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.
A biological muscle's force is nonlinearly constrained by its current state (force, length, and speed) and state history. To investigate if artificial muscles can mimic the complete mechanical state spectrum of biological muscles, this study uses a novel method to characterize twisted coiled polymer actuators (TCPAs) mechanically. Thus, comprehensive and reproducible test procedures are established to verify artificial muscle biomimetics regarding stress, strain, and strain rate combinations intrinsic to biological muscle. A rheometer performs novel high-precision mechanical characterization methods that comprehensively verify biomimetic performance. Sample twist level, torque, length, force, and temperature were controlled and measured during twist-induced coiling, heatsetting/annealing, and mechanical testing. TCPAs were formed from linear low-density polyethylene monofilament. LLDPE TCPAs generate larger stresses than biological muscle through the entire spectrum of strains - contracting more than 40%, exerting more than 0.3 MPa at rest length, and withstanding tension of 8 MPa without damage. Thus, the LLDPE TCPAs attained biological muscle performance statically, but additional tests are required to assess this dynamically. The mechanical performance of LLDPE TCPAs enables biomimetic actuation with an intelligent control and measurement system. Their high-throughput textile manufacturability positions them for advanced biomechatronic applications - including prosthetics and exoskeletons.
Insulin pumps typically use piston-based mechanisms with bulky transmission components to convert rotary motion into the piston's forward motion. These mechanical transmission systems and insulin reservoirs occupy more than one-third of pumps' volume, significantly limiting miniaturization and making pumps cumbersome for daily use. Herein, a compact, magnetically actuated insulin pump is developed that is less than one-quarter the size of piston-based pumps. Instead of bulky mechanical components, the pump uses a magnetic soft actuator to directly compress the insulin chamber, controlled by a precisely tuned electromagnetic field. This innovative design eliminates the need for large transmission systems, enabling a notably smaller form factor. In addition, the fine-tunable magnetic actuation enables a 0.01 μL delivery resolution, significantly surpassing the 0.25 μL resolution of piston-based pumps. This high-resolution mechanism facilitates further miniaturization by allowing the use of high-concentration insulins, thereby reducing the reservoir size. By varying the magnetic field's waveform, amplitude, and duration, the pump's performance can be further enhanced. The reported magnetic insulin pump exhibits superior repeatability and accuracy across single-pulse, basal, and bolus modes compared to commercial insulin pumps. This miniaturized, high-resolution magnetic insulin pump is anticipated to substantially benefit people with diabetes by improving portability, precision, and cost efficiency.
The local force field generated by light endows optical microrobots with remarkable flexibility and adaptivity, promising significant advancements in precise medicine and cell transport. Nevertheless, the automated navigation of multiple optical microrobots in intricate, dynamic environments over extended distances remains a challenge. In this study, we introduce a versatile control strategy aimed at navigating optical microrobotic swarms to distant targets under obstacles of varying sizes, shapes, and velocities. By confining all microrobots within a manipulation domain, we ensure swarm integrity while mitigating the effects of Brownian motion. Obstacle's elliptical approximation is developed to facilitate efficient obstacle avoidance for microrobotic swarms. Additionally, we integrate several supplementary functions to enhance swarm robustness and intelligence, addressing uncertainties such as swarm collapse, particle immobilization, and anomalous laser-obstacle interactions in real microscopic environments. We further demonstrate the efficacy and versatility of our proposed strategy by achieving autonomous long-distance navigation to a series of targets. This strategy is compatible with both optical trapping- and nudging-based microrobotic swarms, representing a significant advancement in enabling optical microrobots to undertake complex tasks such as drug delivery and nanosurgery and understanding collective motions.
Mucus accumulation is a major complication in cystic fibrosis (CF) and chronic obstructive pulmonary disease (COPD), increasing risks of airway obstruction and impairing drug delivery. Maintaining the patency of airway prosthetic devices such as stents and endotracheal tubes also requires effective clearance. Current approaches, primarily blind suction, are ineffective against viscous mucus and risk tissue trauma. Here, it proposes a mucus-cleaning catheter that integrates a rotational soft ciliary brush with suction for efficient and safe clearance on tracheobronchial tissues and within airway prosthetic devices. The soft artificial cilia efficiently wrap up mucus around the suction inlets for removing viscous mucus while a haptic sensing unit provides real-time feedback for operational safety. The device can remove mucus in thin layers without damaging the tissue, typically difficult for pure suction by a therapeutic bronchoscope. This study demonstrates that the device can access airway phantom models, ex vivo ovine lungs, airway stents, and endotracheal tubes, performing on-demand mucus clearance under real-time visualization via an onboard camera. Compared with therapeutic bronchoscopes with pure suction, the catheter provides faster clearance and minimal damage to tissues, offering a promising strategy for safe, efficient, and timely mucus cleaning for patients with excessive mucus accumulation in the airway.
