An Event-based Neural Compressive Telemetry with >11× Loss-less Data Reduction for High-bandwidth Intracortical Brain Computer Interfaces.

Intracortical brain-computer interfaces offer superior spatial and temporal resolutions, but face challenges as the increasing number of recording channels introduces high amounts of data to be transferred. This requires power-hungry data serialization and telemetry, leading to potential tissue damage risks. To address this challenge, this paper introduces an event-based neural compressive telemetry (NCT) consisting of 8 channel-rotating Δ-ADCs, an event-driven serializer supporting a proposed ternary address event representation protocol, and an event-based LVDS driver. Leveraging a high sparsity of extracellular spikes and high spatial correlation of the high-density recordings, the proposed NCT achieves a compression ratio of >11.4×, while consumes only 1 µW per channel, which is 127× more efficient than state of the art. The NCT well preserves the spike waveform fidelity, and has a low normalized RMS error <23% even with a spike amplitude down to only 31 µV.

Comparison and Integration of Voltage and Charge Amplifiers for Capacitive ECG Measurements.

OBJECTIVE: Sensing with capacitive electrodes is of interest for long-term, comfortable bio-potential measurements (e.g., ECG). However, due to the small body-to-electrode capacitance (Ce), the design of the associated front-end amplifier remains a challenge. Both voltage amplifiers (VA) and charge amplifiers (CA) can be employed. While basic comparisons of both typologies were done before, this paper extends the comparison to their responses to artifacts (caused by motion or interference). Further, a VA-CA-switchable amplifier is proposed, allowing to adapt the amplifier type to different situations, and enabling to estimate the body-to-electrode capacitance Ce in a passive way. METHODS: A VA-CA switchable amplifier was implemented in a 180 nm CMOS process. The responses to artifacts for VA and CA were studied by modelling, simulations and experiments using the custom IC. The proposed Ce estimation method was validated by electrical tests and in-vivo tests. RESULTS: VAs are less affected by Ce variation artifacts, while CAs recover faster from triboelectricity artifacts. In a VA, these two artifacts are multiplicative and get modulated if they occur simultaneously, but in a CA they remain independent. CONCLUSION: The combined VA-CA amplifier has the potential for optimal amplifier selection according to the properties of the recorded signal, the value of Ce and the actual presence of artifacts. Moreover, it can estimate Ce without extra hardware. SIGNIFICANCE: The proposed VA-CA switchable structure is superior to an individual VA or CA, thanks its adaptability to signal quality and artifacts, and it provides extra information on the body-to-electrode interface quality (Ce).

Analysis of Interface Material Noise in Non-contact Capacitive Sensing.

The thermal noise due to the resistivity of insulation materials can become a significant noise source in non-contact capacitive sensing, especially when measuring micro-volt-level physiological signals. Since both the impedance and the resistivity of practical insulation materials may be strongly frequency dependent, their thermal noise is often frequency dependent. This paper studies the impedance and noise behavior of different interface materials as function of frequency, by means of modelling, simulations, and experimental measurements. The results show that the inherent resistive noise of some fabrics (e.g., cotton, polyester) could outweigh the typical noise level of circuits for physiological sensing; and as a result, the interface noise can limit the quality of low-amplitude signal detection. Clinical Relevance- This study gives a guideline for material selection from the noise perspective in case of capacitive electrode sensing.

A Prototype System With Custom-Designed RX ICs for Contrast-Enhanced Ultrasound Imaging.

This work presents a prototype system based on a multichannel receiving (RX) integrated circuit (IC) for contrast-enhanced ultrasound (CEUS) imaging. The RX IC is implemented in a 40-nm low-voltage CMOS technology and is designed to interface to a capacitive micromachined ultrasonic transducer array. To enable a direct connection of the RX electronics to the transducer, an analog multiplexer with on-chip protection circuitry is developed. Stress tests confirm the reliability of this arrangement when combined with a high-voltage pulser. The RX IC is equipped with a highly programmable bandpass filter to capture harmonic signals from ultrasound contrast agents (UCAs) while suppressing fundamental components. In order to examine the impact of analog front-end (AFE) bandpass filtering, in vitro acoustic experiments are performed with UCAs. A spatial resolution analysis suggests that the AFE bandpass filtering combined with a pulse inversion (PI) technique can improve the lateral resolution by 38% or 9% compared to the original full-bandwidth approach or a stand-alone PI approach, respectively, while the impact on axial resolution is negligible. A phantom study shows that compared to digital bandpass filtering, the AFE bandpass filtering enables better use of the dynamic range of the RX electronics, resulting in better generalized contrast-to-noise ratio from 0.44/0.53 to 0.57/0.68 without or with PI.

An Implantable Neuromorphic Sensing System Featuring Near-sensor Computation and Send-on-Delta Transmission for Wireless Neural Sensing of Peripheral Nerves.

This paper presents a bio-inspired event-driven neuromorphic sensing system (NSS) capable of performing on-chip feature extraction and "send-on-delta" pulse-based transmission, targeting peripheral-nerve neural recording applications. The proposed NSS employs event-based sampling which, by leveraging the sparse nature of electroneurogram (ENG) signals, achieves a data compression ratio of >125×, while maintaining a low normalized RMS error of 4% after reconstruction. The proposed NSS consists of three sub-circuits. A clockless level-crossing (LC) ADC with background offset calibration has been employed to reduce the data rate, while maintaining a high signal to quantization noise ratio. A fully synthesized spiking neural network (SNN) extracts temporal features of compound action potential signals consumes only 13 µW. An event-driven pulse-based body channel communication (Pulse-BCC) with serialized address-event representation encoding (AER) schemes minimizes transmission energy and form factor. The prototype is fabricated in 40-nm CMOS occupying a 0.32-mm2 active area and consumes in total 28.2 µW and 50 µW power in feature extraction and full diagnosis mode, respectively. The presented NSS also extracts temporal features of compound action potential signals with 10-µs precision.

