. 7B illustrates an implantable device having a miniaturized ultrasound transducer, an integrated circuit, and a first electrode and a second electrode. The integrated circuit includes a modulation circuit configured to receive a signal based on a detected electrophysiological signal and modulate a current through the ultrasound transducer based on the received signal. The integrated circuit further includes an AC-DC power supply circuit including a full-wave rectifier and a voltage doubler, as well as components for referencing or regulating the supplied power. FIG. 7C illustrates an exemplary identifier that can be used with the implantable device. FIG. 7D illustrates an exemplary structure for an AC-coupled amplifier. The electrophysiological signal ("V neural " ) is detected using electrodes and amplified by an amplifier before the signal is sent to a modulation circuit.

[0120] The implantable device is miniaturized, allowing for comfortable and long-term implantation while limiting the tissue inflammation often associated with the implantable device. The body forms the core of the miniaturized implantable device (e.g., ultrasound transducer and integrated circuitry), with electrodes branching out from the body and attached to tissue to sense electrophysiological signals. In some embodiments, the length of the largest dimension of the device is about 5 mm or less, about 4 mm or less, about 3 mm or less, about 2 mm or less, about 1 mm or less, about 0.5 mm or less, or about 0.3 mm or less. In some embodiments, the length of the largest dimension of the implantable device or the body of the implantable device is about 0.2 mm or more, about 0.5 mm or more, about 1 mm or more, about 2 mm or more, or about 3 mm or more. In some embodiments, the length of the largest dimension of the implantable device or the body of the implantable device is about 0.2 mm to about 5 mm, about 0.3 mm to about 4 mm, about 0.5 mm to about 3 mm, about 1 mm to about 3 mm, or about 2 mm.

[0121] In some embodiments, the one or more electrodes are on the body of the device, for example, on a pad on the body of the device. In some embodiments, the one or more electrodes extend from the body of the implantable device at a desired length and can be implanted at any depth in tissue. In some embodiments, the length of the electrodes is about 0.1 mm or more, for example, about 0.2 mm or more, about 0.5 mm or more, about 1 mm or more, about 5 mm or more, or about 10 mm or more. In some embodiments, the length of the electrodes is about 15 mm or less, for example, about 10 mm or less, about 5 mm or less, about 1 mm or less, or about 0.5 mm or less. In some embodiments, the first electrode is exposed from the body of the implantable device and the second electrode extends from the body of the implantable device.

[0122] In some embodiments, the implantable device has a volume of about 5 mm 3 Less than (for example, about 4 mm 3 Below, 3mm 3 Below, 2mm 3 Less than or equal to 1mm 3 In certain embodiments, the implantable device has a volume of about 0.5 mm 3 From about 5 mm 3 , about 1mm 3 From about 5 mm 3 , about 2 mm 3 From about 5 mm 3 , about 3mm 3 From about 5 mm 3 , or about 4 mm 3 From about 5 mm 3 The small size of the implantable device allows it to be implanted using a biopsy needle.

[0123] In some embodiments, the implantable device is implanted in a subject. The subject may be, for example, a vertebrate, such as a mammal. In some embodiments, the subject is a human, a dog, a cat, a horse, a cow, a pig, a sheep, a goat, a chicken, a monkey, a rat, or a mouse.

[0124] In some embodiments, the implantable device or a portion of the implantable device (e.g., the miniaturized ultrasound transducer and the integrated circuit) is encapsulated by a biocompatible material (e.g., a biocompatible polymer), such as a copolymer of N-vinyl-2-pyrrolidone (NVP) and n-butyl methacrylate (BMA), polydimethylsiloxane (PDMS), parylene, polyimide, silicon nitride, silicon dioxide, silicon carbide, alumina, niobium, or hydroxyapatite. The silicon carbide may be amorphous or crystalline silicon carbide. The biocompatible material is preferably water impermeable to avoid damaging or interfering with the electronic circuitry within the device. In some embodiments, the implantable device or a portion of the implantable device is encapsulated by a ceramic (e.g., alumina or titania) or a metal (e.g., steel or titanium). In some embodiments, the electrodes or portions of the electrodes are not encapsulated by a biocompatible material.

[0125] In some embodiments, the miniaturized ultrasonic transducer and the ASIC are disposed on a printed circuit board (PCB). Electrodes may optionally be disposed on the PCB or may be connected to an integrated circuit. Figures 8A and 8B show an exemplary configuration of an implantable device having a PCB. Figure 8A shows a piezoelectric transducer 802 and an ASIC 804 disposed on a first side 806 of a PCB 808. A first electrode 810 and a second electrode 812 are disposed on a second side 814 of the PCB 808. Figure 8B shows a piezoelectric transducer 814 on a first side 816 of a PCB 818 and an ASIC 820 on a second side 822 of the PCB 818. A first electrode 824 originates on the first side 816 of the PCB and a second electrode 826 originates on the second side 822 of the PCB 818. A first electrode 824 and a second electrode 826 may extend from the PCB 818 and be configured to be electrically connected to each other through tissue.

[0126] The miniaturized ultrasound transducer in the implantable device may be a micromachined ultrasound transducer (e.g., a capacitive micromachined ultrasound transducer (CMUT) or a piezoelectric micromachined ultrasound transducer (PMUT)) or may be a bulk piezoelectric transducer. The bulk piezoelectric transducer may be any natural or synthetic material (e.g., quartz, ceramic, or polymer). Examples of bulk piezoelectric transducer materials include barium titanate (BaTiO3), lead zirconate titanate (PZT), zinc oxide (ZO), aluminum nitride (AlN), quartz, berlinite (AlPO4), topaz, langasite (La3Ga5SiO4), and langasite (La3Ga5SiO4). 14 ), gallium orthophosphate (GaPO4), lithium niobate (LiNbO3), lithium tantalate (LiTaO3), potassium niobate (KNbO3), sodium tungstate (Na2WO3), bismuth ferrite (BiFeO3), polyvinylidene fluoride (PVDF), and lead magnesium niobate-lead titanate (PMN-PT).

[0127] In some embodiments, the miniaturized bulk piezoelectric transducer is approximately cubic (i.e., an aspect ratio of about 1:1:1 (length:width:height)). In some embodiments, the piezoelectric transducer is plate-like, with an aspect ratio of about 5:5:1 or more, such as about 7:5:1 or more, or about 10:10:1 or more, in either the length or width aspect. In some embodiments, the miniaturized bulk piezoelectric transducer is long and thin, with an aspect ratio of about 3:1:1 or more, with the largest dimension aligned with the direction of propagation of the ultrasonic carrier wave. In some embodiments, one dimension of the bulk piezoelectric transducer is equal to one-half the wavelength (λ) corresponding to the driving frequency or resonant frequency of the transducer. At the resonant frequency, ultrasonic waves impinging on both faces of the transducer are phase shifted by 180° to reach the opposite phase, creating a maximum offset between the two faces. In some embodiments, the height of the piezoelectric transducer is about 10 μm to about 1000 μm (e.g., about 40 μm to about 400 μm, about 100 μm to about 250 μm, about 250 μm to about 500 μm, or about 500 μm to about 1000 μm). In some embodiments, the height of the piezoelectric transducer is about 5 mm or less (e.g., about 4 mm or less, about 3 mm or less, about 2 mm or less, about 1 mm or less, about 500 μm or less, about 400 μm or less, about 250 μm or less, about 100 μm or less, or about 40 μm or less). In some embodiments, the height of the piezoelectric transducer is about 20 μm or more (e.g., about 40 μm or more, about 100 μm or more, about 250 μm or more, about 400 μm or more, about 500 μm or more, about 1 mm or more, about 2 mm or more, about 3 mm or more, or about 4 mm or more).

[0128] In some embodiments, the ultrasonic transducer has a length in its greatest dimension of about 5 mm or less (e.g., about 4 mm or less, about 3 mm or less, about 2 mm or less, about 1 mm or less, about 500 μm or less, about 400 μm or less, about 250 μm or less, about 100 μm or less, or about 40 μm or less). In some embodiments, the ultrasonic transducer has a length in its greatest dimension of about 20 μm or more (e.g., about 40 μm or more, about 100 μm or more, about 250 μm or more, about 400 μm or more, about 500 μm or more, about 1 mm or more, about 2 mm or more, about 3 mm or more, or about 4 mm or more).

[0129] The miniaturized ultrasound transducer is connected with two electrodes. A first electrode is attached to a first surface of the transducer and a second electrode is attached to a second surface of the transducer, the first surface and the second surface being opposite sides of the transducer along one dimension. In some embodiments, the electrodes comprise silver, gold, platinum, platinum black, poly(3,4-ethylenedioxythiophene (PEDOT), a conductive polymer (e.g., conductive PDMS or polyimide), or nickel. In some embodiments, the transducer operates in a shear mode in which the axis between the metallic surfaces (i.e., electrodes) of the transducer is perpendicular to the motion of the transducer.

[0130] In some embodiments, the implantable device is configured to be attached to neural tissue. In some embodiments, the attachment to the neural tissue does not completely surround the neural tissue. In some embodiments, the implantable device is embedded in tissue (e.g., neural tissue or muscle tissue). That is, the tissue may completely surround the implantable device. In some embodiments, the neural tissue is part of the central nervous system, such as the brain (e.g., the cerebral cortex, basal ganglia, midbrain, medulla oblongata, pons, hypothalamus, thalamus, cerebellum, vulva, or hippocampus, etc.) or spinal cord. In some embodiments, the attachment to brain tissue includes the body of the implantable device being positioned outside the tissue while the electrodes are embedded in the tissue. In some embodiments, the neural tissue is part of the peripheral nervous system, such as a peripheral nerve.

[0131] In some embodiments, the implantable device is used to detect epileptic activity. See, e.g., Mohseni et al., Guest editorial: Closing the loop via advanced neurotechnologies, IEEE Transactions on Neural Systems and Rehabilitation Engineering, vol. 20, no. 4, pp. 407-409 (2012). In some embodiments, the implantable device is used to optimize cochlear implants. See, e.g., Krook-Magnuson et al., Neuroelectronics and biooptics: Closed-loop technologies in neurological disorders, JAMA Neurology, vol. 72, no. 7, pp. 823-829 (2015).

[0132] In some embodiments, the implantable device is attached to a muscle, such as a skeletal, smooth, or cardiac muscle. In some embodiments, electrodes from the implantable device are attached to a muscle, such as a skeletal, smooth, or cardiac muscle.

[0133] <[Implantable device manufacturing] The implantable device can be fabricated by attaching a miniaturized ultrasound transducer (e.g., a bulk piezoelectric transducer, CMUT, or PMUT) to a first electrode on a first side of the piezoelectric transducer and a second electrode on a second side of the transducer (the first and second sides being on opposite sides of the transducer). The first and second electrodes can be attached to an integrated circuit, which can be disposed on a printed circuit board (PCB). The integrated circuit can include a modulation circuit. Two or more electrodes are also attached to the integrated circuit and configured to be electrically connected to each other in the tissue. Attachment of the components to the PCB can include, for example, wire bonding, soldering, flip chip bonding, or gold bump bonding.

[0134] Certain piezoelectric materials, such as metallized PZT sheets of various thicknesses, can be purchased (e.g., PSI-5A4E, Piezo Systems, Woburn, MA, or PZT 841, APC Internationals, Mackeyville, PA). In some embodiments, the sheet of piezoelectric material is diced to the desired size and the diced piezoelectric material is attached to electrodes. In some embodiments, electrodes are attached to the sheet of piezoelectric material and the sheet of piezoelectric material is diced to the desired size with electrodes attached to the piezoelectric material. The piezoelectric material can be diced with a dicing saw having a ceramic blade to cut the sheet of piezoelectric material into individual piezoelectric transducers. In some embodiments, a laser cutter is used to diced or singulate the piezoelectric material. In some embodiments, a pattern etch is used to diced or singulate the piezoelectric material.

[0135] Electrodes can be attached to the top and bottom surfaces of the piezoelectric transducer. The distance between the electrodes is defined as the height of the piezoelectric transducer. Exemplary electrodes can include silver, gold, platinum, platinum black, poly(3,4-ethylenedioxythiophene (PEDOT), conductive polymers (e.g., conductive PDMS or polyimide), or nickel. In some embodiments, the electrodes are attached to the piezoelectric transducer by electroplating or vacuum depositing the electrode material onto the surface of the piezoelectric transducer. In some embodiments, the electrodes are soldered to the piezoelectric transducer using an appropriate solder and flux. In some embodiments, the electrodes are attached to the piezoelectric transducer using an epoxy (e.g., silver epoxy) or low temperature soldering (e.g., with solder paste).

[0136] In an exemplary embodiment, solder paste is applied to the pads on a printed circuit board (PCB) before or after attachment of the integrated circuit to the PCB. The size of the pads on the circuit board depends on the desired size of the piezoelectric transducer. By way of example only, if the desired size of the piezoelectric transducer is approximately 100 μm×100 μm×100 μm, the pads can be approximately 100 μm×100 μm. The pads serve as first electrodes for the implantable device. A piezoelectric material (which may be larger than the pads) is placed on the pads and secured to the pads with applied solder paste to form a piezoelectric PCB assembly. The piezoelectric PCB assembly is heated to harden the solder paste, thereby bonding the piezoelectric transducer to the PCB. If the piezoelectric material is larger than the pads, the piezoelectric material is cut to the desired size, for example, using a wafer dicing saw or a laser cutter. The non-bonded portion of the piezoelectric material (e.g., the portion of the piezoelectric material that does not overlie the pad) is removed. A second electrode is attached to the piezoelectric transducer and the PCB. This is done, for example, by forming a wire bond between the top of the piezoelectric transducer and the PCB, which completes the circuit. The wire bond can be made with a wire made of any conductive material (e.g., aluminum, copper, silver, or gold).

[0137] The integrated circuit and the miniaturized ultrasound transducer may be attached to the same side of the PCB or to opposite sides of the PCB. In some embodiments, the PCB is a flexible PCB and the integrated circuit and the miniaturized ultrasound transducer are attached to the same side of the PCB and the PCB is folded to obtain an implantable device with the integrated circuit and the miniaturized ultrasound transducer on opposite sides of the PCB.

[0138] Optionally, the device or a portion of the device is encapsulated or encapsulated with a biocompatible material (e.g., a biocompatible polymer). The biocompatible material can be, for example, a copolymer of N-vinyl-2-pyrrolidone (NVP) and n-butyl methacrylate (BMA), polydimethylsiloxane (PDMS, e.g., Sylgard 184, Dow Corning, Midland, MI), parylene, polyimide, silicon nitride, silicon dioxide, alumina, niobium, hydroxyapatite, or silicon carbide. The silicon carbide can be amorphous or crystalline silicon carbide. In some embodiments, the biocompatible material (e.g., amorphous silicon carbide) is applied to the device by plasma enhanced chemical vapor deposition (PECVD) or sputtering. PECVD can use precursors such as SiH4 or CH4 to produce silicon carbide. In some embodiments, the implantable device or a portion of the implantable device is encased in a ceramic (eg, alumina or titania) or metal (eg, steel or titanium) suitable for medical implantation.

[0139] FIG. 9 illustrates an exemplary method of manufacturing an implantable device described herein. At step 902, an application specific integrated circuit (ASIC) is attached to a PCB. The PCB may include two or more electrodes for sensing electrophysiological signals in tissue. Solder (e.g., silver epoxy) can be applied to the PCB (e.g., at a first pad located on the PCB) and the ASIC can be placed on the solder. The solder can be cured, for example, by heating the PCB with the ASIC. In some embodiments, the PCB with the ASIC is heated to about 50° C. to about 200° C., e.g., about 80° C. to about 170° C., or to about 150° C. In some embodiments, the PCB with the ASIC is heated for about 5 minutes to about 600 minutes, e.g., about 10 minutes to about 300 minutes, about 10 minutes to about 100 minutes, about 10 minutes to about 60 minutes, about 10 minutes to about 30 minutes, or about 15 minutes. Optionally, the ASIC is coated with additional solder. At step 904, the piezoelectric transducer ("Piezoelectric" in FIG. 9) is attached to the PCB. Solder (e.g., silver epoxy) can be applied to the PCB (e.g., at the second pads located on the PCB) and the piezoelectric material can be placed on the solder. The piezoelectric material can be a fully formed (i.e., "diced") piezoelectric transducer or a sheet of piezoelectric material that is cut to form the piezoelectric transducer once attached to the PCB. The solder can be cured, for example, by heating the PCB with the piezoelectric material. In some embodiments, the PCB with the piezoelectric material is heated to about 50° C. to about 200° C., for example, about 80° C. to about 170° C., or to about 150° C. In some embodiments, the PCB with the piezoelectric material is heated for about 5 minutes to about 600 minutes, for example, about 10 minutes to about 300 minutes, about 10 minutes to about 100 minutes, about 10 minutes to about 60 minutes, about 10 minutes to about 30 minutes, or about 15 minutes. The piezoelectric material can be cut to the desired dimensions using a saw or laser cutter. In some embodiments, the piezoelectric material is a sol-gel (e.g., PZT sol-gel) and the transducer material can be shaped by deep reactive ion etching (DRIE).Although FIG. 9 shows the ASIC attached to the PCB in step 902 and then the piezoelectric material attached to the PCB in step 904, one skilled in the art will appreciate that the ASIC and piezoelectric material may be attached in any order. In step 906, the ASIC and piezoelectric transducer are wire bonded to the PCB. Although the method shown in FIG. 9 shows the ASIC and piezoelectric transducer to the PCB after the ASIC and piezoelectric transducer are attached to the PCB, one skilled in the art will appreciate that the ASIC can be wire bonded to the PCB after it is attached to the PCB, and can be wire bonded either before or after the attachment of the piezoelectric transducer. Similarly, the piezoelectric transducer may be wire bonded to the PCB either before or after the attachment or wire bonding of the ASIC to the PCB. In step 908, at least a portion of the device is covered with a biocompatible material. Preferably, at least the piezoelectric transducer and the ASIC are covered with a biocompatible material. In some embodiments, the sensor, or at least a portion of the sensor, is not covered with a biocompatible material. For example, in some embodiments, the implantable device has a set of electrodes that are not covered with a biocompatible material, such that the electrodes can detect electrophysiological signals, hi some embodiments, the biocompatible material is cured, for example by exposure to UV light or by heating.

[0140] In some embodiments, the implantable device or a portion of the implantable device is encapsulated in an amorphous silicon carbide (a-SiC) film. FIG. 10 illustrates a method of manufacturing an implantable device encapsulated in an a-SiC film. In step 1002, a polyimide layer is applied to a smooth surface. In step 1004, an a-SiC layer is applied to the polyimide layer. This can be done, for example, using plasma enhanced chemical vapor deposition (PECVD) using SiH4 and CH4 as precursors. In step 1006, one or more ports are etched into the a-SiC layer. In some embodiments, the ports are also etched into the polyimide layer. The ports provide access to portions of the implantable device that are not encapsulated in a-SiC, for example, portions of sensors or electrodes that will contact tissue after implantation. In some embodiments, the etching comprises reactive ion etching. In step 1008, the implantable device is attached to the a-SiC layer. The implantable device may be prefabricated before attachment to the a-SiC layer or may be fabricated on the a-SiC. In some embodiments, a printed circuit board (PCB), miniaturized ultrasonic transducer, and sensor are attached to the a-SiC layer. The miniaturized ultrasonic transducer and sensor do not need to be in direct contact with the a-SiC layer, as the miniaturized ultrasonic transducer and sensor may be attached to the PCB. The miniaturized ultrasonic transducer or sensor may be attached to the PCB before or after the PCB is attached to the a-SiC layer. In some embodiments, the miniaturized ultrasonic transducer or sensor may be attached to the PCB by wire bonding the miniaturized ultrasonic transducer or sensor to the PCB. In some embodiments, the sensor includes a portion associated with a port etched in the a-SiC layer. In some embodiments, the ASIC is attached to the PCB, which may be before or after the PCB is attached to the a-SiC layer. In step 1010, the exposed portion of the implantable device is covered with an a-SiC layer. In some embodiments, the exposed portion of the implantable device is covered with an a-SiC layer using PECVD.In step 1012, the encapsulated device is lifted off to release the device from the SiC layer.

[0141] Exemplary embodiments Embodiment 1. A first electrode and a second electrode attached to tissue and configured to sense an electrophysiological signal; an integrated circuit having a multi-transistor circuit and a modulation circuit configured to modulate a current based on the electrophysiological signal; an ultrasound transducer configured to emit ultrasound backscatter based on the modulated current, the backscatter encoding the electrophysiological signal from tissue; 16. An implantable device comprising:

[0142] Embodiment 2. The implantable device of embodiment 1, wherein the multi-transistor circuit is a digital circuit.

[0143] Embodiment 3. An implantable device as described in embodiment 1 or 2, comprising a body having the ultrasonic transducer and modulation circuitry, the length of the maximum dimension of the body being about 5 mm or less.

[0144] Embodiment 4. An implantable device described in any one of embodiments 1 to 3, wherein the first electrode and the second electrode are spaced apart by a distance of at least about 0.5 mm.

[0145] Embodiment 5. An implantable device according to any one of embodiments 1 to 4, having a non-responsive reflector.

[0146] Embodiment 6. An implantable device described in any one of embodiments 1 to 5, wherein the tissue is muscle tissue or nerve tissue.

[0147] Embodiment 7. An implantable device described in any one of embodiments 1 to 6, wherein the tissue is part of the peripheral nervous system.

[0148] Embodiment 8. An implantable device described in any one of embodiments 1 to 7, wherein the tissue is part of the central nervous system.

[0149] Embodiment 9. An implantable device described in any one of embodiments 1 to 8, wherein the tissue is brain tissue.

[0150] Embodiment 10. An implantable device described in any one of embodiments 1 to 6, wherein the tissue is skeletal muscle, cardiac muscle, or smooth muscle.

[0151] Embodiment 11. An implantable device described in any one of embodiments 1 to 10 having three or more electrodes.

[0152] Embodiment 12. An implantable device described in any one of embodiments 2 to 11, wherein the digital circuit is configured to operate a modulation circuit.

[0153] Embodiment 13. An implantable device described in any one of embodiments 2 to 12, wherein the digital circuit is configured to transmit a digitized signal to a modulation circuit, the digitized signal being based on a detected electrophysiological signal.

[0154] Embodiment 14. An implantable device as described in embodiment 13, wherein the digitized signal includes a unique implantable device identifier or a unique electrode pair identifier.

[0155] Embodiment 15. An implantable device as described in embodiment 13 or 14, wherein the digitized signal is compressed by more than five times compared to the analog signal.

[0156] Embodiment 16. An implantable device described in any one of embodiments 1 to 15, wherein the integrated circuit comprises a spike detector.

[0157] Embodiment 17. An implantable device described in any one of embodiments 1 to 16, wherein the integrated circuit has a power supply circuit.

[0158] Embodiment 18. An embedded device described in any one of embodiments 1 to 17, wherein the integrated circuit has an analog-to-digital converter (ADC).

[0159] Embodiment 19. An implantable device described in any one of embodiments 1 to 18, wherein the integrated circuit has an amplifier array.

[0160] Embodiment 20. An implantable device described in any one of embodiments 1 to 20, wherein the modulation circuit has a switch.

[0161] Embodiment 21. An implantable device as described in any one of embodiments 1 to 21, wherein the modulation circuit comprises a field effect transistor (FET).

[0162] Embodiment 22. An implantable device described in any one of embodiments 1 to 21, wherein the length of the maximum dimension of the ultrasonic transducer is about 5 mm or less.

[0163] Embodiment 23. The volume of the body is about 5 mm 3 An implantable device according to any one of embodiments 1 to 22, which is as follows:

[0164] Embodiment 24. The volume of the body is about 1 mm 3 An implantable device according to any one of embodiments 1 to 23, which is as follows:

[0165] Embodiment 25. An implantable device described in any one of embodiments 1 to 24, wherein the ultrasonic transducer is configured to receive ultrasonic waves that power the implantable device.

[0166] Embodiment 26. An implantable device described in any one of embodiments 1 to 25, wherein the ultrasonic transducer is configured to receive ultrasonic waves from a calling device having one or more ultrasonic transducers.

[0167] Embodiment 27. An implantable device described in any one of embodiments 1 to 26, wherein the ultrasonic transducer is a bulk piezoelectric transducer.

[0168] Embodiment 28. An implantable device as described in embodiment 27, wherein the bulk ultrasound transducer is approximately cubic.

[0169] Embodiment 29. An implantable device described in any one of embodiments 1 to 26, wherein the ultrasonic transducer is a piezoelectric micromachined ultrasonic transducer (PMUT) or a capacitive micromachined ultrasonic transducer (CMUT).

[0170] Embodiment 30. An implantable device described in any one of embodiments 1 to 29, which is implanted in a subject.

[0171] Embodiment 31. The implantable device of embodiment 30, wherein the subject is a human.

[0172] Embodiment 32. An implantable device described in any one of embodiments 1 to 31, which is implanted in tissue.

[0173] Embodiment 33. An implantable device described in any one of embodiments 1 to 31, located on or near tissue.

[0174] Embodiment 34. An implantable device as described in embodiment 33, wherein the first electrode and the second electrode are implanted in tissue.

[0175] Embodiment 35. An implantable device as described in embodiment 33, wherein the first electrode and the second electrode are on tissue.

[0176] Embodiment 36. An implantable device described in any one of embodiments 1 to 35, which is at least partially encapsulated by a biocompatible material.

[0177] Embodiment 37. An implantable device as described in embodiment 36, wherein at least a portion of the first electrode and the second electrode are not encapsulated by a biocompatible material.

[0178] Embodiment 38. An implantable device as described in embodiment 36 or 37, wherein the biocompatible material is a copolymer of N-vinyl-2-pyrrolidone (NVP) and n-butyl methacrylate (BMA), polydimethylsiloxane (PDMS), parylene, polyimide, silicon nitride, silicon dioxide, alumina, niobium, hydroxyapatite, silicon carbide, titania, steel, or titanium.

[0179] Embodiment 39. An implantable device as described in embodiment 36 or 37, wherein the biocompatible material is a ceramic or a metal.

[0180] Embodiment 40. A system having one or more implantable devices described in any one of embodiments 1 to 39, and an interrogator having one or more ultrasonic transducers configured to transmit ultrasound to the one or more implantable devices or receive ultrasound backscatter from the one or more implantable devices.

[0181] Embodiment 41. The system described in embodiment 40, wherein the interrogator has a first ultrasonic transducer configured to transmit ultrasonic waves and a second ultrasonic transducer configured to receive ultrasonic backscatter from the one or more implantable devices.

[0182] Embodiment 42. The system described in embodiment 40 or 41, wherein the interrogator includes two or more separate interrogation devices, a first interrogation device configured to transmit ultrasound to the one or more implantable devices, and a second interrogation device configured to receive ultrasound backscatter from the one or more implantable devices.

[0183] Embodiment 43. A system described in any one of embodiments 40 to 42, wherein the interrogator has two or more ultrasonic transducer arrays, each transducer array having two or more ultrasonic transducers.

[0184] Embodiment 44. A system described in any one of embodiments 40 to 43, wherein at least one of the at least one or more ultrasonic transducers is configured to alternately transmit ultrasonic waves to the one or more implanted devices and receive ultrasonic backscatter from the one or more implanted devices, and the configuration of the transducer is controlled by a switch on the interrogator.

[0185] Embodiment 45. A system described in any one of embodiments 40 to 44 having a plurality of embedded devices.

[0186] Embodiment 46. The system described in embodiment 45, wherein the interrogator is configured to beam steer the transmitted ultrasound to alternately focus the transmitted ultrasound on a first portion of the plurality of implantable devices or a second portion of the plurality of implantable devices.

[0187] Embodiment 47. The system described in embodiment 45, wherein the interrogator is configured to simultaneously receive ultrasonic backscatter from at least two implantable devices.

[0188] Embodiment 48. The system described in embodiment 45, wherein the interrogator is configured to transmit ultrasound to or receive ultrasound backscatter from the multiple implantable devices using time division multiplexing.

[0189] Embodiment 49. The system described in embodiment 45, wherein the interrogator is configured to transmit ultrasound to or receive ultrasound backscatter from the multiple implantable devices using spatial multiplexing.

[0190] Embodiment 50. The system described in embodiment 45, wherein the interrogator is configured to transmit ultrasound to or receive ultrasound backscatter from the multiple implantable devices using frequency multiplexing.

[0191] Embodiment 51. A system described in any one of embodiments 45 to 50, wherein the interrogator is configured to be wearable by the subject.

[0192] Embodiment 52. An interrogator having one or more ultrasonic transducers; one or more processors; a computer-readable non-transitory storage medium storing one or more programs configured to be executed by one or more processors; the one or more programs instructions for detecting an ultrasound backscatter-based electrophysiological signal encoding an electrophysiological pulse emitted from the implantable device; and A computer system comprising instructions for determining a position of the implanted device relative to the one or more ultrasound transducers in the interrogator or detecting movement of the implanted device based on ultrasonic backscatter not responsive to electrophysiological pulses emitted from the implanted device.

[0193] Embodiment 53. A computer system as described in embodiment 52, wherein the one or more programs include instructions for determining the position of the implantable device relative to one or more ultrasonic transducers in the interrogator.

[0194] Embodiment 54. A computer system as described in embodiment 52 or 53, wherein the one or more programs include instructions for detecting movement of the embedded device.

[0195] Embodiment 55. The computer system of embodiment 54, wherein the movement includes a lateral movement.

[0196] Embodiment 56. A computer system as described in embodiment 54 or 55, wherein the movement includes an angular movement.

[0197] Embodiment 57. A computer system described in any one of embodiments 52 to 56, wherein the ultrasound backscatter in response to the electrophysiological signal includes a digitized signal encoding the electrophysiological signal.

[0198] Embodiment 58. A computer system described in any one of embodiments 52 to 57, wherein the one or more programs include instructions for attributing detected electrophysiological signals to an implanted device that emitted ultrasonic backscatter encoding electrophysiological signals from multiple implanted devices.

[0199] Embodiment 59. The computer system of embodiment 58, wherein the sensed electrophysiological signals are attributed to an implantable device using time division multiplexing, spatial multiplexing, or frequency multiplexing.

[0200] Embodiment 60. The computer system of embodiment 58, wherein the detected electrophysiological signal is attributed to the implanted device using a unique identifier encoded in the ultrasound backscatter that encodes the electrophysiological pulse.

[0201] Embodiment 61. A method for detecting an electrophysiological signal in tissue, comprising: receiving ultrasound at one or more implantable devices; converting energy from the ultrasound waves into an electrical current flowing through a modulation circuit; sensing said electrophysiological signal; digitizing the electrophysiological signal; modulating a current based on the digitized electrophysiological signal; converting the modulated current into ultrasonic backscatter that encodes information about the electrophysiological signal; and emitting ultrasonic backscatter to an interrogator having one or more transducers configured to receive the ultrasonic backscatter.

[0202] Embodiment 62. A method for detecting an electrophysiological signal in tissue, comprising: transmitting ultrasound waves from an interrogator having one or more ultrasound transducers to one or more implantable devices; and receiving ultrasound backscatter from one or more implanted devices, the ultrasound backscatter encoding digitized information about the electrophysiological signal.

[0203] Embodiment 63. The method of embodiment 61 or 62, comprising receiving the ultrasonic backscatter using an interrogator.

[0204] Embodiment 64. The method of any one of embodiments 61 to 63, wherein the ultrasound powers the one or more implantable devices.

[0205] Embodiment 65. The method of any one of embodiments 61 to 64, wherein the one or more implantable devices comprises a body having an ultrasonic transducer and a modulation circuit, and the length of the maximum dimension of the body is about 5 mm or less.

[0206] Embodiment 66. The method of any one of embodiments 61 to 65, wherein the one or more implantable devices comprise a first electrode and a second electrode attached to tissue and configured to sense electrophysiological signals, and the first electrode and the second electrode are spaced apart by at least about 0.5 mm.

[0207] Embodiment 67. The method of any one of embodiments 61 to 66, wherein the tissue is muscle tissue or nerve tissue.

[0208] Embodiment 68. The method of any one of embodiments 61 to 67, wherein the tissue is part of the peripheral nervous system.

[0209] Embodiment 69. The method of any one of embodiments 61 to 67, wherein the tissue is part of the central nervous system.

[0210] Embodiment 70. The method of any one of embodiments 61 to 67, wherein the tissue is brain tissue.

[0211] Embodiment 71. The method of any one of embodiments 61 to 67, wherein the tissue is skeletal muscle, cardiac muscle, or smooth muscle.

[0212] Embodiment 72. The method of any one of embodiments 61 to 71, comprising reconstructing an electromyogram, an electroneurogram, an electrocardiogram, a compound action potential, multi-unit activity of multiple neurons, a local field potential, or an action potential.

[0213] Embodiment 73. The method of any one of embodiments 61 to 72, comprising implanting the one or more implantable devices.

[0214] Embodiment 74. The method of any one of embodiments 61 to 73, wherein the subject is a human.

[0215] Embodiment 75. The method of any one of embodiments 61 to 74, comprising a step of measuring the position of the one or more implantable devices.

[0216] Embodiment 76. The method of any one of embodiments 61 to 75, comprising measuring angular or lateral movement of the one or more implantable devices.

[0217] Embodiment 77. The method of any one of embodiments 61 to 76, wherein the digitized signal includes a unique implantable device identifier or a unique electrode pair identifier.

[0218] Embodiment 78. The method of any one of embodiments 61 to 77, wherein the one or more implantable devices comprises a plurality of implantable devices.

[0219] Embodiment 79. The method of any one of embodiments 61 to 78, wherein the electrophysiological signal is detected in vivo.

[0220] [Example] Example 1: Manufacture of an implantable device In abbreviated form, the assembly process for the implantable device is as follows. 1. Mount the ASIC on the PCB. 2. Wire bond the ASIC ports to the PCB. 3. Attach the piezoelectric element to the PCB. 4. Wire bond the piezoelectric element ports to the PCB. 5. Encapsulate the entire device except for the leading electrodes.

