A periodic microstrip transmission line is presented, which maximizes phase delay through slow-wave propagation for use in various sensing applications. The dispersion and phase-delay characteristics are analyzed and compared analytically and through full-wave electromagnetic simulations in HFSS at 1GHz. Prototypes are fabricated and experimentally characterized, and measured results show close agreement with simulation and demonstrate improved sensitivity.

This paper presents a concept for a double negative metamaterial (DNM)-based antenna to simultaneously enhance Wireless Power Transfer (WPT) and reduce Specific Absorption Rate (SAR) here for a network of distributed brain microim-plants. The DNM copper coils are integrated in a FR-4 substrate, which has a dielectric constant of 4.3 and tangent loss $(\delta)$ of 0.025. Occupying a $2 \times 2\text{cm}^{2}$ area, the DNM structure is introduced into our target wireless brain-machine interface (BMI) system operating at 915 MHz. Preliminary HFSS simulations show it provides 2 dB WPT enhancement and a 20% SAR reduction. We believe the work has the potential to address the WPT/ SAR co-optimization challenges for biomedical implants in general.

Wireless sub-mm sized distributed brain implants have been proposed as the next frontier of Brain-Machine Interface (BMI) design to achieve untethered, high-density neural recording and stimulation. Simultaneously improving the wireless power transfer (WPT) efficiency and reducing the specific absorption rate (SAR) will be crucial for its clinical success. Towards these goals, we present an EM simulation method, a lumped equivalent circuit model, and a theoretical analysis to accurately predict the power delivered to the recording/ stimulating nodes, as well as the power dissipated in biological tissues and all other lossy elements within the system. This comprehensive framework also explains how increasing the distance between the transmit coil and the scalp can beneficially reduce the SAR without undermining the WPT efficiency. This work presents a rigorous prediction technique for transmission loss and tissue heating towards performance optimization.

Transmitting meaningful information into brain circuits by electronic means is a challenge facing brain-computer interfaces. A key goal is to find an approach to inject spatially structured local current stimuli across swaths of sensory areas of the cortex. Here, we introduce a wireless approach to multipoint patterned electrical microstimulation by a spatially distributed epicortically implanted network of silicon microchips to target specific areas of the cortex. Each sub-millimeter-sized microchip harvests energy from an external radio-frequency source and converts this into biphasic current injected focally into tissue by a pair of integrated microwires. The amplitude, period, and repetition rate of injected current from each chip are controlled across the implant network by implementing a pre-scheduled, collision-free bitmap wireless communication protocol featuring sub-millisecond latency. As a proof-of-concept technology demonstration, a network of 30 wireless stimulators was chronically implanted into motor and sensory areas of the cortex in a freely moving rat for three months. We explored the effects of patterned intracortical electrical stimulation on trained animal behavior at average RF powers well below regulatory safety limits. Transmitting information directly into the brain is a challenge for future brain-computer interfaces. Here, the authors present a patterned electrical microstimulation protocol using an epicortically-implanted network of silicon microchips to target specific areas of the cortex.

We introduce a wireless RF network concept for capturing sparse event-driven data from large populations of spatially distributed autonomous microsensors, possibly numbered in the thousands. Each sensor is assumed to be a microchip capable of event detection in transforming time-varying inputs to spike trains. Inspired by brain information processing, we have developed a spectrally efficient, low-error rate asynchronous networking concept based on a code-division multiple access method. We characterize the network performance of several dozen submillimeter-size silicon microchips experimentally, complemented by larger scale in silico simulations. A comparison is made between different implementations of on-chip clocks. Testing the notion that spike-based wireless communication is naturally matched with downstream sensor population analysis by neuromorphic computing techniques, we then deploy a spiking neural network (SNN) machine learning model to decode data from eight thousand spiking neurons in the primate cortex for accurate prediction of hand movement in a cursor control task.

Wireless sub-mm-sized distributed brain implants could revolutionize Brain-Machine Interface applications by providing untethered cortical recording and/ or stimulation with unprecedented density. A key to achieving this goal is the development of wireless power transfer (WPT) methods that operate in this spatially varying magnetic field. We demonstrate, through circuit simulations and theoretical analysis, over-voltage protection (OVP) realized by circuit resonance auto-tuning will significantly enhance the overall system WPT efficiency by 5dB over the traditional OVP voltage clamping method. The auto-tune algorithm is designed and measured in a completely integrated 65nm CMOS ASIC implant featuring necessarily low power and small size. Additionally, a solution is implemented to avoid an inherent cold start issue that would otherwise significantly constrain auto-tune WPT implementations.

The execution of complex, naturalistic neural tasks relies on the coordinated operation of cortical microcircuits across multiple related functional areas of the brain. Cortical BCI technologies aimed at accessing these distributed computations are thus anticipated to require a large number of spatially diverse, implanted electronic listening posts or nodes, appropriately positioned in physical proximity to the sources of these neural signals. From a neuroengineering perspective, key aspects of the specifications for a next-generation BCI system include considerations of the channel counts that may be safely implanted chronically in vivo subjects, as well as efficient approaches for physical implementations of large arrays of microscale electronic probes. Data rates for extracting brain signals at a useful resolution have to be contemplated in the context of designing a commensurate communication link facilitating low-latency forward transmission for decoding by external computing platforms. This must of course occur in concert with the reverse processes, whereby the same implanted probes would provide a means to "write-in" feedback information into the brain through injection of electronic signals directly into the cortex. This chapter reviews contemporary examples and recent accomplishments in the field, from the viewpoint of systems level engineering, and discusses both the challenges and opportunities ahead to build next generations of brain-computer interfaces.

We introduce a large-scale wireless RF network concept for capturing neural data from spatially distributed autonomous microchip implants. Each sensor chip is assumed to be capable of event detection by transforming time-varying input signals to spike trains. Inspired by the brain’s information processing, we have developed a spectrally efficient, low-error rate asynchronous networking protocol based on a code-division multiple access method. We test the network performance with dozens of sub-millimeter-size silicon microchips in situ on laboratory testbench, complemented by large scale in silico simulations. Through an application example relevant to brain-computer interfaces, we simulate the wireless transmission of spiking data from eight thousand neurons in the primate cortex for accurate hand movement prediction in a cursor control task.

Abstract We describe a wireless RF network for capturing event-driven data from thousands of spatially distributed sensors. As asynchronous devices, each sensor detects events within its local environment. Information acquired by the full network can enable prediction of the time evolution of the system, whether a brain or cardiac circuit in the human body, or an assistive living environment, for example. We develop a communication concept inspired by principles of synaptic information processing in the brain which we mimic by a code-division multiple access strategy in a sparse network. Through extensive simulation, we optimize wireless transmission from ensembles of event-detecting sensors for efficient use of the power and spectrum at low error rates, which is then implemented on-chip to demonstrate the core communication scheme in silico. We also apply the concept to recordings from thirty thousand neurons in the primate cortex, to decode and predict forward state trajectories for hand movement.

A major challenge to high-resolution, closed-loop Brain Computer Interfaces (BCIs) is the availability of implantable technologies facilitating vastly parallel, large-scale access to cortical neural data representing complex, naturalistic tasks or sophisticated therapeutic neuromodulation. The current technological bottleneck is scalability of systems employing intra or epicortical electrode arrays with hard-wired tethers and bulky implant packaging. We address these challenges by employing an approach relying on spatially-distributed, completely wireless clusters of autonomous microscale neural interfaces, where each microdevice provides a single bidirectional channel (read-out and write-in) of neural access, and occupies a volume <0.01 mm2 inclusive of biocompatible packaging for long-term implantation. Wireless power transfer, high-bandwidth bidirectional telecommunications and adaptive networking across multi-areal clusters are managed by a wearable external module to produce an implantable device system with anatomic flexibility and scalability, forming a “cortical internet”.

Multichannel electrophysiological sensors and stimulators—particularly those used to study the nervous system—are usually based on monolithic microelectrode arrays. However, the architecture of such arrays limits flexibility in electrode placement and scaling to a large number of nodes, especially across non-contiguous locations. Here we report wirelessly networked and powered electronic microchips that can autonomously perform neural sensing and electrical microstimulation. The microchips, which we term neurograins, have an ~1 GHz electromagnetic transcutaneous link to an external telecom hub, providing bidirectional communication and control at the individual device level. To illustrate the potential of the approach, we show that 48 neurograins can be individually addressed on a rat cortical surface and used for the acute recording of neural activity. Theoretical calculations and experimental measurements show that the link configuration could potentially be scaled to 770 neurograins using a customized time-division multiple access protocol.

ABSTRACTMultichannel electrophysiological sensors and stimulators, especially those used for studying the nervous system, are most commonly based on monolithic microelectrode arrays. Such architecture limits the spatial flexibility of individual electrode placement, posing constraints for scaling to a large number of nodes, particularly across non-contiguous locations. We describe the design and fabrication of sub-millimeter size electronic microchips (“Neurograins”) which autonomously perform neural sensing or electrical microstimulation, with emphasis on their wireless networking and powering. An ∼1 GHz electromagnetic transcutaneous link to an external telecom hub enables bidirectional communication and control at the individual neurograin level. The link operates on a customized time division multiple access (TDMA) protocol designed to scale up to 1000 neurograins. The system is demonstrated as a cortical implant in a small animal (rat) model with anatomical limitations restricting the implant to 48 neurograins. We suggest that the neurograin approach can be generalized to overcome many scalability issues for wireless sensors and actuators as implantable microsystems.

Scalability of implantable neural interface devices is a critical bottleneck in enhancing the performance of cortical Brain-Computer Interfaces (BCIs) through access to high density and multi-areal cortical signals. This is challenging to achieve through current monolithic constructs with 100-200 channels, often with bulky tethering and packaging, and a spatially distributed sensor approach has recently been explored by a few groups, including our laboratories [1]. In this paper, we describe a microscale (500 μm) programmable neural stimulator in the context of an epicortical wireless networked system of sub-mm "Neurograins" with wireless energy harvesting (near 1 GHz) and bidirectional telemetry. Stimulation neurograins are post-processed to integrate poly(3,4-ethylenedioxythiophene) polystyrene sulfonate (PEDOT:PSS) planar electrodes or intracortical penetrating microwires, and ensembles of microdevices are hermetically encapsulated using liquid-crystal polymer (LCP) thermocompression for chronic implantability. Radio-frequency power and telecommunications management are handled by a wearable external "Epidermal Skinpatch" unit to cater to chronic clinical implant considerations. We describe the stimulation neurograin performance specifications and proof-of-concept in bench top and ex vivo rodent platforms.

A hermetic sealing method of sub-millimeter-sized microelectronic chiplets for wireless body implants is presented by ultrathin and electromagnetically transparent atomic layer deposition (ALD) coatings. Fully 3D conformal encapsulation of wirelessly powered microdevices is demonstrated both with and without opening windows for electrophysiological measurements. The chiplets embedding custom application-specific integrated circuits (ASICs) with radio frequency (RF) transmitters are encapsulated by a stack of alternating layers of hafnium oxide and silicon dioxide to maximize impermeability of water and ionic penetration while minimizing the volume of the packaging material. The hermeticity of the devices is characterized through accelerated aging tests in saline at T = 87 degrees C, while continued functionality is monitored via evaluation of backscattered RF signals (near 1 GHz) to ascertain possible degradation and electronic failure. Earliest failures of wirelessly functional devices occur after more than 180 d of immersion at 87 degrees C. Wireless devices having opening windows through the ALD envelope show no signs of degradation for >100 d. This implies an equivalent lifetime >10 years at T = 37 degrees C. This approach is readily scalable to high throughput batch processing of hundreds of microchiplets, offering a methodology for hermetic packaging of microscale biomedical chronic implants.

