Advanced intelligent systems (Weinheim an der Bergstrasse, Germany)最新文献

Soft robots with multimode locomotion possess great application potential across various engineering fields due to their exceptional motion flexibility and environmental adaptability. Conventional approaches achieve multiple locomotion modes by designing a primary structure capable of different movements and then employing a series of actuators to drive each motion. Alternatively, soft robots made of stimuli-responsive materials usually take the shape of a thin sheet and generate different motions by modulating the external stimuli. This study presents a multimode soft robot based on a single braided tube composed of shape memory alloy wires set at distinct initial configurations. By strategically actuating different wires, the braided tube realizes axial contraction, elongation, and bending. A theoretical model is developed to analyze the underlying deformation mechanisms and to establish a quantitative relationship between the design parameters and the deformation, which is validated by experiments. Building on this, three types of braided soft robots, a crawling robot, a rolling robot, and a multimode robot capable of straight crawling, left/right turning, inchworm crawling, and rolling, are designed and actuated without additional actuators. The proposed structure–actuation integrated design approach provides a new way of developing highly integrated, multifunctional soft robots with enhanced adaptability and performance.

Accurate single-cell analysis is critical for diagnostics, immunomonitoring, and cell therapy, but coincident events, where multiple cells overlap in a sensing zone, can severely compromise signal fidelity. A hybrid framework combining a fully convolutional neural network (FCN) with compressive sensing (CS) to disentangle such overlapping events in 1D sensor data is presented. The FCN, trained on bead-derived datasets, accurately estimates coincident event counts and generalizes to immunomagnetically labeled CD4⁺ and CD14⁺ cells in whole blood without retraining. Using this count, the CS module reconstructs individual signal components with high fidelity, enabling precise recovery of single-cell features, including velocity, amplitude, and hydrodynamic diameter. Benchmarking against conventional state-machine algorithms shows superior performance, recovering up to 21% more events and improving classification accuracy beyond 97%. Explainability via class activation maps and parameterized Gaussian template fitting ensures transparency and clinical interpretability. Demonstrated with magnetic flow cytometry (MFC), the framework is compatible with other waveform-generating modalities, including impedance cytometry, nanopore, and resistive pulse sensing. This work lays the foundation for next-generation nonoptical single-cell sensing platforms that are automated, generalizable, and capable of resolving overlapping events, broadening the utility of cytometry in translational medicine and precision diagnostics, e.g., cell-interaction studies.

According to the World Health Organization, cardiovascular diseases are the leading cause of death worldwide. The electrocardiogram (ECG) is a widely used noninvasive method for detecting these conditions. However, analyzing long-duration ECG signal recordings can be highly time-consuming for medical professionals. Machine learning and deep learning techniques have emerged as valuable tools to assist in diagnosis. However, class imbalance in medical datasets poses a significant challenge. This work presents a comparative analysis of three generative adversarial network (GAN)-based models—deep convolutional GAN, conditional GAN, and Wasserstein GAN with gradient penalty (WGAN-GP)—to generate synthetic ECG spectrograms. The proposed models are evaluated using the Fréchet inception distance and structural similarity index measure. The results indicate that WGAN-GP models outperform the other two models in terms of intraclass diversity and data quality. These findings suggest that GAN-generated spectrograms can help mitigate data imbalance issues and improve ECG classification models.

Memristor-based stateful logic offers a promising solution for in-memory computing by mitigating the von Neumann bottleneck and minimizing data movement between memory and processing units. At the heart of this approach, primitive logic circuits, primarily constructed from resistive switching memory (memristor) units, serve as the foundational elements of stateful logic families. However, the stochastic switching behavior of memristors can compromise computational accuracy, necessitating optimization strategies to ensure reliable and robust logic operations. In this work, we investigate the switching voltage distributions of memristors with an Al₂O₃/TiO_x/TiO_y structure and utilize their multilevel state tunability to propose a novel stateful logic architecture along with an optimization method to enhance operational reliability across various logic types. The proposed optimization strategy is experimentally validated, demonstrating high logic fidelity under all input conditions. Furthermore, a 1-bit full adder, a fundamental arithmetic logic unit, is implemented by cascading the developed stateful logic gates. Finally, this study presents a parallel operation method for stateful logic in a crossbar array, enabling n-bit full adder implementation with a reduced number of computational steps by maximizing parallelism.

Colorectal liver metastases (CRLM) are a significant challenge in oncology, as recurrence after liver resection is frequently observed. Accurate prediction of CRLM recurrence is important to guide specific treatment strategies and improve clinical outcomes. To address this issue, this study proposes a novel framework. To the best of current knowledge, this is the first approach to integrate graph neural networks (GNNs) and causal inference to predict postresection CRLM recurrence using clinical and pathological characteristics. In addition, a GNNExplainer framework is also utilized for the interpretability of the models beyond predictive accuracy. The proposed framework identifies the factors of recurrence and their impact on patient outcomes, not only providing predictions to clinicians but also explaining the underlying reasons. Furthermore, causal inference strengthens the model by confirming factors. The relevance of these variables is also shown through counterfactual and interventional analyses, allowing for more evidence-based choices. The GCN model of theframework exhibits high performance with a test accuracy of 99.40%, an aF1-score of 99.21%, and a receiver operating characteristic area under the curve (ROC AUC) of 99.97%. An extensive evaluation shows the clinical applicability of the proposed framework.