[0221] The ASICs are 450 μm×500 μm×500 μm in size and are fabricated by the 65 nm process of Taiwan Semiconductor Manufacturing Company. Each chip has two transistors, each with five ports (source, drain, gate, center, bulk). Each FET uses the same bulk, so any bulk pad can be connected, but the transistors differ in that the transistors with pads exposed on the top row do not have a resistive bias network, while the transistors with pads exposed on the bottom row do. The chips also have smaller pads for electroplating. The same process can be applied to ASICs with more complex circuits and therefore more pads. These pads were not used in this example. Three versions of the FETs were taped out. Die 1: Long channel FET with threshold voltage 500mV. Die 2: Short channel FET with threshold voltage of 500mV. Die 3: Native FET with threshold voltage 0mV.

[0222] The electrical characteristics of these FETs were measured using a specially designed CMOS characterization board that consisted of one set of pads as wire bond targets and a second set of pads to which wires were soldered. A source meter (2400 Sourcemeter, Keithley Instruments, Cleveland, OH) was used to measure the V DS is supplied to the FET and I DS The V was measured using an adjustable power supply (E3631A, Agilent, Santa Clara, CA). GS The I-V characteristics of the FETs were obtained by modulating the I-V curves of the type 2 die. Uncharacteristic I-V curves were consistently measured for the type 2 die, and impedance measurements showed that the short channel of die 2 would short the FET.

[0223] The piezoelectric elements are lead zirconate titanate (PZT). They were purchased as disks from APC International and diced into 750 μm × 750 μm × 750 μm cubes using a wafer saw (DAD3240, Disco, Santa Clara, CA) with a ceramic blade (PN CX-010-'270-080-H). This moat size was chosen because it maximized power transfer efficiency. For further details, see Seo et al., Neural dust: an ultrasonic, low power solution for chronic brain-machine interfaces, arXiv: 1307.2196v1 (July 8, 2013).

[0224] The implantable device was implanted into the sciatic nerve of Long-Evans rats. The nerve is a large diameter nerve bundle that innervates the hind limb. The nerve is 1-1.4 mm in diameter and is an ideal candidate for device implantation due to its size and accessibility. Several iterations of the implantable device have been made, but the following example will discuss the development of two versions implanted in a rat model.

[0225] The implantable device substrate integrates the ASIC with the piezoelectric and the lead electrodes. The first version of the implantable device used a custom PCB purchased from The Boardworks (Oakland, CA) as the substrate. The PCB was made from FR-4 and was 30 mil thick. The dimensions of the substrate were 3 mm × 1 mm. Since this design was the first attempt at a communications-sensing integrated platform, pad sizes and pad spacing were chosen to facilitate assembly at the expense of larger size. To conserve PCB footprint, each side of the PCB had pads for either the piezoelectric or the ASIC and their respective PCB connections. In addition, two lead pads were placed on the ASIC side of the substrate. All exposed electrodes were plated with ENIG from The Boardworks. The pads on which the ASIC was placed were 500 μm × 500 μm and were chosen to match the size of the die. The size of the pads targeted for wire bonding was chosen to be 200 μm × 200 μm, and was spaced approximately 200 μm from the edge of the die to allow sufficient clearance for wire bonding (described below). Electrode size and spacing were varied and empirically optimized using four electrode pairs spaced 2 mm, 1.5 mm, 1 mm, and 0.5 mm apart from each other. Electrode pairs spaced greater than 1.5 mm apart were found to be optimal for derivation. Minimal signal attenuation was observed between 2 mm and 1.5 mm, with approximately a 33% decrease in signal strength between 1.5 mm and 1 mm.

[0226] The second iteration of the implantable device addressed three main concerns: 1) size, 2) ease of wire bonding, and 3) embedding / communication. First, to reduce the thickness of the substrate, the FR-4 substrate was replaced with a 2 mil thick polyimide flexible PCB (AltaFlex, Santa Clara, CA) and the ASIC (Grinding and Dicing Services Inc., San Jose, CA) was thinned down to 100 μm. To facilitate bonding, the ASIC and PZT coupons were moved to the same side, leaving only the derived electrodes on the back side of the substrate. Having the ASIC and PZT coupons on the same side of the substrate constrained the reduction in substrate size, but the spacing between the electrodes limited the substrate length to at least 2 mm. To push the minimization effort, the ASIC bond pads were reduced to 100 μm × 100 μm, but the 200 μm spacing between the bond pads and the ASIC itself had to be maintained to provide space for wire bonding. The mounting pads for the PZT coupons were also shrunk and placed closer to the edge of the substrate. This is because the PZT coupons do not need to sit completely on the substrate, but can extend beyond it. Additionally, the location of the pads relative to the ASIC was modified to facilitate easier bonding. In the original design, the layout of the bonding pads surrounding the ASIC required two wire bonds to cross each other. While not impossible, it is extremely difficult to avoid shorting the pads together. Therefore, the pad layout was modified to allow the bonds to follow a relatively straight path. Finally, during animal testing, it proved extremely difficult to align the implantable devices. To address this, we added four 1-inch test leads extending from the substrate, two of which were connected directly to the source and drain of the device to obtain measurable power, which we used as an alignment metric. The other two leads were connected to the gate and center port to obtain a ground truth signal. The vias were given a unique shape to prevent confusion about which lead belonged to which port. See Figure 11A.

[0227] There was some fear that the test leads might easily break or shift the moat position when force is applied, so a version with a serpentine trace was designed. Serpentine traces (Figure 11B) have often been used to enable deformable interconnects because their structure allows them to "fold out like an accordion". Conceptually, the serpentine trace design is a series of cantilever beams connected in series via connecting beams.

[0228] Along with the design presented above, we also designed and constructed a miniaturized version of the implantable device using both sides of the substrate. In this design, the substrate is approximately 1.5mm x 0.6mm x 1mm in size. Due to the small size of the substrate, a 5mil silver wire "tail" was attached to the device for recording. This version was not tested in vivo.

[0229] The ASIC and PZT coupons were attached to the PCB substrate using an adhesive. There were three main concerns in the selection of the adhesive: 1) the adhesive needed to hold the ASIC and PZT in place tightly enough that the ultrasonic forces from the wire bonds would not shake the components; 2) the application of the adhesive was relatively precise due to the sub-millimeter scale and pitch of the component / substrate pads; and 3) the adhesive needed to be conductive.

[0230] The ASIC and diced PZT were originally attached to the PCB substrate using a low-temperature curing solder paste. Solder paste consists of powdered metal solder suspended as spheres in a flux. When heat is applied, the solder balls begin to melt and fuse together. However, it was found that the curing of the solder paste often causes the PZT coupon or mote to translate and rotate during reflow. This creates problems for PZT placement and power collection, as well as wire bonding, because the bond pads are no longer properly positioned relative to the chip. However, it was found that a two-part silver epoxy (consisting only of silver particles suspended in epoxy) could be cured without changing the position of the chip or the PZT coupon. Therefore, the ASIC and diced PZT were pasted to the PCB using a two-part conductive silver epoxy (H20E, Epotek, Billerica, MA). The PCB was then secured to a glass slide using Kapton tape (Polyimide Film Tape 5413, 3M, St. Paul, MN) and placed in a 150°C convection oven for 15 minutes to cure the epoxy. Although higher temperatures can speed up the cure (Figure 12), care was taken not to heat the PZT above 160°C (half the Curie temperature of the PZT). Heating the PZT at a higher temperature would risk depolarizing the PZT. Curing at 150°C was found to have no effect on the CMOS performance.

[0231] The connections between the top of the PZT and the PCB, and between the ASIC and the PCB, were made by wire bonding 1 mil Al wire using an ultrasonic wedge bonder (740DB, West Bond, Scotts Valley, CA). In this bonding method, the Al wire is threaded through the wedge of the bond head, and ultrasonic energy "rubs" the Al wire against the substrate, creating heat through friction. This heat welds the two materials together.

[0232] Wire bonding to the ASIC was difficult to avoid shorting due to the size of the CMOS pads and the size of the wire bond ends. This problem was exacerbated by the positioning of the wire bond targets on the first version of the embedded device substrate, where the two bond ends were positioned across the short width of the ASIC pad rather than across its length. Thinner gold wire could have been used for bonding, but the difficulty of thermosonic bonding gold with a wedge bonder made it impractical to use gold wire for bonding with this machine. Furthermore, it is important that the substrate be flat and fixed for effective wire bonding. Thus, our original design of having the ASIC and PZT on different sides of the substrate often caused problems during the wire bonding process on the first version of our embedded substrate. For this reason, board design choices made in the second iteration of the embedded device (moving the ASIC and PZT to the same side and repositioning the pads to provide a straighter path for the wirebond targets) significantly increased the wirebond yield.

[0233] Finally, because we used an ultrasonic bonder, we realized that bonding to the PZT would result in a charge build-up that would damage the chip if the PZT were fully bonded to the substrate. To avoid this, we grounded the source and drain test leads of the device just before wire-bonding the PZT.

[0234] The final step in the assembly of the implantable device is encapsulation. This step accomplishes two objectives: 1) it insulates the PZT, bond pads, and ASIC from the aqueous environment; and 2) it protects the wire bonds between the ASIC / PZT coupon and the PCB. At the same time, some method is needed to remove or prevent the encapsulation material from covering the lead-out electrodes. Furthermore, the encapsulation material must not interfere with the embedding of the device. Finally, it is important, though not essential, to choose an encapsulation material that is visually transparent so that the device can be inspected for physical defects in case any damage occurs during encapsulation.

[0235] The first encapsulation material used was Crystalbond (509', SPI Supplies, West Chester, PA). Crystalbond is an adhesive that is solid at room temperature but begins to soften at 71 °C and melts to a viscous liquid at 121 °C. When the heat is removed from the Crystalbond, it resolidifies within a few minutes and can be well controlled. To encapsulate the implantable devices, a small flake of Crystalbond was scraped off with a razor and placed directly on the device. The substrate was then heated on a hotplate. The temperature was first brought to about 70 °C, where the flake began to deform, and then gradually increased until the Crystalbond was completely liquid. When the edge of the liquid Crystalbond drop had spread past the farthest wirebond but not yet past the lead-out pad, the hotplate was turned off and the substrate was quickly transferred from the plate to a cooling chuck where the Crystalbond was allowed to resolidify.

[0236] Although Crystalbond was effective, a UV-curable epoxide proved to be more selective, biocompatible, and cured more rapidly. First, a photocurable acrylic (3526, Loctite, Dusseldorf; Germany), which cures upon exposure to UV light, was tried. For precision, a sewing needle was used as the applicator and the epoxy was cured with a 405 nm laser point for 2 minutes. This epoxy worked well, but was not medical grade and therefore unsuitable for in vivo implantation. For this reason, a medical grade UV-curable epoxy (OG116-31, EPO-TEK, Billercia, MA) was tried. This epoxy was applied at 365 nm with 92 mW / cm2 light. 2 The epoxy was cured for 5 min in a 1000 nm UV chamber (Flash, Asiga, Anaheim Hills, CA). This epoxy was slightly less viscous than the Loctite epoxy, but selective encapsulation was achieved by again using a sewing needle as the applicator. As an insulation and protection mechanism for the wire bonds, the epoxy was highly effective, but was found to leak when submerged in water for extended periods (~1 h). A second medical grade epoxy was investigated (301-2, EPO-TEK, Billerica, MA) that advertised stability for up to 1 year, but was found to be insufficiently viscous and required baking in an oven to cure. Although the UV epoxy was unstable, its duration of use was adequate for acute in vivo experiments.

[0237] To improve the stability of the encapsulation, Parylene-C was also considered as an encapsulation material. Parylene-C is an FDA-approved biocompatible polymer that is chemically and biologically inert, an excellent barrier and electrical insulator, and highly conformal when deposited. Parylene-C deposition is achieved by vaporizing powdered Parylene-C dimer at temperatures above 150 °C. The gas-phase Parylene-C dimer is then heated at 690 °C to cause pyrolysis, splitting the Parylene-C dimer into monomers. The monomer then fills a chamber that is kept at room temperature. The monomer polymerizes almost instantly once it comes into contact with any surface. For all devices, Parylene-C was deposited using a Parylene deposition system (SCS Labcoter 2 Parylene Deposition System, Specialty Coating Systems, Indianapolis, IN) with the parameters shown in Table 1. Note that Table 1 lists the chamber gauge temperature as 135 °C, which is clear from the actual chamber temperature. Rather, the chamber gauge is simply the vacuum gauge of the process chamber. It is important to keep the temperature at least 135°C to prevent parylene from depositing on the gauge. For the first batch of FR-4 boards, the parylene was selectively treated with Kapton tape to avoid covering the electrodes. However, due to the small pitch of the lead-out electrodes and the ASIC wire-bonding targets, there was not enough surface area for the tape to stick well to the board, and the tape would often slide off, resulting in covering the electrode pads. For the second iteration of the implantable device, we attempted to coat the parylene using a method that involved coating the entire board and then removing the parylene from the electrodes with the tip of a probe. The implantable device was wetted and suspended in air between two layers of glass slides to ensure that the parylene coated the entire device.

[0238] [Table 1]

[0239] The manufacture of the implantable device is discussed in further detail below.

[0240] Before starting with the PCB, ASIC, or PZT coupons, prepare two sample holders for the dust substrate. To do this, take two glass slides (3mm x 1mm x 1mm slides are good) and apply a piece of double-sided tape along the length of the slides. Use the tape to hold the dust mote in place for the remaining steps. On top of one of the slides, on top of the double-sided tape, add a piece of Kapton tape (3M), sticky side up. This slide will be used as a curing slide, as the high temperature of curing can cause problems with the adhesive of the double-sided tape.

[0241] Next, weigh out a 1:1 ratio of parts A and B in a weigh boat and mix in a small amount of silver paste. You don't need a lot of silver epoxy for the assembly process. Shown below is about 10g of epoxy (5g each part), which is more than enough for three boards. Note that the mixed silver epoxy has a shelf life of 2 weeks when kept at 4°C. Therefore, any leftover epoxy can and should be refrigerated when not being used. Additionally, aged epoxy (a few days to a week) tends to be slightly more viscous than fresh epoxy, which makes it easier to apply.

[0242] The boards are panelized and need to be removed. Each board is connected to the panel with several attachment points on the test leads and vias. These attachment points can be cut using a miniature scalpel (Feather Safety Razor Co., Osaka, Japan). Once the PCBs are singulated, place the singulated PCB into the high-temperature sample holder using carbon fiber-tipped tweezers or ESD plastic tweezers.

[0243] The diced / thinned die are transported on a dicing tape, which can make it difficult to remove the die. It is helpful to deform the tape to reduce the adhesion between the die and the tape. Using carbon-tipped or ESD plastic tweezers, gently press the tape and move the tweezers in a circular motion around the die. To see if the die has come off, gently poke the chip with the tip of the tweezers. If the die does not come off easily, continue to press the tape around the chip. Once the chip has come off, carefully place it on a hot sample holder next to its substrate. It is preferable to carry the sample holder over the chip rather than the other way around to prevent the chip from moving. Care must be taken during this process to avoid losing or damaging the die. Do not forcefully remove the die from the tape, as excessive force can blow the chip off the tape.

[0244] Next, the die is attached using silver epoxy. Under a microscope, a pin or similarly fine object is used to apply a small amount of silver epoxy to the CMOS pads on the PCB. It is better to accidentally apply too little epoxy in this process than to accidentally apply too much. You can always apply more silver paste, but removing silver paste is not a trivial task. Small amounts of uncured epoxy can be scraped off using the same tool used to apply it. Make sure to wipe the epoxy off the tool.

[0245] Once the epoxy is on the pads, the ASIC can be placed on top of the epoxy. Due to CAD errors, some of the chips will bounce around. It is important to ensure that the chips are oriented correctly on the board and that no wires are crossed when wire bonding.

[0246] Once the ASIC is correctly positioned on the board, you can cure the silver epoxy by placing it in a 150°C oven for 15 minutes. A different temperature can be used if desired. See Figure 12 for more information. Once the silver epoxy has cured, be sure to gently press each die to check for adhesion. If any die move, a second coat of silver epoxy is required.

[0247] To prepare for wire bonding, move the device from the high temperature sample holder to the regular sample holder. This change is necessary because the double-sided tape has a stronger adhesion than the Kapton tape, so wire bonding will occur more easily. A strip of double-sided tape should be enough to secure the sample holder to the work holder of the wire bonder. It is best that the work holder has not been covered in double-sided tape before, so that the test leads do not accidentally get caught on something. If necessary, cleanroom tape can be used to further secure the sample holder.

[0248] Use the default settings and make a bond on the test board provided to ensure the wire bonder is in good condition. It is important to keep the wire bonder in good condition, as a damaged wedge will not bond well and will only end up damaging the ASIC pads. Forward bonds (bond to die first, then to substrate) should be made in the following order: 1. Gate. 2. Bulk. 3. Center. 4. Drain. 5. Source. It is not important to make the bonds in this order, but this order minimizes the number of substrate reorientations and prevents the bond from being accidentally damaged by the bond head. Small adjustments to the work holder angle may be made to facilitate bonding. It is essential that this bond is as straight as possible. If the end of the second bond lifts off the substrate, it may be successful to reduce the number of bonds to one and rebond the end. If the bond does not adhere properly, one possible solution is to connect the end of the bond to the substrate using silver epoxy. Also, shorts caused by two bond ends touching can be fixed by very carefully cutting the connecting metal with a tiny scalpel.

