Cyborg and bionic systems (Washington, D.C.)最新文献

Recognizing hand gestures from neural control signals is essential for natural human-machine interaction, which is extensively applied to prosthesis control and rehabilitation. However, establishing associations between the neural control signals of motor units and gestures remains an open question. Here, we propose a channel-wise cumulative spike train (cw-CST) image-driven model (cwCST-CNN) for hand gesture recognition, leveraging the spatial activation patterns of motor unit firings to distinguish motor intentions. Specifically, the cw-CSTs of motor units were decomposed from high-density surface electromyography using a spatial spike detection algorithm and were further reconstructed into images according to their spatial recording positions. Then, the resultant cwCST-images were fed into a customized convolutional neural network to recognize gestures. Additionally, we conducted an experiment involving 10 gestures and 10 subjects and compared the proposed method with 2 root-mean-square (RMS)-based approaches and a cw-CST-based approach, namely, RMS-image-driven convolutional neural network classification model, RMS feature with linear discrimination analysis classifier, and cw-CST discharge rate feature with linear discrimination analysis classifier. The results demonstrated that cwCST-CNN outperformed the other 3 methods with a higher classification accuracy of 96.92% ± 1.77%. Moreover, analysis of cw-CST and RMS features showed that the former had better separability across gestures and consistency considering training and testing datasets. This study provides a new solution and enhances the accuracy of gesture recognition using neural drive signals in human-machine interaction.

Quadruped manipulators can use legs to mimic legged animals for crossing unstructured environments. They can also use a bionic arm to execute manipulation tasks. The increasing demands for such robots have pushed research progress. However, there remain challenging works in their usage of a high degree of freedom. To solve this redundant problem, we propose a novel motion coordination framework based on multi-task prioritization and null-space projection. The framework can adaptively generate optimal motion for different parts of the robot considering 3 prioritized tasks. The tasks include end-effector trajectory tracking, motion redistribution to meet physical constraints, and manipulability enhancement. The motion is then executed by a whole-body controller incorporating dynamics, inverse kinematics, multiobjective priorities, and force constraints. Experiments both in simulation and on the robot platform validate the advantages and effectiveness of the algorithm. The robot can finish robust and accurate operational space end-effector tracking with errors less than 3 cm.

This paper presents an earthworm-inspired multimodal pneumatic continuous soft robot enhanced by wire-winding transmission. First, a derived overlapped continuous control law based on multiple peristaltic waves is introduced to effectively improve the motion performance of the robot. Second, by applying the wire-winding transmission method, the extension of one segment is simultaneously transformed into the contraction of other segments, achieving coordinated deformation and making it more similar to real earthworms. In addition, an autonomous obstacle-avoidance control strategy based on contact force sensing is developed to enhance the environmental adaptability of the robot. Based on these methods, an earthworm-inspired soft robot that can perform multimodal movements with autonomous obstacle-avoidance ability and enhanced motion efficiency is developed. A series of experiments including in- and cross-plane crawling, obstacle avoidance steering, and pipeline crawling are conducted to validate the robot's multimodal motion capabilities. The robot can achieve a speed of 6.65 mm/s (36.0 × 10^-3 bl/s) during in-plane crawling movement and 1.66 mm/s (8.97 × 10^-3 bl/s) during pipeline crawling movement. In terms of the in-plane crawling speed, the robot surpasses other robots of the same type. In conclusion, the robot's multimodal capabilities and enhanced motion efficiency demonstrate superior overall performance, and the robot has good potential for medical and industrial applications.

We previously developed a novel needle-free reagent injection method based on electrically induced microbubbles. The system generates microbubbles and applies repetitive mechanical oscillation associated with microbubble dynamics to perforate tissue and introduce a reagent. In this paper, we propose improving the reagent injection depth by reflecting the shock wave through microbubble dynamics. Our results show that the developed shock wave reflection method improves the ability of the electrically induced microbubble injection system to introduce a reagent. The method extends the application potential of electrically induced microbubble needle-free injection.

Recent advancements in robotics and sensor technology have facilitated the development of myoelectric prosthetic hands (MPHs) featuring multiple degrees of freedom and heightened functionality, but their practical application has been limited. In response to this situation, formulating a control theory ensuring the hand dexterity of highly functional MPHs has garnered marked attention. Progress in this field has been directed toward employing machine-learning algorithms to process electromyogram patterns, enabling a broad spectrum of hand movements. In particular, the practical application of 5-finger-driven MPHs with such control functions to real users remains limited, and their attributes and challenges have not been thoroughly examined. In this study, we developed a 5-finger MPH equipped with pattern recognition capabilities. Through a long-term clinical trial, encompassing task assessments and subjective evaluations via questionnaires, we explored the MPH's range of applications. The task assessments revealed an expanded range of achievable tasks as the variety of motions increased. However, this enhanced adaptability was paralleled by a decrease in control reliability. Additionally, findings from the questionnaires indicated that enhancements in task performance with MPHs might be more effective in reducing workplace-related disability than in improving activities in everyday life. This study offers valuable insights into the long-term clinical prospects and constraints associated with multi-degree-of-freedom MPHs incorporating pattern recognition functionality.

Magnetic miniature robots have shown great potential in biomedical applications in recent years. However, a challenge remains in which it is difficult for magnetic miniature robots to achieve balanced capabilities for multimodal locomotion and fluidic manipulation in various environments. Here, we report a magnetic shaftless propeller-like millirobot (MSPM) that possesses the capabilities of rotating-based multimodal 3-dimensional motion and cargo transportation with untethered manipulation. The MSPM utilizes the propulsion and pumping capabilities of the propeller structure to achieve fluidic manipulation. The shaftless propeller structures are designed to achieve omnidirectional locomotion through rolling, propelling, and tumbling. Additionally, the shaftless 3-blade propeller is used to perform a pumping function to achieve controllable transportation of fluids and particles. We anticipate that the MSPM holds great potential as a minimally invasive device for thrombosis treatment and targeted medicine delivery.

Birds achieve remarkable flight performance by flexibly morphing their wings during different flight stages. However, due to the lack of experimental data on the free morphing of wings and the complexity of coupled motion in aerodynamics studies, the intricate kinematic changes and aerodynamic mechanisms of wings during various flight stages still need to be explored. To address this issue, we collected comprehensive data on free-flight pigeons (Columba livia). We categorized the wing kinematic parameters during the takeoff, leveling flight, and landing stages into 5 kinematics parameters: flap, twist, sweep, fold, and bend. Based on this, we established a 3-dimensional pigeon wing model, defined its coupled motion using rotation matrices, and then used the computational fluid dynamics method to simulate the coupled motion in the 3 flight stages. We analyzed and compared the kinematic parameter changes, aerodynamic forces, and flow structures. It is found that, within a wingbeat cycle, pigeons during the takeoff stage cause the leading-edge vortex to attach earlier, enhancing instantaneous lift to overcome gravity and achieve ascending. During the leveling flight stage, the pigeon's average lift becomes stable, ensuring a steady flight posture. In the landing stage, the pigeon increases the wing area facing the airflow to maintain a stable landing posture, achieving a more minor, consistent average lift while increasing drag. This study enhances our understanding of birds' flight mechanisms and provides theoretical guidance for developing efficient bio-inspired flapping-wing aerial vehicles.

Type 2 diabetes is considered as a chronic inflammatory disease in which the dense microvasculature reorganizes with disease progression and is highly correlated with β cell mass and islet function. In this study, we constructed rat models of type 2 diabetes and used ultrasound localization microscopy (ULM) imaging to noninvasively map the pancreatic microvasculature at microscopy resolution in vivo to reflect β cell loss and islet function deterioration, and evaluate the efficacy after anti-cytokine immunotherapy. It was unveiled that ULM morphological and hemodynamic parameters have a strong link with β cell loss and deterioration of pancreatic islet function. This correlation aligns with the observed pathological alterations in the microvessels of islet and demonstrated that ULM can effectively mirror the functionality of β cells during rapid fluctuations in blood glucose levels by observing changes in mean velocity. Furthermore, it was revealed that treatment with anti-cytokine immunotherapy enhances the function and health of β cells by restoring the microvascular environment. Remarkable improvements in vessel morphology (measured by fractal dimension) and hemodynamics (indicated by mean velocity and vessel density) were noted following the anti-cytokine immunotherapy, signifying a significant enhancement at the treatment's conclusion (P < 0.05). These observations suggested that ULM technology holds promise as a visible and efficient tool for monitoring the effectiveness of anti-cytokine immunotherapy in managing type 2 diabetes. Pancreatic microvessel-based ULM may serve as a novel noninvasive method to assess β cells, providing a valuable clinical tool for tracking the progression of type 2 diabetes.

Digital microfluidic chips (DMCs) have shown huge potential for biochemical analysis applications due to their excellent droplet manipulation capabilities. The driving force is a critical factor for characterizing and optimizing the performance of droplet manipulation. Conducting numerical analysis of the driving force is essential for DMC design, as it helps optimize the structural parameters. Despite advances in numerical analysis, evaluating driving forces in partially filled electrodes remains challenging. Here, we propose a versatile electrodynamics simulation model designed to analyze the driving forces of partially filled electrodes to optimize the structural parameters of DMCs. This model utilizes finite element analysis to determine the voltage distribution within the DMC and calculates the driving force acting on the droplets using the principles of virtual work. Using this electrodynamics simulation model, we evaluated the effects of various structural parameters, including the dielectric constant and thickness of the dielectric layer, the dielectric constant and conductivity of the droplet, and substrate spacing, on the droplet driving force. This evaluation helps to optimize the structural parameters and enhances the droplet manipulation of DMCs. Measurements of droplet acceleration demonstrated that the droplet acceleration on the partially filled electrode aligns with the simulated driving force trend, which verified the effectiveness of the proposed electrodynamics simulation model. We anticipate that the electrodynamics simulation model is capable of evaluating the driving force in partially filled electrodes within complex DMCs, offering unprecedented possibilities for future structural designs of DMCs.

The presence of cellular defects of multifactorial nature can be hard to characterize accurately and early due to the complex interplay of genetic, environmental, and lifestyle factors. With this study, by bridging optically-induced dielectrophoresis (ODEP), microfluidics, live-cell imaging, and machine learning, we provide the ground for devising a robotic micromanipulation and analysis system for single-cell phenotyping. Cells under the influence of nonuniform electric fields generated via ODEP can be recorded and measured. The induced responses obtained under time-variant ODEP stimulation reflect the cells' chemical, morphological, and structural characteristics in an automated, flexible, and label-free manner. By complementing the electrokinetic fingerprint of the cell centroid motion with data on the dynamics of electro-deformation and orientation, we show that subtle differences at the single-cell level can be elucidated. Specifically, here, we demonstrate, for the first time, the ability of the combined ODEP-based robotic and automatic analysis platform to discriminate between primary endometrial stromal cells obtained from fertile patients and patients with disrupted receptivity/selectivity equilibrium. When multiple cells were considered at the patient level, the performance achieved an average accuracy of 98%. Single-cell micro-operation and analysis systems may find a more general application in the clinical diagnosis and management of patients with pathological alterations at the cellular level.