A vastly enhanced capability to bi-directionally interface with cortical microcircuits in a clinically viable way is the ultimate aspiration in neuroengineering. This necessitates a paradigm shift in neural interface system design beyond current bulky, monolithic constructs which are challenging to scale past 100-200 channels due to anatomic and engineering design constraints. A neural interface system relying on a spatially-distributed network of wireless microscale implantable sensors offers a highly scalable, robust and adaptive architecture for next-generation neural interfaces. We describe the development of a wireless network of sub-mm, untethered, individually addressable, fully wireless "Neurograin" sensors, in the context of an epicortical implant. Individual neurograin chiplets integrate a ~ 1 GHz wireless link for energy harvesting and telemetry with analog and digital electronics for neural signal amplification, on-chip storage, and networked communications via a TDMA protocol. Each neurograin thus forms a completely self-contained single channel of neural access and is implantable after post-process atomic layer deposition of thin-film (100 nm thick) barriers for hermetic sealing. Finally, ensembles of implantable neurograins form a fully wireless cortico-computer communication network (utilizing their unique device IDs). The implanted network is coordinated by a compact external "Epidermal Skinpatch" RF transceiver and data processing hub, which is implemented as a wearable module in order to be compatible with clinical implant considerations. We describe neurograin performance specifications and proof-of-concept in bench top and ex vivo and in vivo rodent platforms.

To dramatically increase the scale and spatial resolution for future chronic electrocorticography (ECoG) applications, we propose a wireless brain-machine interface (BMI) system based on a high number (up to 1000) of freely distributed, sub-mm sized (0.25 mm 2 ) IC implants. The chip features an onchip antenna for RF energy harvesting at 900 MHz and data backscattering at 10 Mbps. In order to synchronize and time-multiplex the uplink data transmission of the untethered chips, while allowing their oscillators to free-run to save power, a robust Mbps ASK-PWM downlink data protocol based on digital counters was implemented. To the best of our knowledge, this paper presents the first experimental validation of simultaneous wireless power transfer and bi-directional RF data communications on a network of (32) brain implant ICs over a single inductive coupling link.

This brief studies the frequency drift due to temperature variation in LC CMOS quadrature oscillators. The quadrature oscillators, which can oscillate in two modes at different frequencies, have different temperature behavior in the two modes, with one mode showing lower sensitivity to temperature variation. Simulation results of a series coupled quadrature oscillator confirm that the mode with higher frequency is more stable to temperature variation. The measurement results of the fabricated chip in a 90-nm CMOS technology verifies the analysis. The optimized frequency drift is +/- 100 ppm across temperature (0-60 degrees C) for an oscillation frequency of 5 GHz with -118 dBc/Hz phase noise at 1 MHz offset frequency while consuming 8 mA from a power supply of 1.2 V.

We report a new correlation-based direct sequence spread-spectrum technique that estimates low-pass-equivalent Volterra coefficients for weakly nonlinear RF systems. The methodology provides robust coefficient estimates at test signal amplitudes well below the operating signal level during normal system operation, demonstrating that the technique can be used for “background” nonlinearity measurement. We demonstrate the methodology by measuring third-order Volterra kernels for a power amplifier chain operated in compression with a 10 MHz bandwidth signal at 1960 MHz.

A novel Brain-Machine Interface (BMI) system based on a distributed network of implantable wireless sensors was proposed. Small CMOS "Neurograin" chips (0.5x0.5 mm2) with on-chip antenna are designed to harvest near-field RF energy at ~1 GHz, and backscatter 10 Mbps BPSK modulated data asynchronously and periodically. A "Skinpatch" software-defined radio (SDR) receiver is realized on a commercial USRP running GNU Radio programs. It down-converts the reflected waves from the Neurograins and performs data recovery. In this BMI prototype demonstration, 32 Neurograins will be wirelessly powered, while a Skinpatch USRP will recover their backscattered packets in real-time.