[0249] Known practical bonding parameters are listed in Table 2 below. These parameters are only guidelines and should be modified if necessary. Requiring excessive power (above 490) usually indicates a problem. The substrate fixturing (both the PCB to the glass slide and the CMOS to the PCB), the condition of the wedge, and the condition of the pads should all be checked. In the case of pad condition, pads damaged from previous wire bonds usually require more power. In some cases the device can be salvaged, but failed attempts to bond at powers above 600 usually end up damaging the pads too much to make a good bond.

[0250] [Table 2]

[0251] After wire bonding, the device should be electrically tested to ensure that the bonds are made properly. If a Type 1 die is used, the I-V characteristics should be approximately as shown in Table 3.

[0252] [Table 3]

[0253] If the I-V characteristics do not seem correct, a valuable troubleshooting method is to check the drain-to-center, source-to-center, and drain-to-source resistances. If the die is working properly, you can expect the drain-to-center and source-to-center resistances to be around 90kΩ, and the drain-to-source resistance to be around 180kΩ.

[0254] After making sure the FETs are properly connected, the PZT coupon should be attached. This is done in a similar fashion to attaching the ASIC. Using a sewing needle, place a small amount of silver epoxy on the appropriate pad. It is best to place a small amount of epoxy at the back end of the pad (towards the edge of the board) because the PZT coupon will not be centered on the pad and will be pushed back and the coupon will come off the board. Note that the polarity of the PZT coupon only slightly affects its efficiency. To determine if the coupon is in the correct position, check to see if the bottom side is larger than the top side. Due to the path of the dicing saw, the bottom of the coupon is slightly larger than the top. Therefore, the bottom edge can be seen from above. The coupon is now oriented in the same way it was when the disk was diced.

[0255] The wirebonding of the PZT is done in a similar manner to the ASIC (forward bonding, PZT bonded to PCB). However, one important difference is that the drain and source vias are grounded. There is an earth port next to the Westbond, accessible through a banana connector. As a guideline, the parameters shown in Table 4 have been found to work.

[0256] [Table 4]

[0257] It may take several bonding attempts to make a successful bond, depending on how well the PZT coupon was attached to the substrate. The more bonding attempts, the worse the mechanical structure of the PZT becomes (the silver coating is damaged). Therefore, it is best to optimize the process very quickly. Bonds that fail due to a peeled end generally indicate that the power was not enough. Bonds that fail due to a broken wire generally indicate that the power was too high.

[0258] Once a successful bond has been formed, further electrical testing may be performed to ensure that the PZT bond has not damaged the ASIC.

[0259] The final step is to solder test wires into the vias and encapsulate the device. The test wires are 3 mil silver wire. These wires are insulated. The insulation can be removed by holding the wires near a flame (but not in the flame) and watching as the plastic melts and retreats.

[0260] After the wires are soldered, the device may be encapsulated. The encapsulation material is OG116-31 medical grade UV curable epoxy and should be applied with a sewing needle. An effective method is to place a large drop of epoxy on top of the PZT coupon and a large drop on top of the ASIC. Using a clean needle, press a small drop onto the board so that the entire top surface of the board is coated. The epoxy will wet the board but will not spill due to its surface tension. Once the body of the board is coated, coat the vias and the sides of the piezo elements as well. The board can then be cured in a UV chamber for approximately 5 minutes. It has been found that test wires can sometimes come into contact with something in the UV chamber and short out the ASIC. Therefore, before placing the board in the chamber, it is a good idea to wrap the wires or place the board on some tape to insulate the wires from any surface of the chamber.

[0261] After curing, the backside should be coated, especially the exposed PZT coupon covering the board, the backside of the test via, and the backside of the two vias on the backside of the board that connect the electrodes to the topside of the board. This part can be a little difficult because the space between the vias and the electrodes on the backside is small, so it is best to start with a very small amount of epoxy, place it near the edge of the board, and then pull the epoxy down towards the vias. The backside of the board should cure in the same way as the topside. Once the board is fully encapsulated, a final electrical test should be performed. If the test passes, the implantable device is complete.

[0262] Example 2: Setup for testing implantable devices Testing implantable devices has always been difficult due to the thin test leads that extend from the board. Clipping these vias on and off for I-V measurements often resulted in pulling the leads off the body of the device. Additionally, the test leads make it difficult to perform water tank test measurements because electronics exposed to water will short out. To circumvent this issue, a PCB was designed to serve as a testbed for implantable device measurements. The PCB (Bay Area Circuits, Fremont, CA) was fabricated from FR-4 and was 60 mils thick. The PCB has four vias arranged on the board to match the layout of the Version 2 implantable device board.

[0263] Gold header pins (Pin Strip Header, 3M, Austin, TX) were soldered to the vias, extending from both sides of the board. This allowed us to place our device on the testbed and access the header pins to contact the embedded device. We then 3D printed polyethylene terephthalate (PETG) plastic caps (Flashforge Creator X, FlashForge, Jinhua, China) to insulate the vias. These caps were printed with grooves, and an O-ring was fitted inside the groove to create a watertight seal around the header pins. The caps were connected to the board, and a micromill (47158, Harbor Freight, Camarillo, CA) was used to drill 2 mm holes in the PCB and caps, and the caps and board were screwed together to compress them. Wires extending from the testbed were soldered to the header pins, which then encapsulated the pins. To measure the effectiveness of the seal, the board was immersed in a 6 M NaCl solution and the resistance between the pins was measured using a Keithley 2400. A MATLAB script was written to automatically record and plot the resistance over time. A drop in resistance indicates that the seal has been broken. As an additional test, a piece of litmus paper was also placed under the plastic cap with the intention that if the cap leaked, the litmus paper would change color. The pins were encapsulated using the same medical grade epoxy used to encapsulate the implantable device board, and parylene was evaporated onto the epoxy on the back side of the test board to create a complete waterproof barrier. The resistance between two adjacent pins of the test bed immersed in a salt water solution was measured as a function of time with epoxy insulation only and with epoxy + parylene insulation. Without the parylene barrier, the epoxy began to leak and the test bed pins were shorted by the salt water.

[0264] One version of the implantable device consisted of a 1 mm x 3 mm x 1 mm PCB made of FR-4 with PZT piezoelectric elements and silicon ASICs encapsulated using Crystalbond. These were implanted into the sciatic nerve of adult male Long-Evans rats anesthetized with a mixture of ketamine and xylazine IP. Ground truth measurements were obtained as a reference using tungsten microwires with 28G stainless steel needle electrodes in the legs of the animals. Neural activity was evoked by electronic stimulation, and backscatter data was obtained by transmitting and receiving pressure waves using a transducer (V323-SU-F1 Olympus, Waltham, MA).

[0265] The original signal traversing the dust mote was later calculated from the backscattered data using MATLAB. A representative trace of the reconstructed signal versus the ground truth is shown in Figure 13.

[0266] The reconstructed implanted device data followed the general profile of the ground truth and captured the compound neural action potential, but some features in the reconstructed data (e.g., the "dip" seen from seconds 1 to 3) were unexplained.

[0267] The second version of the implantable device was approximately 0.8mm x 3mm x 1mm and used a polyimide substrate with a medical grade UV-curable epoxy as the encapsulation. The key change was the addition of a length 1 test lead, which was captured by the derivation electrode to allow measurement of the voltage across the piezoelectric element and ground truth measurements. The same device implantation protocol used in version 1 was used in version 2, but the reconstruction of the backscattered signal was done on the fly using a custom transceiver board.

[0268] Example 3: Silicon carbide encapsulated implantable device Silicon carbide (SiC) may be a more effective material for insulating and protecting implantable devices than epoxy encapsulants. SiC is formed by covalent bonds between Si and C, forming tetrahedral molecules with short bond lengths, thus providing high bond strength and chemical and mechanical stability. Amorphous SiC (a-SiC) has been welcomed as a coating material in the biomedical community because a-SiC can be deposited at temperatures much lower than those typically required for crystalline SiC and is an electrical insulator. Deposition of a-SiC is typically performed by plasma-enhanced chemical vapor deposition (PECVD) or sputtering. Ongoing research with sputtered a-SiC has shown that it is difficult to form pinhole-free SiC layers. Instead, PECVD using SiH4 and CH4 as precursors can produce impressive pinhole-free SiC films.

[0269] Moreover, the implanted a-SiC has demonstrated remarkable biocompatibility: previous studies have shown that a-SiC-coated 50 μm iridium shafts implanted in the cerebral cortex of rabbits for ~20 days did not elicit the usual chronic inflammatory response with macrophages, lymphocytes, and monocytes recruited to the insertion site. See Hess et al., PECVD silicon carbide as a thin film packaging material for microfabricated neural electrodes, Materials Research Society Symposium Proceedings, vol. 1009, doi: 10.1557 / PROC-1009-U04-03 (2007).

[0270] It is interesting to consider an implantable device approach that involves creating a device with a silicon carbide encapsulant on silicon for a truly long-term implant. A possible process is shown in Figure 14. One of the biggest challenges here is to ensure that the PECVD of SiC does not depole the piezoelectric material. It is important to deposit at a minimum temperature of 200°C, below the Curie temperature of the piezoelectric transducer, to obtain a contamination-free film.

[0271] Example 4: Power transmission and backscattering of miniaturized ultrasonic transducers A set of experiments was performed with PZT due to the relative ease of obtaining PZT crystals with various structures. Metallized PZT sheets of several thicknesses were obtained (PSI-5A4E, Piezo Systems, Woburn, MA and PZT84, APC Internationals, Mackeyville, PA). The minimum thickness of the PZT was 127 μm. For biocompatibility, the PZT was fully encapsulated in PDMS silicone.

[0272] The most commonly used method of dicing PZT ceramics is to cut the PZT sheet into individual PZT crystals using a wafer dicing saw with an appropriate ceramic blade. The minimum resolution of the cut is determined by the cutting kerf of the blade and can be as small as 30 μm.

[0273] Another possible method is to use a laser cutter. Unlike a dicing saw, laser cutting involves focusing a high-powered laser beam into the material, which melts, vaporizes, removes, or cuts pieces. The precision of laser cutting can be as fine as 10 μm and is limited to the wavelength of the laser. However, when dealing with sensitive samples such as PZT ceramics, the temperature at the cut site can damage the piezoelectric performance of the material. Excimer laser cutting of ceramics uses a UV laser to cut with a rare gas excimer, but such laser cutters are extremely expensive and currently there is no suitable equipment available. Therefore, a dicing saw was used for all cuts.

[0274] To drive or extract electrical energy from the PZT, electrical connections are made to both the top and bottom plates. The materials typically used as electrodes for PZTs are silver or nickel. Silver is typically used for many types of non-magnetic AC applications, where the silver is suspended in a glass frit in flake form, usually screened onto the ceramic and fired. For high field DC applications, silver has a high chance of migrating and bridging between the two plates. Therefore, nickel, which has high corrosion resistance and does not readily migrate, can be electroplated or evaporated as an alternative material.

[0275] Both materials can be soldered with the appropriate solder and flux. For example, silver is soluble in tin, but a silver-containing solder can be used to prevent the trapping of silver in the electrode. The phosphorus content of the nickel plating makes soldering difficult, but the correct flux can be used to prevent surface oxidation. However, when soldering, the soldering temperature must be between 240°C and 300°C to avoid exceeding the Curie point and depoling the PZT sample. Even at these temperatures, care must be taken not to exceed soldering times of 2-4 seconds, as PZT is also pyroelectric.

[0276] Alternatively, the electrical connection may be made with silver epoxy or by low-temperature soldering using solder paste. Standard two-component silver epoxy can provide sufficient electrical conductivity and can be cured overnight at room temperature. However, the joints tend to be brittle and can easily break during testing. The bond can be reinforced using a non-conductive epoxy encapsulation, but this additional layer can mechanically load the PZT and significantly reduce its quality factor. Low-temperature solder paste, on the other hand, undergoes a phase change at temperatures between 150°C and 180°C and can provide an excellent electrical connection and a bond strength comparable to that achieved with flash soldering. Therefore, the low-temperature soldering approach was used.

[0277] Wafer dicing can cut PZT into small crystals on the order of 10 μm. However, samples with dimensions smaller than 1 mm are extremely difficult to handle and bond with tweezers. Furthermore, impedance spectroscopy measurements were inconsistent due to different lengths of wires used to connect to the top and bottom plates of the PZT crystals (and therefore different parasitic inductances and capacitances introduced by the wires) and different amounts of solder paste dispensed among the multiple samples.

[0278] Therefore, a 31 mil thick, two-layer FR-4 PCB was fabricated with all electrical interconnects shorting and de-embedding parasitic vibrations from the wires and the substrate. The fabricated board (with various test structures and modules individually characterizing PZT crystals with thicknesses of 127 μm, 200 μm, and 250 μm) is shown with dimensions in Figure 15. Each unit cell in the test module has two pads of specific dimensions on one side of the PCB for connection to the PZT crystals, and on the other side pads for the individual elements for backscatter communication. The pitch between the unit cells is limited by the size of the individual elements and is approximately 2.3 mm x 2 mm.

[0279] To avoid directly handling the extremely small PZT crystals, Figures 16A-16E outline a scalable process flow for bonding PZT onto a PCB. As shown in Figure 16A, solder paste is dispensed using a pump under constant pressure for a limited time onto one of the top pads. The pads are then bonded to a thickness of 250 μm based on the thickness of the PZT used. 2 , 200μm 2 , or 127 μm 2 FIG. 16B shows a larger PZT piece (easier to handle) placed on top to cover the pads. The substrate and piezo assembly are baked in an oven to harden the solder paste. Thus, the PZT crystals are now bonded to the raised electrodes before soldering. FIG. 16C shows a wafer dicing saw making a total of four cuts along the edge of the pads, using the solder paste as an alignment marker on the substrate, and trimming away the unbonded areas, leaving an array of small PZT crystals bonded to the PCB. FIG. 16D shows a single wire bond making electrical contact between the top plate of the PZT and the electrode on the PCB, completing the circuit. Finally, FIG. 16E shows the entire assembly being encapsulated in PDMS (Sylgard 184, Dow Corning, Midland, MI) to protect and insulate the wire bonds.

[0280] Since piezoelectric materials are electromechanical structures, their electrical and mechanical properties were measured. Below, the test setup and techniques for making such measurements are detailed.

[0281] Any electrical device can be modeled as a black box using a mathematical construct called two-port network parameters. The characteristics of the circuit are specified by a matrix of numbers, and the response of the device to a signal applied to its input can be easily calculated without solving for all the internal voltages and currents in the network. There are several different types of two-port network parameters (e.g., Z-parameters, Y-parameters, S-parameters, ABCD-parameters, etc.), and conversion between different parameters can be easily performed. The device that allows the extraction of these parameters is called a Vector Network Analyzer (VNA). The VNA contains a directional coupler that breaks down the voltage at each port into incident and reflected waves (based on impedance mismatching) and calculates the ratio between these waves to calculate the scattering or S-parameters.

[0282] Before performing measurements with the VNA, the instrument needs to be calibrated because the internal directional coupler is non-ideal. Calibration allows to move the reference plane of the measurement to the end of the cable, i.e. to calibrate out the parasitic vibrations from the cable. There are several calibration standards, but the most commonly used is the open, short and load calibration procedure. A schematic diagram of the measurement is shown in Figure 17. The top / bottom plate is connected using double clips soldered to the end of the coaxial cable. The parasitic vibrations from the clips are not much lower than 100 MHz.

[0283] As an example, a VNA (E5071C ENA, Agilent Technologies, Santa Clara, CA) was used to measure the distance (250 μm). 3The electrical properties of the PZT crystal were measured. It was noticed that the measured capacitance of the PZT crystal was significantly different from that expected from a single parallel plate capacitance model due to the large parasitic capacitance from the PCB and fixtures (clips and connectors). Since the VNA coefficients of the calibration process outlined above only moved the measurement plane to the tip of the cable, the parasitic vibrations of the board and fixtures were included using an open / short / load calibration structure made on the same board. The measured PZT response matched the expected response after calibration.

[0284] Using this calibration technique, the impedance of the PZT can be plotted as a function of frequency (see Figure 18B). However, from this plot it is extremely difficult to determine if there is any electromechanical resonance. By overlaying the simulation results obtained with air backing (no mechanical clamping), it was noticed that the impedance spectroscopy matches the measurements well at low and high frequencies, with the exception of a prominent peak at the resonance frequency of about 6 MHz and its harmonics. It was observed that the resonance peak due to the air backing was significantly weakened when one side of the PZT with a PCB (FR-4) was clamped and loaded. Despite the absence of observable resonance in the measurements, a small blimp at about 6 MHz was observed. Mechanical Quality Factor Q m can be calculated using the following formula:

[0285]

number

[0286] In the formula, f a and f r represents the anti-resonance frequency (where the impedance is maximum) and the resonance frequency (where the impedance is minimum), and Z r represents the impedance at resonance, and C prepresents the low frequency capacitance. The quality factor calculated from the measurements is about 4.2 compared to 5.1 from the simulation. According to the datasheet, the unloaded Q of the PZT is ~500, which indicates that the FR-4 backing and wire bonds significantly degrade the quality factor. Despite the drastic reduction in the mechanical Q of the PZT crystal, experiments showed that the backscattered signal level only decreased by about ~19.