An RX AFE With Programmable BP Filter and Digitization for Ultrasound Harmonic Imaging.

This paper presents a front-end integrated circuit for ultrasound (US) harmonic imaging, interfacing to a one-dimensional capacitive micromachined ultrasonic transducer (CMUT). It contains a complete ultrasound receiving chain, from analog front-end (AFE) to gigabit/s data link. A two-stage self-biased inverter-based transimpedance amplifier (TIA) is proposed in this work to improve tradeoffs between power, noise, and linearity at the first stage. To improve harmonic imaging performance, the design is further equipped with a 4[Formula: see text]-order highly programmable bandpass filter, which has a tunable bandwidth from 2 MHz to 15 MHz. An 8 b 80 MS/s SAR ADC digitizes the signal, which is further encoded and serialized into an LVDS data link, enabling a reduction in the number of output cables for future systems with multiple ADCs. The design is realized in a 40 nm CMOS technology. Electrical measurements show it consumes 2.9 mW for the AFE and 2.1 mW for the ADC and digital blocks. Its overall dynamic range varies from 61 dB to 69 dB, depending on the reception bandwidth. The imaging capability of this design is further demonstrated in a US transmission and reception imaging system. The acoustic measurements prove successful ultrasound harmonic acquisition, where the on-chip bandpass filter can improve the lateral resolution by more than 30%.

Dedicated Algorithm for Unobtrusive Fetal Heart Rate Monitoring Using Multiple Dry Electrodes.

Multi-channel measurements from the maternal abdomen acquired by means of dry electrodes can be employed to promote long-term monitoring of fetal heart rate (fHR). The signals acquired with this type of electrode have a lower signal-to-noise ratio and different artifacts compared to signals acquired with conventional wet electrodes. Therefore, starting from the benchmark algorithm with the best performance for fHR estimation proposed by Varanini et al., we propose a new method specifically designed to remove artifacts typical of dry-electrode recordings. To test the algorithm, experimental textile electrodes were employed that produce artifacts typical of dry and capacitive electrodes. The proposed solution is based on a hybrid (hardware and software) pre-processing step designed specifically to remove the disturbing component typical of signals acquired with these electrodes (triboelectricity artifacts and amplitude modulations). The following main processing steps consist of the removal of the maternal ECG by blind source separation, the enhancement of the fetal ECG and identification of the fetal QRS complexes. Main processing is designed to be robust to the high-amplitude motion artifacts that corrupt the acquisition. The obtained denoising system was compared with the benchmark algorithm both on semi-simulated and on real data. The performance, quantified by means of sensitivity, F1-score and root-mean-square error metrics, outperforms the performance obtained with the original method available in the literature. This result proves that the design of a dedicated processing system based on the signal characteristics is necessary for reliable and accurate estimation of the fHR using dry, textile electrodes.

A Low-Voltage Chopper-Stabilized Amplifier for Fetal ECG Monitoring With a 1.41 Power Efficiency Factor.

This paper presents a low-voltage current-reuse chopper-stabilized frontend amplifier for fetal ECG monitoring. The proposed amplifier allows for individual tuning of the noise in each measurement channel, minimizing the total power consumption while satisfying all application requirements. The low-voltage current reuse topology exploits power optimization in both the current and the voltage domain, exploiting multiple supply voltages (0.3, 0.6 and 1.2 V). The power management circuitry providing the different supplies is optimized for high efficiency (peak charge-pump efficiency = 90%).The low-voltage amplifier together with its power management circuitry is implemented in a standard 0.18 µm CMOS process and characterized experimentally. The amplifier core achieves both good noise efficiency factor (NEF=1.74) and power efficiency factor (PEF=1.05). Experiments show that the amplifier core can provide a noise level of 0.34 µVrms in a 0.7 to 182 Hz band, consuming 1.17 µW power. The amplifier together with its power management circuitry consumes 1.56 µW, achieving a PEF of 1.41. The amplifier is also validated with adult ECG and pre-recorded fetal ECG measurements.

A 680 nA ECG acquisition IC for leadless pacemaker applications.

A sub- µW ECG acquisition IC is presented for a single-chamber leadless pacemaker applications. It integrates a low-power, wide dynamic-range ECG readout front end together with an analog QRS-complex extractor. To save ASIC power, a current-multiplexed channel buffer is introduced to drive a 7 b-to-10 b self-synchronized SAR ADC which utilizes 4 fF/unit capacitors. The ASIC consumes only 680nA and achieves CMRR > 90 dB, PSRR > 80 dB, an input-referred noise of 4.9 µVrms in a 130 Hz bandwidth, and has rail-to-rail DC offset rejection. Low-power heartbeat detections are evaluated with the help of the ASIC acquiring nearly 20,000 beats across 10 different records from the MIT-BIH arrhythmia database. In the presence of muscle noise, both the average Sensitivity (Se) and Positive Predictivity (PP) show more than 90% when the input SNR > 6 dB.