Highly Reliable 3D Channel Memory and Its Application in a Neuromorphic Sensory System for Hand Gesture Recognition.

Brain-inspired neuromorphic computing systems, based on a crossbar array of two-terminal multilevel resistive random-access memory (RRAM), have attracted attention as promising technologies for processing large amounts of unstructured data. However, the low reliability and inferior conductance tunability of RRAM, caused by uncontrollable metal filament formation in the uneven switching medium, result in lower accuracy compared to the software neural network (SW-NN). In this work, we present a highly reliable CoOx-based multilevel RRAM with an optimized crystal size and density in the switching medium, providing a three-dimensional (3D) grain boundary (GB) network. This design enhances the reliability of the RRAM by improving the cycle-to-cycle endurance and device-to-device stability of the I-V characteristics with minimal variation. Furthermore, the designed 3D GB-channel RRAM (3D GB-RRAM) exhibits excellent conductance tunability, demonstrating high symmetricity (624), low nonlinearity (ßLTP/ßLTD ∼ 0.20/0.39), and a large dynamic range (Gmax/Gmin ∼ 31.1). The cyclic stability of long-term potentiation and depression also exceeds 100 cycles (105 voltage pulses), and the relative standard deviation of Gmax/Gmin is only 2.9%. Leveraging these superior reliability and performance attributes, we propose a neuromorphic sensory system for finger motion tracking and hand gesture recognition as a potential elemental technology for the metaverse. This system consists of a stretchable double-layered photoacoustic strain sensor and a crossbar array neural network. We perform training and recognition tasks on ultrasonic patterns associated with finger motion and hand gestures, attaining a recognition accuracy of 97.9% and 97.4%, comparable to that of SW-NN (99.8% and 98.7%).

Reflection-Boosted Wearable Ring-Type Pulse Oximeters for SpO2 Measurement with High Sensitivity and Low Power Consumption.

In this study, we demonstrated a Monte Carlo simulation to model a finger structure and to calculate the intensity of photons passing through tissues, in order to determine optimal angular separation between a photodetector (PD) and a light-emitting diode (LED), to detect SpO2. Furthermore, our model was used to suggest a mirror-coated ring-type pulse oximeter to improve the sensitivity by up to 80% and improve power consumption by up to 65% compared to the mirror-uncoated structure. A ring-type pulse oximeter (RPO) is widely used to detect photoplethysmography (PPG) signals for SpO2 measurement during sleep and health-status monitoring. Device sensitivity and the power consumption of an RPO, which are key performance indicators, vary greatly with the geometrical arrangement of PD and LED within the inner surface of an RPO. We propose a reflection-boosted design of an RPO to achieve both high sensitivity and low power consumption, and determine an optimal configuration of a PD and LED by performing a 3D Monte Carlo simulation and confirming its agreement with experimental measurement. In order to confirm the reflection-boosted performance in terms of signal-to-noise ratio, R ratio, and perfusion index (PI), RPOs were fabricated with and without a highly reflective coating, and then used for SpO2 measurement from eight participants. Our simulation allows the numerical calculation of the intensity of photon passing and scattering through finger tissues. The reflection-boosted RPO enables reliable measurement with high sensitivity, resulting in less power consumption for the LED and longer device usage than conventional RPOs without any reflective coating, in order to maintain the same level of SNR and PI. Compared to the non-reflective reference RPO, the reflection-boosted RPO design greatly enhanced both detected light intensity (67% in dc and 322% in ac signals at a wavelength λ1 = 660 nm, and also 81% and 375% at λ2 = 940 nm, respectively) and PI (23.3% at λ1 and 25.5% at λ2). Thus, the reflection-boosted design not only enhanced measurement reliability but also significantly improved power consumption, i.e., by requiring only 36% and 30% power to drive the LED sources with λ1 and λ2, respectively, to produce the device performance of a non-reflective RPO reference. It is expected that our proposed RPO provides long-term monitoring capability with low power consumption and an enhanced PI for SpO2 measurement.

Laser-generated focused ultrasound transducer using a perforated photoacoustic lens for tissue characterization.

We demonstrate a laser-generated focused ultrasound (LGFU) transducer using a perforated-photoacoustic (PA) lens and a piezoelectric probe hydrophone suitable for high-frequency ultrasound tissue characterization. The perforated-PA lens employed a centrally located hydrophone to achieve a maximum directional response at 0° from the axial direction of the lens. Under pulsed laser irradiation, the lens produced LGFU pulses with a frequency bandwidth of 6-30 MHz and high-peak pressure amplitudes of up to 46.5 MPa at a 70-µm lateral focal width. Since the hydrophone capable of covering the transmitter frequency range (∼20 MHz) was integrated with the lens, this hybrid transducer differentiated tissue elasticity by generating and detecting high-frequency ultrasound signals. Backscattered (BS) waves from excised tissues (bone, skin, muscle, and fat) were measured and also confirmed by laser-flash shadowgraphy. We characterized the LGFU-BS signals in terms of mean frequency and spectral energy in the frequency domain, enabling to clearly differentiate tissue types. Tissue characterization was also performed with respect to the LGFU penetration depth (from the surface, 1-, and 2-mm depth). Despite acoustic attenuation over the penetration depth, LGFU-BS characterization shows consistent results that can differentiate the elastic properties of tissues. We expect that the proposed transducer can be utilized for other tissue types and also for non-destructive evaluation based on the elasticity of unknown materials.