Remote manipulation devices extend human capabilities over vast distances or in inaccessible environments, removing constraints between patients and treatment. The integration of therapeutic and assistive devices with the Internet of Things (IoT) has demonstrated high potential to develop and enhance intelligent rehabilitation systems in the e-health domain. Within such devices, soft robotic products distinguish themselves through their lightweight and adaptable characteristics, facilitating secure collaboration between humans and robots. The objective of this research is to combine a textile-based sensorized glove with an air-driven soft robotic glove, operated wirelessly using the developed control system architecture. The sensing glove equipped with capacitive sensors on each finger captures the movements of the medical staff's hand. Meanwhile, the pneumatic rehabilitation glove designed to aid patients affected by impaired hand function due to stroke, brain injury, or spinal cord injury replicates the movements of the medical personnel. The proposed artificial intelligence-based system detects finger gestures and actuates the pneumatic system, responding within an average response time of 48.4 ms. The evaluation of the system further in terms of accuracy and transmission quality metrics verifies the feasibility of the proposed system integrating textile gloves into IoT infrastructure, enabling remote motion sensing and actuation.
Non-small cell lung cancer (NSCLC) comprises the largest subtype of lung cancer with the most cases. Lung adenocarcinoma and lung squamous cell carcinoma are two NSCLC subtypes that pose challenges for accurate diagnosis using conventional methods, including histological examination and imaging, which can be slow and inconclusive. To address these concerns, RPSLearner is proposed, which combines random projection (RP) for dimensionality reduction and stacking ensemble learning to accurately predict lung cancer subtypes. Specifically, multiple independent RP matrices are first generated to project the high-dimensional RNA-seq data into a lower-dimensional space, whose features are subsequently concatenated. After that, the concatenated RP features are fed into a stack of diverse base classifiers, and integrated the predictions from base models via a deep linear layer network. Benchmarking tests on 1 333 NSCLC patients demonstrated that RPSLearner outperformed state-of-the-art approaches for lung cancer subtype classification. Specifically, RPSLearner efficiently preserved sample-to-sample distances even after significant dimension reduction, and the meta-model in RPSLearner yielded consistently higher scores than individual base models. In addition, the feature fusion method outperformed conventional score ensemble methods. We believe RPSLearner is a promising model for downstream lung cancer clinical diagnosis, and it holds the potential to be extended to subtyping of other types of cancer.
Recent advances in dynamic and reversible polymer networks have led to self-healing soft robots that can restore their physical and electrical properties after damage. However, in most cases, human intervention remains essential for the healing process. This poses a challenge, especially in working environments with limited human access or where human involvement cand hinder efficiency. To address this gap, in this article, first, the different phases of the healing process in soft robotics are discussed and then the technologies that are or can be integrated into self-healing soft robots to allow each individual phase to be performed autonomously with minimal human involvement are reviewed. Finally, in this article, the challenges of integrating all phases into self-healing soft robots are discussed and the perspectives on achieving fully autonomous self-healing in the future are offered. These phases are classified into five: damage detection, damage cleaning, damage closure, stimulus-triggered material healing, and recovery assessment. Achieving these attributes requires employing physical intelligence at the material level through the use of stimuli-responsive materials or utilizing embodied intelligence at the system level by integrating healing-assistive subsystems or a synergistic combination of both. Consequently, self-healing soft robots can achieve self-sufficiency in their healing capabilities, rendering them a sustainable solution for broader applications.
SuperResNET is an integrated machine learning-based analysis software for visualizing and quantifying 3D point cloud data acquired by single-molecule localization microscopy (SMLM). SuperResNET computational modules include correction for multiple blinking of single fluorophores, denoising, segmentation (clustering), feature extraction used for cluster group identification, modularity analysis, blob retrieval, and visualization in 2D and 3D. Here, a graphical user interface version of SuperResNET was applied to publicly available direct stochastic optical reconstruction microscopy (dSTORM) data of nucleoporin Nup96 and Nup107 labeled nuclear pores that present a highly organized octagon structure of eight corners. SuperResNET effectively segments nuclear pores and Nup96 corners based on differential proximity threshold analysis from 2D and 3D SMLM datasets. SuperResNET quantitatively analyzes features from segmented nuclear pores, including complete structures with eightfold symmetry, and from segmented corners. SuperResNET modularity analysis of segmented corners from 2D SMLM distinguishes two modules at 10.7 ± 0.1 nm distance, corresponding to two individual Nup96 molecules. SuperResNET is therefore a model-free tool that can reconstruct network architecture and molecular distribution of subcellular structures without the bias of a specified prior model, attaining molecular resolution from dSTORM data. SuperResNET provides flexibility to report on structural diversity in situ within the cell, providing opportunities for biological discovery.
Developing microrobotic systems for accurate and fast manipulation of microobjects or living cells has the potential to significantly advance biomedical and microfabrication applications. Despite recent progress in this field, comprehensive multistimuli responsive, fast, and precisely controllable microrobots remain limited. In this study, automated position and speed control of acoustically powered, bubble-based, magnetically steerable microrobots is demonstrated, along with micromanipulation of mammalian cells using these microswimmers. Enhanced control of the microswimmers is achieved by designing and implementing a closed-loop control system that guides the microrobots along a predetermined path while modulating their speed by adjusting the acoustic frequency near the resonant value. The microrobots are guided to cells, enabling cell manipulation by pulling them with the microrobots. Overall, the results highlight the capability and controllability of these magnetically and acoustically responsive microrobots for future cell-based applications, including manipulation, delivery, and microsurgery.
Atomic force microscopy (AFM) is extensively applied to measure the nanomechanical properties of living cells. Despite its popularity, some applications on mechanobiology are limited by the low throughput of the technique. Currently, the analysis of AFM-nanoindentation data is performed by model fitting. Model fitting is slow, data intensive, and prone to error. