Future high-density and high channel count neural interfaces that enable simultaneous recording of tens of thousands of neurons will provide a gateway to study, restore and augment neural functions. However, building such technology within the bit-rate limit and power budget of a fully implantable device is challenging. The wired-OR compressive readout architecture addresses the data deluge challenge of a high channel count neural interface using lossy compression at the analog-to-digital interface. In this article, we assess the suitability of wired-OR for several steps that are important for neuroengineering, including spike detection, spike assignment and waveform estimation. For various wiring configurations of wired-OR and assumptions about the quality of the underlying signal, we characterize the trade-off between compression ratio and task-specific signal fidelity metrics. Using data from 18 large-scale microelectrode array recordings in macaque retina ex vivo, we find that for an event SNR of 7-10, wired-OR correctly detects and assigns at least 80% of the spikes with at least 50× compression. The wired-OR approach also robustly encodes action potential waveform information, enabling downstream processing such as cell-type classification. Finally, we show that by applying an LZ77-based lossless compressor (gzip) to the output of the wired-OR architecture, 1000× compression can be achieved over the baseline recordings.

Current retinal prostheses provide electrical stimulation without feedback from the stimulated neurons. Incorporation of multichannel recording electronics would typically require trans-scleral cables for power supply and data transmission. In this work, we explore a wireless, optoelectronic, miniature, modular, and distributed electro-neural interface for recording, which we call Sensory Particles with Optical Telemetry (SPOT). It can be used in an advanced, bi-directional retinal prosthesis and other sensory applications. Emphasis is placed on the novel telemetry stage. SPOTs are powered by near-infrared light and transmit information by light. As a proof of concept, we designed and built a low-power, small-footprint linear transconductance circuit utilizing chopper stabilization in 130nm CMOS. Our design achieved 57 mS transconductance within 3.5 kHz bandwidth, and a near-infrared (NIR) power density of 0.5 mW/mm², well within the ocular and thermal safety limits. The telemetry circuit consumes 0.015 mm² area, and each SPOT can be powered by a single photovoltaic (PV) supply of area 0.0056 mm². Electrical spikes transmitted by an 850nm LED were detected with 15 dB SNR, at the output of the optical link.

Modeling linear periodically time-varying (LPTV) circuits is challenging due to the presence of frequency translation. Many approaches have been proposed that simplify the analysis and provide intuition into the operation of these circuits. It is critical to select the proper model when designing LPTV systems: too complex, and intuition is lost; too simple, and numerical accuracy degrades. This work shows how a conversion matrix-based model can be used for mixer-first receivers with complex feedback in the presence of finite switch transitions. This model accurately predicts S11 below -10 dB for all tested transition times, in contrast with prior models, which are shown to be invalid with transitions beyond 2% of the clock period. As a design tool, this approach models gain, harmonic rejection ratio, and noise figure within 0.1 dB of simulation with switch transitions even at 5% of the clock period.

The efficacy of wireless intracortical brain–computer interfaces (iBCIs) is limited in part by the number of recording channels, which is constrained by the power budget of the implantable system. Designing wireless iBCIs that provide the high-quality recordings of today’s wired neural interfaces may lead to inadvertent over-design at the expense of power consumption and scalability. Here, we report analyses of neural signals collected from experimental iBCI measurements in rhesus macaques and from a clinical-trial participant with implanted 96-channel Utah multielectrode arrays to understand the trade-offs between signal quality and decoder performance. Moreover, we propose an efficient hardware design for clinically viable iBCIs, and suggest that the circuit design parameters of current recording iBCIs can be relaxed considerably without loss of performance. The proposed design may allow for an order-of-magnitude power savings and lead to clinically viable iBCIs with a higher channel count.

Neural interfaces of the future will be used to help restore lost sensory, motor, and other capabilities. However, realizing this futuristic promise requires a major leap forward in how electronic devices interface with the nervous system. Next generation neural interfaces must support parallel recording from tens of thousands of electrodes within the form factor and power budget of a fully implanted device, posing a number of significant engineering challenges. In this paper, we exploit sparsity and diversity of neural signals to achieve simultaneous data compression and channel multiplexing for neural recordings. The architecture uses wired-OR interactions within an array of single-slope A/D converters to obtain massively parallel digitization of neural action potentials. The achieved compression is lossy but effective at retaining the critical samples belonging to action potentials, enabling efficient spike sorting and cell type identification. Simulation results of the architecture using data obtained from primate retina ex-vivo with a 512-channel electrode array show average compression rates up to  ∼ 40× while missing less than 5% of cells. In principle, the techniques presented here could be used to design interfaces to other parts of the nervous system.

Always-on sound classification is a desirable but power-intensive function for a variety of emerging Internet of Everything applications. This work explores the accuracy-complexity tradeoff by using summary statistics for classifying semi-stationary sounds. Compared to contemporary solutions including deep learning, this approach requires one to three orders of magnitude fewer parameters and can therefore be trained over ten times faster. We propose a mixed-signal design using N-path filters for feature extraction to further improve energy efficiency without incurring a large accuracy penalty for a binary classification task (less than 2.5% area reduction under receiver operating characteristic curve).

This letter presents a low-noise integrated potentiostat for affinity-free molecular detection in applications for personalized medicine. The affinity-free sensing technique uses a digital classifier to identify molecules through unique vibrational signatures. The sensing mechanism relies on coherent interference of electron wave functions at the interface between a nanoscale working electrode and a liquid electrolyte. Coherence at the sensing interface is enabled by low-noise feedback, which reduces the effective temperature of the electrons. The described three-channel potentiostat IC uses chopping and correlated double sampling to achieve an input-referred voltage noise of 12 nV/rt-Hz at 30 Hz and a current resolution of 1.8 pA_rms with 0.5-s averaging time. Each channel consumes 5 mW and occupies 0.41 mm² in 65-nm CMOS.

This paper describes an architecture for the massively parallel digitization of neural action potentials. The scheme achieves simultaneous data compression and channel multiplexing through wired-OR interactions within an array of single-slope A/D converters. The achieved compression is lossy but effective at retaining the critical samples belonging to action potential spikes. Simulation results using ex-vivo experimental data from a 512-channel array show compression rates up to ∼73x while maintaining ≥90% reconstruction coverage for parasol cells in the primate retina.

In this paper some local stability results for an autonomous Long Short-Term Memory neural network model with respect to the origin are provided. In particular, it is shown through linearization that the local asymptotic stability conditions with respect to the origin only depend on one of the weight matrices. Simulations indicate that these local stability conditions greatly influence the behavior of the autonomous four-dimensional neural network in the region where each variable's values vary between minus one and one. Finally, some sufficient stability conditions for the nonlinear model are formulated as a convex program involving linear matrix inequalities.