[0287] In the electrical characterization setup, the VNA has a built-in signal generator that provides the input required for characterization. To perform acoustic characterization of the PZT, sound waves are generated and launched into the sample and used as the input. This can be done using a commercially available broadband ultrasonic transducer.

[0288] Figure 19 shows a typical transducer configuration consisting of a piezoelectric active element, a support and a backing plate. The support is usually made of a material with high damping and high density, thereby controlling the vibration of the transducer by absorbing energy from the back side of the active element, while the backing plate is used to protect the transducer element from the test environment and act as a matching layer.

[0289] Ultrasonic power transfer tests were performed using a home-made setup shown in Figure 20. A 5 MHz or 10 MHz single-element transducer (6.3 mm and 6.3 mm active area, respectively; focal length ∼30 mm; Olympus, Waltham, MA) was mounted on a computer-controlled two-axis translation stage (VelMex, Bloomfield, NY). The transducer output was calibrated using a hybrid capsule hydrophone (HGL-0400, Onda, Sunnyvale, CA). A prototype of the assembly was placed in a container of water such that the transducer was immersed in the water at a distance of approximately 3 cm directly above the prototype. Using a programmable pulse generator (33522B, Agilent Technologies Santa Clara, CA) and a radio frequency amplifier (A150, ENI, Rochester, NY), the transducer was driven at a specific frequency with a 10-cycle sinusoidal pulse train and a pulse repetition frequency (PRF) of 1 kHz. The received signal was amplified by a radio frequency amplifier (BT00500-AlphaS-CW, Tomco, Stepney, Australia) connected to an oscilloscope (TDS3014B, Tektronix, Beaverton OR), and the ultrasound signal was collected and recorded using MATLAB.

[0290] Figures 21A and 21B show representative measurements of the output power of a 5 MHz transducer as a function of distance (z-axis) between the transducer face and the hydrophone. The peak pressure in water was obtained at ~33 mm from the transducer face (Figure 21A), while the derating peak (at 0.3 dB / cm / MHz) was at ~29 mm (Figure 21B). Figure 22A shows an XZ scan of the derating transducer output, showing the near and far beam patterns and the Rayleigh distance or focus at ~29 mm, matching the derating peak in Figure 21B. Figure 22B shows an XY cross-sectional scan of the beam at the focus at ~29 mm, with a 6 dB beam width of approximately 2.2 mm.

[0291] The total integrated acoustic output power of the transducer at various frequencies over the 6 dB bandwidth of the beam is nominally the spatial peak temporal average I SPTA 29.2μW / cm 2 The total output power at the focus was kept at ∼1 μW, with a peak sparsification pressure of 25 kPa and a mechanical index (MI) of 0.005. SPTA Both the FDA regulatory limit of 720 mW / cm 2 and 1.9 (FDA 2008), respectively.

[0292] Figure 18A shows the measured power delivery efficiency of a fully assembled cable-free prototype calibrated for various implantable device transducer sizes compared to analytical predictions for this same setup. The measured results closely matched the behavior of the simulation model for all transducer sizes, with the exception of a few small transducer dimensions. This is likely due to sensitivity to transducer position and ultrasound beam width. The smallest PZT crystals (127 μm) 3 The efficiency of the link is measured as 2.064 × 10 -5 This resulted in a nominal received power of 20.64 pW at the transducer. The transmitted output power density was 720 mW / cm 2 When the power input is kept at 100 MHz, a maximum of 0.51 μW can be recovered at the transducer. Such low power levels obtained with the PZT are mainly due to the high inefficiency of the broadband transducers used in the experiments. Dedicated custom-made transducers with optimized electrical input impedance for each transducer dimension allowed a significant improvement of the obtained power levels by more than two orders of magnitude, as predicted by the simulation model.

[0293] (250μm) 3The frequency response of the resulting voltage across the PZT crystal is shown in Figure 18C. The measured resonant frequency was 6.1 MHz, which is consistent with the predicted resonant frequency shift for a cube due to Poisson's ratio and the corresponding mode coupling between the resonant modes along each of the cube's three axes. Furthermore, the calculated Q of 4 matched the electrically measured Q of PZT.

[0294] The above experimental results show that the analytical model of power coupling to very small PZT nodes using ultrasound is accurate at least down to the scale of ~100 μm and possibly even smaller. How small a transducer can be fabricated before loss of function remains to be seen. Note that measurements of even smaller nodes (<127 μm) were limited by the commercial availability of PZT substrates, not by the prototype assembly process. Moving forward, a significant amount of research and technology has looked at microelectromechanical and nanoelectromechanical RF resonators (see Sadek et al., Wiring nanoscale biosensors with piezoelectric nanomechanical resonators, Nano Lett., vol. 10, pp. 1769-1773 (2010); Lin et al., Low phase noise array-composite micromechanical wine-glass disk oscillator, IEEE Elec. Dev. Meeting, pp. 1-4 (2005)) and thin film piezoelectric transducers (Trolier-McKinstry et al., Thin film piezoelectrics for MEMS, J. Electroceram., vol. 12, pp. 7-17 (2004)), facilitating extremely small (10s of μm) transducers and truly evaluating the scaling theory.

[0295] Example 5: Beamforming using an interrogator ultrasonic transducer array In this example, we present an ultrasound beamforming system capable of interrogating individual implantable sensors via backscatter in a distributed, ultrasound-based recording platform. A custom ASIC drives a 7x2 PZT transducer array with three cycles of a 32V square wave with specific programmable time delays to focus the beam on an 800μm neural dust mote located 50mm away. The measured sound-to-electricity conversion efficiency of the receive mote in water is 0.12%, and the entire system delivers 26.3% of the power from a 1.8V supply to the transducer drive output, consumes 0.75μJ in each transmit phase, and changes the backscatter by 0.5% for every volt applied to the input of the backscatter circuit. Further scaling of both the transmit array and the receive mote paves the way for wearable long-term sensing and neuromodulation systems.

[0296] In this highly distributed and asymmetric system, where the implanted devices greatly outnumber the interrogation transceivers, beamforming can be used to efficiently interrogate a large number of implanted devices. By studying beamforming algorithms, tradeoffs, and performance on an implanted device platform, it is demonstrated that cooperation between different interrogators is useful to sufficiently suppress interference from nearby implanted devices. See Bertrand et al., Beamforming approaches for untethered ultrasonic neural dust motes for cortical recording: a simulation study, IEEE EMBC, 2014, pp. 2625-2628 (Aug. 2014). This example demonstrates a hardware implementation of an ultrasonic beamforming system for the interrogator and implanted device system shown in FIG. 2A. The ASIC (see, e.g., Tang et al., Integrated ultrasonic system for measuring body-fat composition, 2015 IEEE International Solid-State Circuits Conference - (ISSCC) Digest of Technical Papers, San Francisco, CA, 2015, pp. 1-3 (Feb. 2015); Tang et al., Miniaturizing Ultrasonic System for Portable Health Care and Fitness, IEEE Transactions on Biomedical Circuits and Systems, vol. 9, no. 6, pp. 767-776 (Dec. 2015)) has seven identical channels, each with 6-bit delay control with 5 ns resolution for transmit beamforming, and integrates high-voltage level shifters and receive / transmit switches that isolate any feedthrough.

[0297] The ASIC runs on a single 1.8V supply and generates a 32V square wave to drive the piezoelectric transducers using integrated charge pumps and level shifters. The system delivers ~32.5% of the power from the 1.8V supply to the 32V output voltage and ~81% of the power from the 32V supply to the output load (each transducer element is 4.6pF). A block diagram of the ASIC is shown in Figure 2A. Details of the circuitry that enables such low energy consumption per measurement are given in Tang et al., Integrated ultrasonic system for measuring body-fat composition, 2015 IEEE International Solid-State Circuits Conference - (ISSCC) Digest of Technical Papers, San Francisco, CA, 2015, pp. 1-3 (Feb. 2015). The ASIC is fabricated in 0.18μm CMOS with high-voltage transistors. The chip area is 2.0mm 2 and includes the entire system except for the digital controller, ADC, and two off-chip blocking capacitors.

[0298] The design of a transducer array is a strong function of the desired penetration depth, aperture size, and element size. Quantitatively, the Rayleigh distance R of the array can be calculated as

[0299]

number

[0300] where D is the size of the aperture and λ is the wavelength of ultrasound in the propagation medium. By definition, the Rayleigh distance is the distance at which the beam emitted by the array is fully formed. In other words, the pressure field converges to a natural focus at the Rayleigh distance, and to maximize the received power, it is preferable to place the receiver at the Rayleigh distance where the beam divergence is minimal.

[0301] The frequency of operation is optimized to match the size of the element. Preliminary studies in a water tank have shown that it is possible to achieve a frequency of 100 kHz (800 μm). 3 The maximum energy efficiency is achieved with PZT crystals (800μm) 3 The PZT crystal has a resonant frequency of 1.6 MHz after encapsulation, resulting in a λ of ~950 μm. The pitch between each element is chosen to be an odd multiple of a half wavelength for efficient beamforming. As a result, for this demonstration of beamforming capabilities, the overall aperture is ~14 mm, and the Rayleigh distance is 50 mm. At 50 mm, each element is far enough out of the field (R=0.17 mm) for an element size of 800 μm. Thus, the beam pattern of the individual elements is omnidirectional enough to allow beamforming.

[0302] Several transmit and receive beamforming techniques are possible. Herein, a time-delay-sum transmit beamforming algorithm is chosen to constructively direct the signals to the target direction. This algorithm can demonstrate beam steering and maximum power delivery to a variety of implanted devices. To accommodate backscatter communication to a large number of implanted devices simultaneously, more sophisticated algorithms may be required. These algorithms include delay-sum beamforming, linearly constrained minimum variance beamforming, convex optimization beamforming for a single beam, "multicast" beamforming with convex optimization, maximum kurtosis beamforming, minimum variance distortionless response robust adaptive beamforming, multinomial tensor decomposition, and deconvolution of mote impulse responses from multiple Rx channel time domain data. A detailed treatment of one aspect of this problem is described in Bertrand et al., Beamforming approaches for untethered ultrasonic neural dust motes for cortical recording: a simulation study, IEEE EMBC, 2014, pp. 2625-2628 (Aug. 2014).

[0303] Each of the seven channels is driven by three cycles of a 32V square wave with a specific programmable time delay, focusing the energy at an observation distance of 50mm. The time delay applied to each channel is calculated based on the difference in the propagation distance from the center of the array to the focal point and the difference in the propagation time of ultrasound in the medium.

[0304] Ultrasim was used to characterize the ultrasonic propagation behavior in water from the 1D array described above. The simulated XY (FIG. 24A) and XZ (FIG. 24B) cross-sectional beam patterns closely match the measurements shown, despite not modeling the PDMS encapsulation.

[0305] Water is used as the medium for the beamforming system measurements since it exhibits acoustic properties similar to tissue. Premetallized lead zirconate titanate (PZT) sheets (APC International, Mackeyville, PA) are diced with a wafer saw to yield 800 μm × 800 μm × 800 μm crystals (each with a parallel capacitance of 4.6 pF), the size of each transmit element. Each PZT element is electrically connected to a corresponding channel in the ASIC using conductive copper foil and epoxy on the bottom terminals and wire bonds on the top terminals. The array is encapsulated in PDMS (Sylgard 184, Dow Corning, Midland, MI) to protect and insulate the wire bonds. The quality factor of the PZT crystals after encapsulation is ~7. The array is organized into seven groups of 2 × 1 elements with a pitch of ~5 / 2λ ~2.3 mm. The size of the array is approximately 14 mm × 3 mm. Finally, the entire assembly is placed into a cylindrical tube 25 mm in diameter and 60 mm in height, and the tube is filled with water.

[0306] The 2D beam pattern and output of the transducer array are calibrated using a capsule hydrophone (HGL-0400, Onda, Sunnyvale, CA). The hydrophone is mounted on a computer-controlled 2D translation stage (VelMex, Bloomfield, NY). The hydrophone has an acceptance angle of 30° (-6 dB at 5 MHz), which is sufficient to capture the beam provided with a transmission distance of 50 mm and a scanning range of ±4 mm.

[0307] The array is overlaid on the measured XY cross-sectional beam pattern shown in Figure 23A. The delay applied to each transducer (element) in the array is shown in Figure 23B. The -6 dB beam width at the focal point is 3.2 mm ~ 3λ. The flexibility of the ASIC allows both wide and granular programming of the delay. The peak pressure levels of the array at 50 mm before and after beamforming are ~ 6 kPa and ~ 20 kPa, respectively. The 3X in transmit output pressure wave after beamforming matches the simulation. The simulation also shows that the Rayleigh distance of the array is 50 mm (see Figure 23C).

[0308] Furthermore, to validate the performance of interrogating multiple implantable devices, we validated the beam steering performance of the array as shown in FIG. 24A, which shows beam steering at three different positions in the XY plane. The time delay at each beam position is shown below in FIG. 24B. The 1D beam steering closely approximates the simulation (see FIG. 24C). Note that the beam steering range is limited to ±4 mm by the mechanical configuration of the array, rather than the electronic performance.

[0309] The hydrophone is replaced with an implantable device (with a 800 μm × 800 μm × 800 μm bulk piezoelectric transducer) and placed at a transmit distance of 50 mm to verify the power link. The open circuit peak-to-peak voltage measured at the mote is 65 mV for a transmit pulse duration of 2.56 μs. The spatial peak average acoustic power integrated into the -6 dB beamwidth at the focus is 750 μW, which is 0.005% of the FDA-defined safety limit. The maximum obtainable power at the mote is 0.9 μW, and the measured sound-to-electrical conversion efficiency is 0.12%. The measurements are consistent with the link model (Seo et al., Model validation of untethered ultrasonic neural dust motes for cortical recording, J. Neurosci. Methods, vol. 244, pp. 114-122 (2015)). The above system delivers 26.3% of the power from a 1.8V supply to the transducer drive output (defined as the drive efficiency) and consumes 0.75 μJ for each transmit phase.

[0310] The ultrasound backscatter communication performance of the system is validated by measuring the difference in backscatter voltage levels as input to the backscatter circuit (Seo et al., Model validation of untethered ultrasonic neural dust motes for cortical recording, J. Neurosci. Methods, vol. 244, pp. 114-122 (2015)) and regulated with a DC power supply. The transmit time and period of the system are 3 μs and 80 μs, leaving a window of ~77 μs for reception. The central 2 × 1 element of the array is used to receive the backscatter. The output of the receiving crystal is amplified and digitized for processing. The measured backscatter sensitivity is ~0.5% per volt applied to the input of the backscatter circuit, which is consistent with simulations. The overall performance of the system is summarized in Table 5.

[0311] [Table 5]

[0312] Our measurements with an ultrasonic beamforming system show that transmit beamforming alone can provide sufficient signal-to-noise ratio (SNR) for multiple sensor interrogation in the neural dust platform. The SNR reduction due to dust mote shrinking can be largely mitigated by implementing receive beamforming. Furthermore, alternative means of multiplexing can be used to increase the interrogation ratio. An example of multiplexing is spatial multiplexing, where multiple motes are simultaneously interrogated with the same transmit beam. However, it is important to consider the trade-off between processing / communication load and power consumption in the system design. Furthermore, interference from nearby dust motes needs to be sufficiently suppressed to achieve the required SNR.

[0313] The current overall system efficiency is 0.12% acoustic-electrical efficiency.

[0314]

number

[0315] Despite this low power link efficiency, it still meets the FDA safety regulation of ~1% (spatial peak average 1.9 W / cm2). 2 ) it is possible to obtain a peak-to-peak voltage of up to 0.92 V and up to 180 μW for an 800 μm ultrasonic transducer 50 mm away in water.

[0316] Furthermore, the low power link efficiency in this demonstration is a result of such a large transmission distance, which is determined by the array aperture and element size. For peripheral nerve interference, for example, the desired transmission distance, including the thickness of skin, tissue, etc., is about 5 mm. For the long range of the array, the aperture should be ~4.4 mm. By further scaling each element, the overall dimensions of the array aperture and transmission distance can be reduced to the desired 5 mm. Simulations show that acoustic-to-electrical efficiencies of up to 1% can be achieved with a 100 μm receiving ultrasound transducer in water.

[0317] For intra-tissue transmission, we assume that there is a loss of 3 dB / cm / MHz in the tissue. Figure 25 shows the scaling of both link efficiency and received power level for operation at 1% of the FDA safety limit. Despite this rather modest loss, simulations show that at 100 μm, a peak-to-peak voltage of up to 0.6 V and up to 75 μW can be obtained. Thus, intra-tissue wireless power delivery using this platform is feasible. Furthermore, this power level is sufficient to operate highly efficient and low-power energy consumption circuits and charge pumps (similar to the ASICs mentioned above) to output voltages suitable for electrically stimulating nearby neurons and detecting physiological conditions using sensors.

[0318] (Example 6) Wireless recording of the peripheral nervous system using ultrasonic neural dust The following example demonstrates an implantable device system that records neural signals. This example shows that ultrasound is efficient for delivering power to mm-scale devices in tissue. Similarly, electromyograms (EMG) and electroneurograms (ENG) can be transmitted with high fidelity from anesthetized rats by passive battery-free communication using backscatter. These results highlight the potential of ultrasound-based neural interface systems to advance bioelectronic-based therapies in the future. This example further provides a method for determining the position and movement of an implantable device.