Soft robotics holds immense promise for applications requiring adaptability and compliant interactions. However, the lack of sufficiently fast and accurate simulation environments for soft robots has hindered progress, particularly in linking with reinforcement learning (RL) applications. Traditional finite element method (FEM) models provide precise insights into soft robot dynamics but are computationally intensive and impractical for accelerated simulation. This work introduces a novel framework that integrates high-fidelity FEM simulations with computationally efficient physics-based simulations through a surrogate model tailored for RL. The surrogate model, trained on real-world and FEM-generated datasets, captures complex dynamics while maintaining efficiency. Sim2real experiments validate the framework, implementing the trajectory tracking and the force control tasks with high accuracy. These results demonstrate the framework's ability to bridge the simulation gap, enabling its application to advanced tasks, such as manipulation and interaction in unstructured environments.

As the most common pediatric malignancy, B-cell acute lymphoblastic leukemia (B-ALL) has multiple distinct subtypes characterized by recurrent and sporadic somatic and germline genetic alterations. Identifying B-ALL subtypes can facilitate risk stratification and enable tailored therapeutic design. Existing methods for B-ALL subtyping primarily depend on immunophenotyping, cytogenetic tests, and genomic profiling, which can be costly, complicated, and laborious. To overcome these challenges, RanBALL (an ensemble random projection-based model for identifying B-ALL subtypes) is presented, an accurate and cost-effective model for B-ALL subtype identification. By leveraging random projection (RP) and ensemble learning, RanBALL can preserve patient-to-patient distances after dimension reduction and yield robustly accurate classification performance for B-ALL subtyping. Benchmarking results based on >1700 B-ALL patients demonstrate that RanBALL achieves remarkable performance (accuracy: 0.93, F1-score: 0.93, and Matthews correlation coefficient: 0.93), significantly outperforming state-of-the-art methods like ALLSorts in terms of all performance metrics. In addition, RanBALL performs better than t-SNE in terms of visualizing B-ALL subtype information. We believe RanBALL will facilitate the discovery of B-ALL subtype-specific marker genes and therapeutic targets to have consequential positive impacts on downstream risk stratification and tailored treatment design is believed. To extend its applicability and impacts, a Python-based RanBALL package is available at https://github.com/wan-mlab/RanBALL.

Soft wearable robots have gained widespread interest across various disciplines; however, they remain insufficient in overcoming the physical limitations of the human body. In particular, enhancing vertical jump height, a commonly used indicator of physical capability, requires improved actuator power density, stroke length, and soft structure efficiency. To address these challenges, the Jump-Enhancing Textile Suit is proposed, which integrates the Pneumatic Energy-Storing Propulsion Actuator (PESPA) and the Triarticular Kinetic-Chained Structure (TKiCS) to assist jump performance. PESPA stores elastic energy under pneumatic pressure and releases it during the propulsive phase to augment human movement. TKiCS uses the kinetic chain mechanism to reduce anchoring points and fully harness the high stiffness region, thereby improving force transmission efficiency. Controlled vertical jump experiments with healthy adult participants are conducted. The suit increases jump height by 3.74 cm on average and up to 9.04 cm maximum, while also enhancing hip, knee, and ankle torques. Under isotonic testing, PESPA achieves a power density of 2298.69 W kg⁻¹ and outperforms conventional pneumatic actuators. A dynamic model enables accurate force prediction and precise timing for effective assistance. These findings establish a practical foundation for pneumatic wearable robotics and suggest applications in jump augmentation, rehabilitation, and athletic performance.

Soft robotics requires compliant actuators for safe human interaction. While McKibben artificial muscles are popular for their high force output, their reliance on bulky, noisy pumps limits their use in wearable devices. Electrohydrodynamic (EHD) pumps offer a compact and silent alternative, but existing designs struggle with dielectric discharge and fabrication issues, which compromise reliability and power density. This study introduces a novel EHD pump featuring 0.1 mm copper wire electrodes in a diagonal arrangement within a laser-cut acrylic frame. This design improves dielectric resilience, minimizes deformation, and allows for compact integration. A new simplified fabrication process results in sample variation under 5%. The pump demonstrates remarkable performance, achieving 107 kPa pressure and an 88 mL min⁻¹ flowrate, doubling the power density of the previous model while retaining 88% of its flowrate after 50 discharge events. An automated self-recovery mechanism is also implemented, enabling the pump to instantly restore function after a discharge. When paired with a McKibben muscle, the system achieves a 2 s contraction time, a tenfold improvement over the prior EHD-driven system. This work presents a significant advancement in fast, resilient, and scalable actuation, paving the way for next-generation wearable robotics and assistive technologies.

Macrophages play a central role in modulating different biological and physiological events. The behaviors and functions of macrophages may be regulated by a host of factors, including their viability, proliferation rate, and population density. Specifically, the population density of macrophages has been increasingly reported to be correlated with their activities. It is, however, still unclear if changes in macrophage population density will alter the biophysical attributes of these cells, notably their morphology. Herein, label-free phase-contrast microscopy is coupled with machine learning to interrogate the relationship between the population density and morphological features of macrophages. Through a systematic approach, variations in the morphological phenotypes of macrophages, which are dependent on their population density, are revealed. In parallel, through unsupervised clustering, the presence of single-cell morphological heterogeneity within each macrophage population and subpopulation is elucidated. Next, discriminative morphological attributes which can be leveraged to distinguish between macrophages from different groups are identified through feature scoring. Finally, high-performing explainable supervised machine learning algorithms that can be employed to predict the population density of macrophages based on their size and shape features are identified. This work is anticipated to offer a deeper understanding of the association between macrophage population density and morphologyas well as the potential use of morphological attributes as predictive metrics for analyzing cell populations.