FPGA-based Lightweight QDS-CNN System for sEMG Gesture and Force Level Recognition.

Deep learning (DL) has been used for electromyographic (EMG) signal recognition and achieved high accuracy for multiple classification tasks. However, implementation in resource-constrained prostheses and human-computer interaction devices remains challenging. To overcome these problems, this paper implemented a low-power system for EMG gesture and force level recognition using Zynq architecture. Firstly, a lightweight network model structure was proposed by Ultra-lightweight depth separable convolution (UL-DSC) and channel attention-global average pooling (CA-GAP) to reduce the computational complexity while maintaining accuracy. A wearable EMG acquisition device for real-time data acquisition was subsequently developed with size of 36mm×28mm×4mm. Finally, a highly parallelized dedicated hardware accelerator architecture was designed for inference computation. 18 gestures were tested, including force levels from 22 healthy subjects. The results indicate that the average accuracy rate was 94.92% for a model with 5.0k parameters and a size of 0.026MB. Specifically, the average recognition accuracy for static and force-level gestures was 98.47% and 89.92%, respectively. The proposed hardware accelerator architecture was deployed with 8-bit precision, a single-frame signal inference time of 41.9µs, a power consumption of 0.317W, and a data throughput of 78.6 GOP/s.

Wearable and flexible electrochemical sensors for sweat analysis: a review.

Flexible wearable sweat sensors allow continuous, real-time, noninvasive detection of sweat analytes, provide insight into human physiology at the molecular level, and have received significant attention for their promising applications in personalized health monitoring. Electrochemical sensors are the best choice for wearable sweat sensors due to their high performance, low cost, miniaturization, and wide applicability. Recent developments in soft microfluidics, multiplexed biosensing, energy harvesting devices, and materials have advanced the compatibility of wearable electrochemical sweat-sensing platforms. In this review, we summarize the potential of sweat for medical detection and methods for sweat stimulation and collection. This paper provides an overview of the components of wearable sweat sensors and recent developments in materials and power supply technologies and highlights some typical sensing platforms for different types of analytes. Finally, the paper ends with a discussion of the challenges and a view of the prospective development of this exciting field.

Multichannel Flexible Pulse Perception Array for Intelligent Disease Diagnosis System.

Pressure sensors with high sensitivity, a wide linear range, and a quick response time are critical for building an intelligent disease diagnosis system that directly detects and recognizes pulse signals for medical and health applications. However, conventional pressure sensors have limited sensitivity and nonideal response ranges. We proposed a multichannel flexible pulse perception array based on polyimide/multiwalled carbon nanotube-polydimethylsiloxane nanocomposite/polyimide (PI/MPN/PI) sandwich-structure pressure sensor that can be applied for remote disease diagnosis. Furthermore, we established a mechanical model at the molecular level and guided the preparation of MPN. At the structural level, we achieved high sensitivity (35.02 kPa-1) and a broad response range (0-18 kPa) based on a pyramid-like bilayer microstructure with different upper and lower surfaces. A 27-channel (3 × 9) high-density sensor array was integrated at the device level, which can extract the spatial and temporal distribution information on a pulse. Furthermore, two intelligent algorithms were developed for extracting six-dimensional pulse information and automatic pulse recognition (the recognition rate reaches 97.8%). The results indicate that intelligent disease diagnosis systems have great potential applications in wearable healthcare devices.