[0319] An in vivo implantable device system was used to report electroneurogram (ENG) from the sciatic nerve and electromyogram (EMG) from the gastrocnemius muscle of the peripheral nervous system of a subject rat. The system had an external ultrasound transducer that powered and communicated with a millimeter-scale sensor that was not implanted in the nerve or muscle. See Figure 26A. The implantable device had a piezoelectric crystal, a single custom transistor, and a pair of derived electrodes. See Figures 26B, 26C, and 26D.

[0320] During operation, the external transducer alternates between a) emitting a series of six 540 ns pulses every 100 μs and b) listening for any reflected pulses. The entire sequence of transmission, reception, and reconstruction events is detailed in FIG. 27A-H. This sequence is repeated every 100 μs during operation. In essence, a pulse of ultrasonic energy emitted by the external transducer is partially reflected back to the external transducer. In addition, some of the ultrasonic energy vibrates the piezoelectric crystal. This causes the piezoelectric crystal to convert the mechanical force of the ultrasonic waves into electrical power that is fed to the transistor. Any changes in the extracellular voltage between the two output electrodes modulate the gate of the transistor, changing the amount of current flowing between the terminals of the crystal. These changes in turn change the vibration of the crystal and the intensity of the reflected ultrasonic energy. In this way, the shape of the reflected ultrasonic pulse encodes the electrophysiological voltage signal seen at the implanted electrodes, which can be reconstructed externally. Table 6 summarizes the performance specifications of Neural Dust compared to other state-of-the-art systems.

[0321] [Table 6]

[0322] The implantable device was fabricated on a 50 μm thick polyimide flexible printed circuit board (PCB) with the ultrasound transducer piezoelectric crystal (0.75 mm × 0.75 mm × 0.75 mm) and custom transistor (0.5 mm × 0.45 mm) attached to the top side of the substrate with conductive silver paste. Electrical connections between elements are made with aluminum wire bonds and conductive gold traces. Gold lead-out pads (0.2 mm × 0.2 mm) exposed on the bottom of the substrate are spaced 1.8 mm apart and contacted with nerves or muscles to record electrophysiological signals. The recorded signal is sent through microvias to the input of the transistor. Additionally, some implants had flexible, compliant leads 0.35 mm wide and 25 mm long that had test points for simultaneous measurement of the voltage across the piezoelectric crystals and direct wired measurement of the extracellular potential between the electrode pairs used by the ultrasound transducer (this direct wired recording of the extracellular potential as a ground truth measurement is described below. It is used as a control for the ultrasound reconstruction data). The entire implant was encapsulated in medical grade UV-curable epoxy to protect and insulate the wire bonds. The size of a single implantable device is approximately 0.8 mm x 3 mm x 1 mm. The size of the implant is only limited by our use of commercially available polyimide backplane technology, which is available to anyone to purchase. If we resorted to a more aggressive assembly technique with in-house polymer patterning, implants not much larger than the piezoelectric crystal dimensions could be made (~1 mm). 3 (generating implants).

[0323] The implantable device assembly is detailed further. Sintered lead zirconate titanate (PZT) sheets with ~12 μm of silver (841, APC Int., Mackeyvile, PA) were diced to the desired dimensions using a dicing saw (DAD3240, Disco, Santa Clara, CA) with a ceramic blade (PN CX-010-270-080-H). The diced PZT coupons, along with the custom transistors, were attached to a displacement gold-plated 50 μm thick polyimide flexible PCB (Altaflex, Santa Clara, CA) using a thin film of 1:1 mixed 2-part silver epoxy (H20E, Epotek, Billerica, MA). The substrate was cured for 10 min at 150 °C (which is well below the melting temperature of polyimide and the Curie temperature of PZT). The custom transistors were wire bonded to the pre-patterned target using an aluminum ultrasonic wire bonder (7400B, West Bond, Scotts Valley, CA). To prevent charge build-up on the PZT due to contact with the wedge, the top and bottom contacts of the PZT were discharged to a thin metal sheet before the top contact of the PZT was wire bonded to close the circuit. A medical grade UV-curable epoxy (OG116-31, Epotek) was used to protect and insulate the wire bonds. The platform was then powered with a 92 mW / cm 2 Cured in a UV chamber @ 365 nm (Flash, Asiga, Anaheim Hills, CA) for 3 minutes.

[0324] A custom integrated circuit drives an external transceiver board to enable low-noise interrogation. The external ultrasound transceiver board communicates with the implanted device by providing power (transmit (TX) mode) and receiving reflected signals (receive (RX) mode). The system is a low-power, programmable, and portable transceiver board that drives a commercially available external ultrasound transducer (V323-SU, Olympus, Waltham, MA). The transceiver board showed a reduced pressure focus at ~8.9 mm (Figure 28A). The XY cross-sectional beam pattern clearly demonstrated the transition from near-field to far-field propagation of the beam. The narrowest beam was at the Rayleigh distance (Figure 28B). The transducer was driven with a 5 V peak-to-peak voltage signal at 1.85 MHz. The reduced peak rarefaction pressure was measured to be 14 kPa, resulting in a mechanical index (MI) of 0.01. 6.37 mW / cm at 10 kHz pulse repetition. 2 and 0.21 mW / cm 2 Reduced spatial pulse peak average (I SPPA ) and spatial peak temporal average (I SPTA ) were 0.0034% and 0.03% of the FDA regulatory limits, respectively. The transceiver board was capable of outputting a peak-to-peak voltage of up to 32 V, and the output voltage increased linearly with the input voltage (Figure 28C).

[0325] Reflections from non-piezoelectric crystal interfaces provide a built-in reference for motion artifacts and temperature drift. The entire system was characterized in a custom-made water tank with a manual 6-degree-of-freedom (DOF) linear translation-rotation stage (Thorlabs Inc., Newton, NJ). Distilled water was used as the propagation medium. Distilled water has a similar acoustic impedance to tissue at 1.5 MRayls. For an initial calibration of the system, a current source (2400 LV, Keithley, Cleveland, OH) was used to pass electrical current of varying current density through a 0.127 mm thick platinum wire (773000, AM Systems, Sequim, WA) immersed in the tank to mimic extracellular signals. The implantable device was immersed in the current path between the electrodes. As current was applied between the wires, a potential difference was raised between the electrodes of the implant. This potential difference was used to mimic extracellular electrophysiological signals during tank testing.

[0326] The electrical and ultrasonic characterization of the underwater assembly is further detailed. The custom transistor was electrically tested using a precision current meter (2400-LV, Keithley) and a DC power supply (3631A, Agilent, Santa Clara, CA). To characterize the piezoelectric crystal prior to assembly, impedance plots were obtained using two-terminal measurements with an open / short / load calibration scheme on an impedance analyzer (4285A, Agilent). The impedance of exposed gold lead-out pads (0.2 mm × 0.2 mm) spaced 1.8 mm apart on the bottom of the PCB was measured using an electrochemical impedance spectrometer (nanoZ, White Matter LLC, Mercer Island, WA) in phosphate buffer solution (PBS1X). The device served as the active electrode and a silver wire served as the reference electrode. Ultrasonic characterization of the transducer was performed in a custom-built water tank. A capsule hydrophone (HGL-0400, Onda Corp., Sunnyvale, CA) with a 20 dB preamplifier (AH-2020, Onda Corp.) was mounted on a computer-controlled 2D translation stage (XSlide, VelMex Inc., Bloomfield, NY) and used to calibrate the output pressure and characterize the beam pattern of a 2.25 MHz single-element transducer (V323-SU, Olympus). Ultrasound transmission and communication sensitivity was verified in a smaller water tank using a manual translation and rotation stage (Thorlabs Inc.). The contours of the implantable device were patterned by UV laser on an extruded acrylic piece, and the implantable device was clamped to the acrylic stage with nylon screws. The position and angle of the transducer relative to the mote were manually adjusted until the voltage across the piezoelectric crystal was maximized. Cable capacitance and parasitic capacitance were carefully accounted for by adjusting the series capacitance in a high impedance probe (N2863B, Agilent).A current source (2400-LV, Keithley) was used to pass a current of varying current density through two 0.127 mm thick platinum wires (773000, AM systems) immersed in the water tank to generate an electric field in the water tank. The transceiver board consisted of a custom integrated circuit (IC) in a QFN-64 package. The custom integrated circuit (IC) achieved an on-chip 1.8 V-32 V charge pump efficiency of 33%, a system latency of 20 ns, and consumed 16.5 μJ for each transmit cycle (Tang et al., 2015). During receive mode, a high-voltage switch was closed to amplify the signal by 28 dB. Both operations were performed on-chip. The output signal from the chip was digitized by an off-chip 10-bit, 100 MHz analog-to-digital converter (ADC) (LTC2261-12, Linear Technology, Milpitas, CA). The output of the ADC was fed back to a field programmable gate array (FPGA) and USB 3.0 interrogation module (XEM6310-LX45, Opal Kelly, Portland, OR) and transmitted to a laptop computer. An FPGA-USB module was also used to serially program the IC.

[0327] To interrogate the implantable device, six 540 ns pulses were transmitted by the external transducer every 100 μs. See FIG. 27. These transmitted pulses reflected off the neural dust mote and generated backscattered pulses towards the external transducer. The reflected backscattered pulses were recorded by the same transceiver board. The received backscattered waveform shows four predefined regions. These are the pulses reflected from four distinct interfaces (FIG. 28D): 1) the water-polymer encapsulation boundary; 2) the top surface of the piezoelectric crystal; 3) the piezoelectric-PCB boundary; and 4) the back surface of the PCB. As expected, the backscattered amplitude of the signal reflected from the piezoelectric crystal (second region) changed as a function of the potential change at the derivation electrode. The reflected pulses from the other interfaces did not respond to the potential change at the derivation electrode. Importantly, the pulse from the other non-responsive region was used as a signal level reference to make the system robust to artifacts created by movement or heat (since the pulse reflected from all interfaces changes with physical or thermal disturbance of the neural dust mote, but only the pulse from the second region changes as a function of the electrophysiological signal). In a water tank, the system showed a linear response to changes in the electrode potential recording and a noise floor of ~0.18 mVrms (Figure 28E). The overall dynamic range of the system was limited by the input range of the transistor, which is >500 mV (i.e., there is only an incremental change in current when the transistor is fully on (the input exceeds its threshold voltage) or fully off). The noise floor rose with the measured power drop of the beam. A misaiming of 0.7 mm caused the beam to drop by a factor of 2 (N=5 devices; Figure 28F). This increased noise floor caused by lateral misaiming poses the greatest challenge to performing neural recordings without a beam steering system (i.e., without an external transducer that can focus the ultrasound beam onto the implanted dust mote and therefore be aligned on-axis), where the implanted device converted input acoustic power into electrical power across a resistive piezoelectric load with an efficiency of ~25%.Figure 28G plots the off-axis voltage and power drop at Rayleigh distances for one of the transducers used in this example, while Figure 28H plots the change in effective noise floor as a function of angular misaiming.

[0328] EMG and ENG can be recorded in vivo in rodents without restraint. EMG responses from the gastrocnemius muscle of anesthetized adult Long-Evans rats were recorded using an implantable device system. The implantable device ("Dust") was placed on the surface of the exposed muscle, then the skin and surrounding connective tissue were replaced, and the wound was closed with surgical sutures (Figure 29A). The ultrasound transducer was placed 8.9 mm (one Rayleigh distance of the external transducer) from the implant, and a commercial ultrasound gel (Aquasonic 100, Parker Labs, Fairfield, NJ) was used to improve coupling. A manual manipulator was used to align the system by maximizing the voltage obtained on the piezoelectric crystal as measured from the flexible lead wires. For bulk stimulation of muscle fiber responses, an Ag / AgCl wire hook electrode was placed approximately 2 cm distal to the trunk of the sciatic nerve. Stimulation pulses lasting 200 μs were applied every 6 seconds, and data were recorded 20 ms before and after the stimulation window (Figure 29B). The power spectral density (PSD) of the reconstructed data, which has some harmonics due to edges in the waveform, is shown in Figure 29C. This process can continue indefinitely, within the limits of the anesthesia protocol. Comparison of data after 30 minutes of continuous recording showed no appreciable degradation in recording quality (Figure 29D).

[0329] EMG recruitment curves were obtained by varying the stimulus amplitude for both ground truth and wireless dust backscatter (FIGS. 30A and 30B). The reconstruction of the EMG signal from wireless backscatter data was sampled at 10 kHz, while the wired ground truth measurements were sampled at 100 kHz with a noise floor of 0.02 mV. The two signals at the response saturation stimulus amplitude (100%) matched with R=0.795 (FIG. 30C). The difference between the wireless and wired data was within ±0.4 mV (FIG. 30D). The salient features of the implanted device EMG response were approximately 1 ms narrower than the ground truth, which resulted in the largest error in the difference plots (FIGS. 30C and 30D). The response from the skeletal muscle fibers occurred 5 ms after stimulation and lasted for 5 ms. The EMG peak-to-peak voltage shows a sigmoidal response as a function of stimulus intensity (FIG. 30E). Error bars indicate the measurement uncertainty obtained from two rats and 10 samples (each for each stimulation amplitude). The minimum signal detected by the implanted device was approximately 0.25 mV, which is in good agreement with the noise floor measurements made in a water tank.

[0330] A similar setup was used to measure electroneurogram (ENG) responses from the main branch of the sciatic nerve in anesthetized rats. The sciatic nerve was exposed by releasing the hamstring muscles, and a neural dust mote was placed and sewn to the nerve, with the lead electrode in contact with the nerve sheath. By varying the amplitude of the stimulation current sent to the bipolar stainless steel electrodes on the leg, similar grades of response were measured for both ground truth and radio backscatter (Figures 31A and 31B). The two signals at the response saturation stimulation amplitude (100%) matched with R=0.886 (Figure 31C). The average error was within ±0.2 mV (Figure 31D). The peak-to-peak ENG voltage showed a sigmoidal response, and the error bars of the sigmoidal response show the uncertainty obtained from two rats and 10 samples (each for each stimulation amplitude). The minimum signal detected by the implanted device was again 0.25 mV (Figure 31E).

[0331] The experimental setup and surgical procedures are described in further detail. All animal procedures were performed in accordance with University of California Berkeley Animal Care and Use Committee regulations. Adult male Long-Evans rats were used for all experiments. Before starting the surgery, the animals were anesthetized with a mixture of ketamine (50 mg / kg) and xylazine (5 mg / kg) IP. The fur over the surgical site was shaved and cleaned. For EMG recording, a portion of the gastrocnemius muscle measuring approximately 10 mm × 5 mm was exposed by removing the overlying skin and fascia. The implantable device was then placed over the exposed muscle, the skin and fascia were replaced, and the wound was closed with 5 / 0 surgical sutures. For ENG recording, an incision was made in the knee from the sciatic notch and the sciatic nerve was exposed by releasing the hamstring muscles. The implantable device was then placed in contact with the nerve sheath of the main branch of the sciatic nerve bundle and sewn to the nerve using 10 / 0 microsurgical sutures. At the end of the experiment, the animals were euthanized. A continuous current stimulus was delivered using an isolated pulse stimulator (2100, AM Systems). Stimuli were delivered at various current amplitudes using a single biphasic pulse with a pulse width of 2 ms. Electrophysiological responses from 10 stimuli (i.e., samples) were recorded per experiment. An FPGA-USB module generated triggers for the stimulator every 6 s. For EMG experiments, a bipolar Ag-AgCl hook electrode placed around the trunk of the sciatic nerve was used for stimulation. To evoke ENG activity, a 28 G stainless steel needle electrode was placed on the leg with an electrode spacing of approximately 5 mm. The wired signal was amplified (100x) with a battery-powered operational amplifier with a built-in bandpass filter (DAM50, WPI, Sarasota, FL) set to 10 Hz–1 kHz. The ground reference for the amplifier was a 28 G stainless steel needle electrode placed on the leg contralateral to the recording setup. The amplifier output was connected to a multichannel digitizer, sampled at 100 kHz, and recorded on a computer. The implantable device was placed 1 Rayleigh distance (8.9 mm) from the transducer, which corresponds to a 5.9 μs transmission time, assuming an acoustic velocity in water of ∼1500 m / s.Six cycles of a 1.85 MHz sine wave with a peak voltage of 5 V were emitted every 100 μs (pulse repetition frequency (PRF) 10 kHz). The total transmit pulse width was approximately 3.3 μs, which was small enough not to overlap with the first voltage measurement taken at 5.9 μs. Given that the first reflection (i.e., backscatter) returning to the transducer occurred at approximately 11.8 μs (twice the transmit time) and lasted for 3.3 μs, the maximum PRF (i.e., in this context, sample rate) was ~66 kHz. As bulk peripheral nerve responses occur below 1 kHz, a PRF of 10 kHz was chosen to adequately capture the changes. To sample the backscatter waveform at 1.85 MHz without loss of signal fidelity, the off-chip ADC on the transceiver board was strongly oversampled at 50 MHz. This resulted in ~8 Mbits of data for 10 ms of neural recording, which was stored in a 128 Mbyte, 16-bit wide, synchronous DDR2 DRAM (MT47H64M16HR-3, Micron Technology, Boise, ID). After recording, the raw waveforms were sent to a laptop computer via a USB interface. The raw waveforms were simultaneously recorded using an 8-bit digitizer (USB-5133, National Instruments, Santa Clara, CA) for comparison. The raw backscatter waveforms sampled at 50 MHz from each experiment were sliced, time-aligned, and sample-averaged. The averaged signals were band-pass filtered from 10 Hz to 1 kHz using a symmetrical fourth-order Butterworth filter. Distinct features of the backscatter waveform (Figure 28D) were used as templates to identify regions of interest. The signals were then rectified and the integral of the region was calculated to evaluate the input voltage signal. The input voltage signal showed a linear response (Figure 28E). The signal multiplication factors were extracted from ground truth measurements.

