National Microelectronics Institute

Clinical-ready neuromodulation devices for real-time monitoring and treatment of neurological disorders require a neural interface integrated circuit (IC) that supports therapeutic neurostimulation together with concurrent recording of neural activities from large brain regions. To address this need, this article presents an IC with 32 high dynamic range (HDR) electrocorticography (ECoG) recording channels and two programmable neurostimulators allowing bi-directional neural interfacing. The recording channel uses a 2nd-order continuous-time <inline-formula xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink"> <tex-math notation="LaTeX">$\Delta \Sigma $</tex-math> </inline-formula> modulator (CTDSM) architecture with 12-tap finite impulse response digital-to-analog converter (FIRDAC) feedback facilitated by a pre-emptive current-steering excess loop delay (ELD) compensation scheme. By employing a high-performance linearized low-noise/low-power transconductance amplifier (L<sup xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">3</sup>TA) in the input integration stage, the recording channel achieves 83-dB dynamic range (DR) with up to 222-mV<sub xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">pp</sub> linear input range, and thus can accommodate large electrode dc offsets (EDOs) and stimulation artifacts. Thanks to the FIRDAC that relaxes component matching requirements, each recording channel occupies only 0.028 mm<sup xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">2</sup> and consumes <inline-formula xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink"> <tex-math notation="LaTeX">$5.39~{\mu }$</tex-math> </inline-formula>W while being manufactured in standard 180-nm CMOS technology. The concurrent neural recording and stimulation capabilities of the chip have been validated in a mouse experiment in vivo, which shows that the recording channel can tolerate real stimulation artifacts and the recorded ECoG signals can be separated from the artifacts with high fidelity.

This paper presents a high-resolution, single-photon LiDAR imaging camera capable of performing simultaneous depth and color estimation. It features a time-gated CMOS single-photon avalanche diode detector array with a large fill-factor, integrated with a metasurface-based plasmonic color filter array for three-dimensional color imaging. A sliding detector gate enables millimeter-scale depth resolution using as few as two binary planes per step, and high-fidelity color reconstruction is demonstrated from as few as two photons per pixel. This compact, photon-efficient system offers depth-sensitive, multispectral imaging possibilities for many diverse applications such as space exploration, autonomous navigation, augmented/virtual reality, and biomedical imaging.

Taking inspiration from human tactile system, a sensitive biomimetic multimodal tactile sensor for discrimination of static and dynamic forces is presented in this article. The multimodal tactile sensor has a piezoelectric-capacitive tandem for responding to the dynamic and static forces, respectively. Sensor can cater to normal direction dynamic force signals with a piezoelectric part operating in the <inline-formula xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink"> <tex-math notation="LaTeX">${d}_{{33}}$ </tex-math></inline-formula> mode and static force with a capacitive part. The capacitive sensing part has a unique configuration with a top electrode and two sets of differential pairs electrodes for the force measurement in x and y shear axis and one electrode for normal force measurement. The experimental characterization of the sensor was performed for static, quasi-static, and dynamic forces. Along with the static forces, the sensor was also able to cater to dynamic forces up to 60 Hz. The force sensitivity of the sensor for the normal force is 0.084 pF/N and 0.035 V/N from the capacitive and piezoelectric part, respectively, for a force range of 10 N. Also, in the shear X- and Y-directions, the sensor exhibited a sensitivity of 0.027 and 0.029 pF/N, respectively, in the force range of 1.2 N. Through the vibrotactile data, the sensor showed an ability to discriminate between two texture samples through a neural network classifier. The presented sensor owing to its dimension, performance, and capabilities can find its application in minimally invasive robotic surgery, robotics, wearable devices, and prosthetics.