[0332] In vivo ultrasound transmission. Six pulses were generated at 1.85 MHz using a 2.25 MHz single-element transducer (V323-SU, Olympus NDT, Waltham, MA). The transducer had a measured full width at half maximum (HPBW) of >2.5 MHz. To measure transmission loss through tissue, various thicknesses of skin found near the gastrocnemius muscle of male Long-Evans rats were placed between the transducer and the implanted device. Voltage was obtained at the piezoelectric crystal in the presence and absence of tissue. 8.9 mm of tissue resulted in tissue attenuation of 10 dB.

[0333] ENG recordings with different electrode spacing. Derivation electrodes with various spacings were fabricated on a 50 μm thick polyimide flexible printed circuit board (PCB). There were a total of five electrodes, each with a size of 0.2 mm × 0.2 mm, one of which was used as a reference electrode. The other electrodes were spaced 0.3 mm, 0.8 mm, 1.3 mm, and 1.8 mm away from the reference electrode, respectively. The spacing board was placed against the nerve sheath of the main branch of the sciatic nerve bundle (distal) and sewn to the nerve. A bipolar Ag-AgCl hook electrode placed around the trunk of the sciatic nerve (proximal) was used for stimulation. Continuous current stimulation of a single biphasic pulse lasting 0.5 ms every 1 s was performed using an insulated pulse simulator (2100, AM Systems, Sequim, WA). Signals recorded with various electrode spacings were amplified (100x) with a battery-powered differential amplifier with a built-in bandpass filter (DAM50, WPI, Sarasota, FL) set between 10 Hz and 1 kHz (Figure 32A). As expected, the peak-to-peak voltage recorded at the electrodes increased at least quadratically with spacing. The amplitude saturated after a spacing of 1.3 mm, confirming that an electrode spacing of 1.8 mm on the recording sensor was sufficient to capture the maximally saturated ENG response (Figure 32B).

[0334] Acoustic intensity calculations. To assess the safety of ultrasound systems, the American Society of Ultrasound in Medicine and the National Electronics Manufacturers Administration (NEMA) have established several parameters. The acoustic output of diagnostic ultrasound systems is limited by the attenuation of the spatial-peak pulse-average intensity (ISPPA), spatial-peak temporal-average intensity (ISPTA), and mechanical index (MI). These attenuations are calculated by multiplying the measurements in water by an attenuation constant of 0.3 dB / cm / MHz to simulate the effect on tissue. The output pressure of a 2.25 MHz single-element transducer (V323-SU, Olympus NDT) was calibrated using a capsule hydrophone (HGL-0400, Onda Corp) with a 20 dB preamplifier (AH-2020, Onda Corp., Sunnyvale, CA) mounted on a computer-controlled 2D translation stage (XSlide, VelMex Inc., Newton, NJ) and immersed in a custom-made water tank. Six cycles of a 1.85 MHz sine wave with a peak input voltage of 5 V were launched into the transducer every 1 ms (pulse repetition frequency (PRF) 10 kHz). The hydrophone was placed one Rayleigh distance (8.9 mm) away from the transducer. The pulse intensity integral (PII) was calculated as:

[0335]

number

[0336] where p is the instantaneous peak pressure and z0 is the characteristic acoustic impedance of the medium. For water, z0 is estimated to be 1.5 MRayl. I SPPA teeth

[0337]

number

[0338] where PD is the pulse duration defined as (t)(0.9×PII-0.1×PII)×1.25, as outlined by the standards established by NEWA. SPPA I SPPA = PII × PRF, where PRF is the pulse repetition frequency.

[0339]

number

[0340] where p r is the peak densification pressure and f is the acoustic frequency.

[0341] Alternatively, in vivo acute recordings in an anesthetized and immobilized rat model were used to retrieve compound action potentials from the main branch of the sciatic nerve and evoked EMG from the gastrocnemius muscle. The performance of the system was equivalent to conventional electrophysiological recordings using microelectrodes and cabled electronics. One of the major powers of the demonstrated technology is that, unlike conventional radio frequency technology, ultrasound-based systems can be scaled down to sizes of <100 μm, opening up new avenues of technology in implantable electronics. How small a high performance radio frequency receiver can be is physically constrained by the long wavelength (millimeters to centimeters) of radio frequency energy and the high degree of absorption of radio frequency energy in tissue (which heats the tissue and limits the total energy that can be delivered to the implant). Ultrasound systems work much more in both dimensions, allowing the design of extremely small receiving devices. Furthermore, the extremely miniaturized low power electronics allow useful recording electronics to fit into such a small package. A flat, thin piezoelectric transducer with appropriate impedance matching would allow for a wearable transceiver substrate small enough for the neurophysiology of awake, mobile rodents. Additionally, a wearable, battery-powered, multi-element array would allow for beam steering of the ultrasound beam. This has several advantages: 1) the implantable device can be kept on-axis despite relative movement between the implantable device and the external transducer; 2) a large number of implantable devices can be interrogated by electronically scanning the focused beam; and 3) it would be easier to adjust the position of the implantable device after surgery. Further denoising of the transceiver driving electronics would also help to reduce the noise floor. Furthermore, computational scaling predicts that implantable devices on the scale of <500 μm are feasible. To achieve this, there are numerous material and microfabrication challenges, including the use of microfabricated backplanes, solder microbump assembly of components (instead of the traditional wire-bonding approach used here) and thin-film encapsulants such as Parylene (instead of medical grade epoxies).A transition from PZT piezoelectric crystals to biocompatible BaTiO3 single crystal transducers is also conceivable. Taken together, these developments may pave the way for longitudinal studies of neural and muscle tissue recordings.

[0342] Example 7: Digital communication between an implantable device and an interrogator The system, including the implantable device and the interrogator with transducer array, is verified using a benchtop setup that mimics an in vivo environment. The ultrasound coupling gel serves as a tissue phantom because its acoustic impedance is similar to that of the target biological tissue (approximately 1.5 MRayl). The implantable device has a bulk piezoelectric transducer that is directly connected to two electrodes that are in contact with the transducer. The implantable device is placed in the tissue phantom and the interrogator transducer array is coupled to the gel. Both elements are mounted on a precision-calibrated stage for precise placement. The transducer array is placed 14 mm away from the dust mote. This corresponds to a round-trip flight time of 18.6 μs assuming an acoustic velocity of 1,540 m / s in the ultrasound coupling gel. The transducer array is excited with six 1.8 MHz, 0-32 V square pulses and the backscattered signals are digitized with 2000 samples at 17 Msps and 12-bit resolution. For time domain backscatter inspection, the full backscatter waveform is filtered in real time on the device and sent to the client via a wired serial connection. In normal operation, a full modulation extraction algorithm is applied in real time to the backscatter data on the device to compress the backscatter signal to 4 bytes. The processed data is sent to the remote client via Bluetooth® SSP protocol and streamed in real time via a GUI.

[0343] Figure 33A shows the filtered backscattered signal collected by the experimental setup described above. The signal is collected while the dust mote's piezoelectric crystal electrodes are in the shorted and open configurations. The impedance change due to the switch activity results in a backscattered peak amplitude that is 11.5 mV larger in the open switch configuration and a modulation depth of 6.45% (Figure 33B). The longer echo from the mote indicates that the transducer is resonating despite the weakened backing layer. As long as the backscatter from the implanted device is captured in the ROI, while the poorly weakened transducer system response spreads out the backscattered signal in the time domain, the modulation is successful.

[0344] The backscatter sensor mote is modulated to send a predetermined 11-character ASCII message ("hello world") using pulse amplitude modulation non-return to zero level coding. Modulation of the device's acoustic impedance is achieved by shorting the piezoelectric transducer to a digitally controlled switch where a high level corresponds to an open configuration and a low level corresponds to a closed configuration. Figure 34 shows the modulation values ​​for the transducer and the corresponding extracted modulation values ​​for the interrogator. The absolute value and noise margin of the extracted signal value depends on various factors (e.g., distance, direction, size of the mote). However, the extracted waveform continues to represent the modulated signal to the dust mote and scales linearly.

[0345] Wireless transmission of the extracted backscatter values ​​of the implanted device modulated with "hello world" demonstrates real-time communication of the device with the implanted device. Interrogation of the two-state backscatter system firmly demonstrates the system's wireless communication link with both the implanted sensor and a remote client. This wireless communication link invites development towards a closed-loop neuromodulation system connecting the brain with external devices. [Brief description of the drawings]

[0346] [Figure 1] FIG. 1 is a schematic diagram of the Neural Dust system, including an external transceiver, a subdural interrogator, and a Neural Dust mote, as disclosed in Seo et al., Neural dust: an ultrasonic, low power solution for chronic brain-machine interfaces, arXiv: 1307.2196v1 (July 8, 2013). [Figure 2A] 2A is a block diagram of an exemplary interrogator in the systems disclosed herein. The illustrated interrogator includes an ultrasonic transducer array including multiple ultrasonic transducers. Each ultrasonic transducer in the ultrasonic transducer array is activated by a channel. The channel includes a switch that alternately configures the transducer to receive or transmit ultrasonic waves. [Figure 2B] 2B is a schematic diagram of another exemplary interrogator in the system disclosed herein. The illustrated interrogator includes two ultrasonic transducer arrays, each of which includes a plurality of ultrasonic transducers. The interrogator also includes an integrated circuit. The integrated circuit may include digital circuitry, which may include a processor. The integrated circuit is connected to a user interface (which may include a display, a keyboard, buttons, etc.), a storage medium (i.e., non-transitory storage), an input / output device (which may be wireless (e.g., Bluetooth®)), and a power source (e.g., a battery). [Figure 3A] 3A depicts a block diagram of an exemplary interrogator that may be worn by a subject. The interrogator includes a wireless communication system (in the illustrated example, a Bluetooth® radio). The wireless communication system may be used to communicate with a computer system. [Figure 3B] 3B depicts an exploded view of a wearable interrogator that includes a battery, a wireless communication system, and a transducer array. [Figure 3C] FIG. 3C depicts the wearable interrogator shown in FIG. 3B fully assembled with a harness for attachment to a subject. [Figure 3D] 3D shows a wearable interrogator attached to a subject, i.e., a rodent (but it could be any animal, such as a human, dog, cat, horse, cow, pig, sheep, goat, chicken, monkey, rat, mouse, etc.). The interrogator includes a transducer array. The transducer array is secured to the subject's body with an adhesive. [Figure 3E] FIG. 3E shows a cross-sectional view of the transducer array of the interrogator shown in FIGS. 3A to 3D. [Figure 4] 4 is a schematic diagram illustrating communication between a transducer in an interrogator and an implantable device having a miniaturized ultrasonic transducer. The interrogator transmits ultrasonic waves to the implantable device, the miniaturized ultrasonic transducer emits ultrasonic backscatter that is modulated by a sensor, and the interrogator receives the backscatter. [Diagram 5]FIG. 5A shows a series of cycles of ultrasonic pulses emitted from the interrogator. When the transceiver board of the interrogator receives a trigger from the interrogator (e.g., FPGA), it generates a series of transmit pulses. At the end of the transmit cycle, the switch on the ASIC disconnects the transmit module and connects to the receive module. Each cycle is 100 microseconds. FIG. 5B shows an enlarged view of the transmit pulse sequence (i.e., one cycle) shown in FIG. 5A. In the cycle, there are six ultrasonic pulses of 1.85 MHz, and the ultrasonic pulses are repeated every 540 nanoseconds. FIG. 5C shows ultrasonic backscatter emitted from the implantable device. The ultrasonic backscatter arrives at the interrogator transducer at about 2t Rayleigh. FIG. 5D shows an enlarged view of the ultrasonic backscatter, which can be analyzed. The analysis of the ultrasonic backscatter includes the steps of filtering, rectifying, and integrating the ultrasonic backscatter wave. FIG. 5E shows an enlarged view of the filtered ultrasonic backscatter wave. The backscattered waves include a responsive region that responds to changes in impedance to the miniaturized ultrasonic transducer and a non-responsive region that does not respond to changes in impedance to the miniaturized ultrasonic transducer. [Figure 6] FIG. 6 illustrates one embodiment of an implantable device having a miniaturized transducer and a modulation circuit configured to modulate the current flowing in the miniaturized ultrasound transducer based on an electrophysiological signal sensed by an electrode pair. [Figure 7A] FIG. 7A illustrates one embodiment of a miniaturized ultrasound transducer (referred to as a "piezoelectric element") connected to an ASIC. The ASIC has a power supply circuit and a modulation circuit (or "backscatter circuit"). The power supply circuit has an energy storage capacitor ("cap"). The electrodes may be embedded in tissue. [Figure 7B] 7B illustrates an implantable device having a miniaturized ultrasound transducer, an integrated circuit, and an electrode pair. The integrated circuit has a modulation circuit, an AC composite amplifier string, and a power supply circuit. The power supply circuit has a full wave rectifier, a voltage doubler, a reference, and a regulator. [Figure 7C] FIG. 7C illustrates an example rectifier that may be used in the integrated circuit shown in FIG. 7B. [Figure 7D] FIG. 7D illustrates an exemplary amplifier string that may be used in the integrated circuit shown in FIG. 7B. [Figure 8] 8A shows a schematic diagram of an exemplary implantable device including a miniaturized ultrasound transducer and an ASIC on a printed circuit board (PCB). FIG. 8B shows a schematic diagram of another exemplary implantable device including a miniaturized ultrasound transducer and an ASIC on a printed circuit board (PCB). [Figure 9] FIG. 9 illustrates a method of manufacturing the implantable devices described herein. [Figure 10] FIG. 10 is a flow chart of a method for encapsulating an implantable device comprising amorphous silicon carbide. [Figure 11] Figure 11A shows different shaped vias used to connect components of an implantable device, and Figure 11B shows a serpentine trace arrangement for a deformable interconnect. [Figure 12] Figure 12 shows the time vs. temperature relationship for curing silver epoxy, which is one example of a material for attaching wire bonds in the manufacture of implantable devices. [Figure 13] Figure 13 shows electroneurograms (ENG) recorded using the implanted device. The dotted line shows the signal recorded by the ground truth electrodes. From the obtained data, general features including compound action potentials were reconstructed. The general features match the ground truth features. [Figure 14] FIG. 14 shows a schematic for encapsulating an implantable device in silicon carbide. [Figure 15] Figure 15 shows the prototype assembly schematic and PCB. [Figure 16A]Figures 16A-E show the process steps to ensure a miniaturized ultrasonic transducer (PZT) of the desired size is assembled onto a PCB. In Figure 16A, epoxy solder paste is placed onto the substrate. In Figure 16B, the piezoelectric material is attached to the PCB. In Figure 16C, the piezoelectric material is diced to form a bulk piezoelectric ultrasonic transducer of the desired size. In Figure 16D, the ultrasonic transducer is wire bonded to the PCB. In Figure 16E, the PCB and ultrasonic transducer are encapsulated in PDMS. [Figure 16B] Figures 16A-E show the process steps to ensure a miniaturized ultrasonic transducer (PZT) of the desired size is assembled onto a PCB. In Figure 16A, epoxy solder paste is placed onto the substrate. In Figure 16B, the piezoelectric material is attached to the PCB. In Figure 16C, the piezoelectric material is diced to form a bulk piezoelectric ultrasonic transducer of the desired size. In Figure 16D, the ultrasonic transducer is wire bonded to the PCB. In Figure 16E, the PCB and ultrasonic transducer are encapsulated in PDMS. [Figure 16C] Figures 16A-E show the process steps to ensure a miniaturized ultrasonic transducer (PZT) of the desired size is assembled onto a PCB. In Figure 16A, epoxy solder paste is placed onto the substrate. In Figure 16B, the piezoelectric material is attached to the PCB. In Figure 16C, the piezoelectric material is diced to form a bulk piezoelectric ultrasonic transducer of the desired size. In Figure 16D, the ultrasonic transducer is wire bonded to the PCB. In Figure 16E, the PCB and ultrasonic transducer are encapsulated in PDMS. [Figure 16D]Figures 16A-E show the process steps to ensure a miniaturized ultrasonic transducer (PZT) of the desired size is assembled onto a PCB. In Figure 16A, epoxy solder paste is placed onto the substrate. In Figure 16B, the piezoelectric material is attached to the PCB. In Figure 16C, the piezoelectric material is diced to form a bulk piezoelectric ultrasonic transducer of the desired size. In Figure 16D, the ultrasonic transducer is wire bonded to the PCB. In Figure 16E, the PCB and ultrasonic transducer are encapsulated in PDMS. [Figure 16E] Figures 16A-E show the process steps to ensure a miniaturized ultrasonic transducer (PZT) of the desired size is assembled onto a PCB. In Figure 16A, epoxy solder paste is placed onto the substrate. In Figure 16B, the piezoelectric material is attached to the PCB. In Figure 16C, the piezoelectric material is diced to form a bulk piezoelectric ultrasonic transducer of the desired size. In Figure 16D, the ultrasonic transducer is wire bonded to the PCB. In Figure 16E, the PCB and ultrasonic transducer are encapsulated in PDMS. [Figure 17] FIG. 17 shows a schematic diagram of electrical impedance measurement using a vector network analyzer (VNA). [Figure 18A] FIG. 18A shows that the measured power transfer efficiency in bulk piezoelectric ultrasonic transducers of different sizes is consistent with the simulated behavior. [Figure 18B] FIG. 18B shows that the measured impedance spectroscopy of the PZT crystal is consistent with the simulation. [Figure 18C] FIG. 18C shows that the frequency response of the miniaturized ultrasonic transducer versus the resulting power is approximately 6.1 MHz. [Figure 19] FIG. 19 is a schematic diagram of an exemplary ultrasonic transducer that can be used as part of an interrogator. [Figure 20]20 is a schematic diagram of an acoustic characterization setup using a calibrated ultrasonic transducer for verification of power delivery. The ultrasonic receiver is separate from the ultrasonic transmitter. [Figure 21A] FIG. 21A shows the output of a 5 MHz transducer as the hydrophone is moved away from the transducer surface. [Figure 21B] FIG. 21B shows that the peak of the abatement output is shifted to the left compared to the peak in water. [Figure 22A] FIG. 22A shows an XZ cross section of the transducer output, illustrating the Rayleigh distance and the sharp transition from near to far distance propagation. [Figure 22B] FIG. 22B shows an XY beam cross section showing the 6 dB bandwidth of a 2.2 mm beam. [Figure 23A] Figure 23A shows the focused 2D beam pattern from the transducer array in the XY plane. The measured beam is close to the simulated beam in both the X and Y directions. [Figure 23B] FIG. 23B shows the delay time applied to each transducer element in an ultrasound transducer array. [Figure 23C] FIG. 23C shows the simulated two-dimensional XZ cross-sectional beam pattern. [Figure 24A] FIG. 24A illustrates beam steering of an ultrasound beam transmitted from a transducer array. [Figure 24B] Below each beam pattern is the delay required for each transducer in the transducer array to obtain the measured beam pattern, as shown in FIG. 24B. [Figure 24C] Figure 24C shows the one-dimensional beam patterns in the X-axis for each of the beam patterns shown in Figure 24A. The measured beam patterns are very close to the simulated beam patterns. [Diagram 25] FIG. 25 shows the coupling efficiency and received power of a simulated miniaturized ultrasound transducer in tissue at a transmission distance of 5 mm. [Figure 26A] 26A-D show an overview of an exemplary system with an implantable device. FIG. 26A shows an external transducer that powers and communicates with an implantable device placed locally in the body. Driven by a custom transceiver board, the transducer alternates between sending a series of pulses that power the device and receiving reflected pulses that are modulated by the electrophysiological signal. FIG. 26B shows the implantable device fixed to the sciatic nerve of an anesthetized rat. The insert in FIG. 26B shows the implantable device with optional test leads. FIG. 26C shows components of an exemplary implantable device. The implantable device is assembled on a flexible PCB and includes a piezoelectric crystal, one custom transistor, and a recording electrode pair. FIG. 26D shows a close-up of the implantable device with calibration leads for measuring the electrophysiological signal (ground truth) and voltage obtained on the piezoelectric crystal on a flexible PCB. The calibration leads were removed during in vivo experiments. [Figure 26B] 26A-D show an overview of an exemplary system with an implantable device. FIG. 26A shows an external transducer that powers and communicates with an implantable device placed locally in the body. Driven by a custom transceiver board, the transducer alternates between sending a series of pulses that power the device and receiving reflected pulses that are modulated by the electrophysiological signal. FIG. 26B shows the implantable device fixed to the sciatic nerve of an anesthetized rat. The insert in FIG. 26B shows the implantable device with optional test leads. FIG. 26C shows components of an exemplary implantable device. The implantable device is assembled on a flexible PCB and includes a piezoelectric crystal, one custom transistor, and a recording electrode pair. FIG. 26D shows a close-up of the implantable device with calibration leads for measuring the electrophysiological signal (ground truth) and voltage obtained on the piezoelectric crystal on a flexible PCB. The calibration leads were removed during in vivo experiments. [Figure 26C] 26A-D show an overview of an exemplary system with an implantable device. FIG. 26A shows an external transducer that powers and communicates with an implantable device placed locally in the body. Driven by a custom transceiver board, the transducer alternates between sending a series of pulses that power the device and receiving reflected pulses that are modulated by the electrophysiological signal. FIG. 26B shows the implantable device fixed to the sciatic nerve of an anesthetized rat. The insert in FIG. 26B shows the implantable device with optional test leads. FIG. 26C shows components of an exemplary implantable device. The implantable device is assembled on a flexible PCB and includes a piezoelectric crystal, one custom transistor, and a recording electrode pair. FIG. 26D shows a close-up of the implantable device with calibration leads for measuring the electrophysiological signal (ground truth) and voltage obtained on the piezoelectric crystal on a flexible PCB. The calibration leads were removed during in vivo experiments. [Figure 26D] 26A-D show an overview of an exemplary system with an implantable device. FIG. 26A shows an external transducer that powers and communicates with an implantable device placed locally in the body. Driven by a custom transceiver board, the transducer alternates between sending a series of pulses that power the device and receiving reflected pulses that are modulated by the electrophysiological signal. FIG. 26B shows the implantable device fixed to the sciatic nerve of an anesthetized rat. The insert in FIG. 26B shows the implantable device with optional test leads. FIG. 26C shows components of an exemplary implantable device. The implantable device is assembled on a flexible PCB and includes a piezoelectric crystal, one custom transistor, and a recording electrode pair. FIG. 26D shows a close-up of the implantable device with calibration leads for measuring the electrophysiological signal (ground truth) and voltage obtained on the piezoelectric crystal on a flexible PCB. The calibration leads were removed during in vivo experiments. [Figure 27] FIG. 27 illustrates the communication between an exemplary interrogator and an implantable device. The top of FIG. 27 is a schematic of the information flow. The bottom of FIG. 27 shows the time course of signals at each step represented by the diagram shown at the top of the figure. In FIG. 27A, the FPGA from the interrogator generates a trigger signal to start recording. FIG. 27B shows the extracellular electrophysiological potential sent to the recording electrodes on the implantable device. FIG. 27C shows the transceiver board generates a series of transmit pulses upon receiving the trigger from the FPGA. At the end of the transmit cycle, a switch on the interrogator ASIC disconnects from the transmit module and connects with the receive module. FIG. 27D shows a close-up of the successive transmit pulses, showing 6 pulses at 1.85 MHz. FIG. 27E shows the backscatter from the implantable device. The backscatter reaches the transducer at about 2t Rayleigh distance. FIG. 27F shows a close-up of the backscatter waveform. The backscatter waveform contains a large saturating signal, whose overlap with the transmit pulse feeds through electrically and is ignored. As the backscatter pulse returns, it can be seen following the transmit window. FIG. 27G shows the filtered and rectified backscatter waveform and the area under the curve calculated to reconstruct the waveform. FIG. 27H shows the reconstructed waveform sampled at 10 kHz. Each portion of the reconstructed waveform is calculated by calculating the area under the curve of the properly reflected pulse received every 100 μs. [Figure 28A] FIG. 28A shows the reduced normalized peak pressure as a function of distance from the surface for an exemplary transducer (interrogator), with a reduced focus at approximately 8.9 mm at 1.85 MHz. [Figure 28B] Figure 28B shows the XY cross-sectional beam patterns at near, Rayleigh, and far distances and the corresponding 1D voltage plot at y = 0. These patterns show that the beam is focused at the Rayleigh distance. [Figure 28C] FIG. 28C shows that the output pressure of the transducer is a linear function of the input voltage (up to 32V peak-to-peak voltage). [Figure 28D] FIG. 28D (a reproduction of FIG. 5E shown in a different context) is an exemplary backscatter waveform shown in different regions. It can be seen that the backscatter waveform is flanked (in time) by regions corresponding to reflections originating from non-responsive regions; these regions correspond to pulses reflected from other device components. Measurements in non-responsive regions may be taken as a reference, where the non-responsive regions do not read biological data. This discriminatory measurement allows subtraction of the behavior of all components related to the external transducers in the experiment. [Figure 28E] FIG. 28E is a calibration curve obtained from the custom water tank fixture, showing a noise floor of 0.18 mVrms. [Figure 28F] FIG. 28F shows the effect of the noise floor as a function of lateral mispointing following the falloff in beam pattern power. [Figure 28G] FIG. 28G shows a one-dimensional plot of the off-axis voltage and output drop of the transducer at y=0 at the Rayleigh distance. [Fig. 28H] Figure 28H shows a plot of degradation versus angular mispointing. Angular mispointing distorts the beam pattern, i.e., the beam pattern becomes elliptical instead of circular. This lengthens the radius of the focal spot (and spreads the energy over a wider area). In other words, the distortion of the focal spot relaxes the constraint on mispointing. [Figure 29A] Figure 29A shows an in vivo experimental setup for recording EMG from the gastrocnemius muscle of a rat. An implantable device was placed on the exposed muscle surface and the wound was closed with surgical sutures. An external transducer couples ultrasound to the implantable device and wireless data is recorded and displayed on a computer system (e.g., a laptop computer). [Figure 29B] Figure 29B shows a comparison of ground truth measurements and reconstructed EMG signals across multiple trials. 20 ms samples were recorded with a 6 s interval between stimuli. [Figure 29C] Figure 29C shows the power spectral density (PSD) of the recorded EMG signal, which was 4.29e4μV2 / Hz and 3.11e4μV2 / Hz at 107Hz for the ground truth and reconstructed dust data, respectively, with some harmonics visible at the edges of the waveform. [Figure 29D] FIG. 29D shows that the radio backscatter data recorded at t=0 min and t=30 min agreed with R=0.901. [Figure 30A] FIG. 30A shows EMG signals of various intensities recorded in vivo with electrodes on a PCB and varying stimulation intensities. [Figure 30B] FIG. 30B shows that EMG responses of similar slopes were recorded wirelessly using the implanted device. [Figure 30C] FIG. 30C shows that EMG signals reconstructed from the ground truth and radio backscatter data agreed with R=0.795 for the response-saturating stimulation amplitude (100%) (R=0.60, 0.64, 0.67, and 0.92 for 54%, 69%, 77%, and 89%, respectively). [Figure 30D] In FIG. 30D, quantitative comparison showed agreement with the prominent features to less than 0.4 mV. [Figure 30E] In FIG. 30E, EMG peak-to-peak voltage showed the expected sigmoidal correlation with stimulation intensity. [Figure 31A] FIG. 31A shows ENG signals of various intensities recorded in vivo with electrodes on a PCB and varying stimulation intensities. [Figure 31B] FIG. 31B shows that an ENG response of similar slope was recorded wirelessly using a mote. [Figure 31C]FIG. 31C shows that the ENG signals reconstructed from the ground truth and radio backscatter data agreed with R=0.886 for the response-saturating stimulation amplitude (100%) (R=0.822, 0.821, 0.69, 0.918, 0.87 for 44%, 61%, 72%, 83%, and 89%, respectively). [Figure 31D] In FIG. 31D, quantitative comparison showed agreement with prominent features to within 0.2 mV. [Figure 31E] In FIG. 31E, ENG peak-to-peak voltages showed the expected sigmoidal correlation with stimulation intensity. [Figure 32A] FIG. 32A shows the ENG response recorded over time at different electrode spacings. [Figure 32B] FIG. 32B shows peak-to-peak ENG at various electrode spacings. [Figure 33A] 33A shows ultrasound backscatter from an implanted device with the device embedded in ultrasound coupling gel used to mimic tissue, including transmit penetration and exit centered at 26 microseconds, and miniaturized ultrasound transducer backscatter centered around 47 microseconds. [Figure 33B] Figure 33B is a close-up of the backscatter region (response area) of a miniaturized ultrasound transducer, showing that the signal input to the implantable device results in amplitude modulation. [Diagram 34] FIG. 34 shows digital data corresponding to the ASCII characters "hello world" wirelessly obtained from an implantable device by unipolar encoding and modulation of the backscattered amplitude pulse.

Claims

1. a first electrode and a second electrode attached to the tissue and configured to sense an electrophysiological signal; A storage medium storing the program; 1. An integrated circuit comprising a digital circuit and a modulation circuit, an integrated circuit, the digital circuitry configured to execute a program stored in the storage medium and operate the modulation circuitry to modulate a current based on the electrophysiological signal; an ultrasound transducer configured to transmit digital ultrasound backscatter encoding the electrophysiological signal from the tissue based on the modulated current; 1. An implantable device comprising:

2. a main body including the ultrasonic transducer and the modulation circuit; The maximum dimension of the body is 5 mm or less. The implantable device of claim 1 .

3. The first electrode and the second electrode are spaced apart by 0.5 mm or more. An implantable device according to claim 1 or 2.

4. Equipped with a non-responsive reflector, An implantable device according to any one of claims 1 to 3.

5. The tissue is muscle tissue or nerve tissue. An implantable device according to any one of claims 1 to 4.

6. The tissue is part of the peripheral nervous system or part of the central nervous system, An implantable device according to any one of claims 1 to 4.

7. The tissue is brain tissue or peripheral nerve tissue. An implantable device according to any one of claims 1 to 4.

8. The tissue is skeletal muscle, cardiac muscle or smooth muscle. An implantable device according to any one of claims 1 to 4.

9. having three or more electrodes, An implantable device according to any one of claims 1 to 8.

10. the digital circuitry is configured to transmit a digital signal to the modulation circuitry; the digital signal is based on the sensed electrophysiological signal; An implantable device according to any one of claims 1 to 9.

11. The implantable device of claim 10, which satisfies the following (i) or (ii): (i) the digital signal includes a unique embedded device identifier; (ii) the implantable device includes a plurality of electrode pairs, and the digital signal includes a unique electrode pair identifier.

12. the integrated circuit includes a spike detector; An implantable device according to any one of claims 1 to 11.

13. The integrated circuit includes a power supply circuit. An implantable device according to any one of claims 1 to 12.

14. the integrated circuit includes an analog-to-digital converter (ADC); An implantable device according to any one of claims 1 to 13.

15. The integrated circuit includes an amplifier array. An implantable device according to any one of claims 1 to 14.

16. a main body including the ultrasonic transducer and the modulation circuit; The volume of the body is 5 mm 3 Below is the An implantable device according to any one of claims 1 to 15.

17. The ultrasonic transducer is configured to receive ultrasonic waves that power the implantable device. An implantable device according to any one of claims 1 to 16.

18. The ultrasonic transducer is configured to receive ultrasonic waves from the interrogator, The interrogator includes one or more ultrasonic transducers. An implantable device according to any one of claims 1 to 17.

19. The ultrasonic transducer of the implantable device comprises a bulk piezoelectric transducer, a piezoelectric micromachined ultrasonic transducer (PMUT) or a capacitive micromachined ultrasonic transducer (CMUT). An implantable device according to any one of claims 1 to 18.

20. The implantable device is at least partially encapsulated by a biocompatible material. An implantable device according to any one of claims 1 to 19.

21. At least a portion of the first electrode and the second electrode are not encapsulated by the biocompatible material.

21. The implantable device of claim 20.

22. the integrated circuit is configured to compress data; An implantable device according to any one of claims 1 to 21.

23. The integrated circuit is configured to filter spurious electrophysiological signals. An implantable device according to any one of claims 1 to 22.

24. The modulation circuit includes an on / off switch. An implantable device according to any one of claims 1 to 23.

25. One or more implantable devices according to any one of claims 1 to 24; An interrogator comprising one or more ultrasonic transducers, an interrogator configured to transmit ultrasound to or receive ultrasound backscatter from one or more of the implantable devices; The system comprises:

26. The system includes a plurality of implantable devices.

26. The system of claim 25.

27. The interrogator is configured to beam steer transmitted ultrasound waves to alternately focus the transmitted ultrasound waves on a first portion of the plurality of implantable devices or a second portion of the plurality of implantable devices.

27. The system of claim 26.

28. the interrogator is configured to simultaneously receive ultrasound backscatter from two or more implantable devices; 27. The system of claim 26.

29. the interrogator is configured to transmit ultrasound to or receive ultrasound backscatter from the plurality of implantable devices; In the transmission or reception, time division multiplexing, spatial multiplexing or frequency multiplexing is utilized.

27. The system of claim 26.

30. The calling device is configured to be wearable by the subject. A system according to any one of claims 25 to 29.

Citation Information

Patent Citations

  • Biological information detector, biological information presentation system and electronic apparatus

    JP2015211801A

  • In-air ultrasonic rangefinding and angle estimation

    US20140253435A1

  • System and method for telemetrically providing intrabody spatial position

    US6239724B1