Electrical impedance tomography (EIT)-based tactile sensors offer cost-effective and scalable solutions for robotic sensing, especially promising for soft robots. However, a major issue of EIT-based tactile sensors when applied to highly deformable objects is their performance degradation due to surface deformations. This limitation stems from their inherent sensitivity to strain, which is particularly exacerbated in soft bodies, thus requiring dedicated data interpretation to disentangle the parameter being measured and the signal deriving from shape changes. This has largely limited their practical implementations. This article presents a machine learning-assisted tactile sensing approach to address this challenge by tracking surface deformations and segregating this contribution in the signal readout during tactile sensing. We first capture the deformations of the target object, followed by tactile reconstruction using a deep learning model specifically designed to process and fuse EIT data and deformation information. Validations using numerical simulations achieved high correlation coefficients (0.9660–0.9999), peak signal-to-noise ratios (SNRs) (28.7221–55.5264 dB), and low relative image errors (RIEs) (0.0107–0.0805). Experimental validations, using a hydrogel-based EIT e-skin under various deformation scenarios, further demonstrated the effectiveness of the proposed approach in real-world settings. The findings could underpin enhanced tactile interaction in soft and highly deformable robotic applications.

We propose a fully asynchronous peak detection approach for SPAD-based direct Time-of-Flight (dToF) flash LiDAR, enabling pixel-wise event-driven depth acquisition without global synchronization. By allowing pixels to independently report depth once a sufficient signal-to-noise ratio is achieved, the method reduces latency, mitigates motion blur, and increases effective data rate compared to frame-based systems. The framework is validated under two hardware implementations: an offline <inline-formula xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink"> <tex-math notation="LaTeX">$256\times 128$</tex-math> </inline-formula> SPAD array with PC based processing and a real-time FPGA proof-of-concept prototype with <inline-formula xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink"> <tex-math notation="LaTeX">$2.4\mu $</tex-math> </inline-formula>s latency for on-chip integration. Experiments demonstrate robust depth estimation, reflectivity reconstruction, and dynamic event-based representation under both static and dynamic conditions. The results confirm that asynchronous operation reduces redundant background data and computational load, while remaining tunable via simple hyperparameters. These findings establish a foundation for compact, low-latency, event-driven LiDAR architectures suited to robotics, autonomous driving, and consumer applications. In addition, we have derived a semi-closed-form solution for the detection probability of the raw-peak finding based LiDAR systems that could benefit both conventional frame-based and proposed asynchronous LiDAR systems.

This work presents L-Sort, the first localization-based spike sorting accelerator on a miniaturized chip hardware (FPGA). Spike sorting, crucial for neural activity analysis, involves detecting, extracting, and clustering spikes. L-Sort uses geometric data to localize spikes, reducing the need for extensive waveform retention. The system includes a ZCU104 FPGA board and a computer for real-time visualization. Visitors will feed raw Neuropixel data into the FPGA and observe real-time sorted spike trains.

not-yet-known not-yet-known not-yet-known unknown A Memristor-controlled CMOS reconfigurable true time delay circuit is introduced in this paper. The delay circuit includes 3 stages of g m -C all pass filter delay elements connected in cascade, and me

Dielectrophoresis (DEP) is a fast and reliable nanoparticle recovery method that utilizes nonuniform electric fields to manipulate particles based on their material composition and size, enabling recovery of biologically-derived nanoparticles from plasma for diagnostic applications. When applying DEP to undiluted human plasma, collection of endogenous albumin proteins was observed at electric field gradients much lower than predicted by theory to collect molecular proteins. To understand this collection, nanoparticle tracking analysis of bovine serum albumin (BSA) dissolved in 0.5× phosphate-buffered saline was performed and showed that albumin spontaneously formed aggregate nanoparticles with a mean diameter of 237nm. These aggregates experienceda dielectrophoretic force as a function of aggregate radius rather than the diameter of individual protein molecules which contributed to their collection. In high conductance buffer (6.8mS/cm), DEP was able to move these aggregates into regions of high electric field gradient, and in lower conductance buffer (0.68mS/cm), these aggregates could be moved into high or low gradient regions depending on the applied frequency. Disruption of BSA aggregates using a nonionic detergent significantly decreased the particle diameter, resulting in decreased dielectrophoretic collection of albumin which increased the collection consistency of particles of interest. These results provide techniques to manipulate albumin aggregates via DEP, which impacts collection of diagnostic biomarkers.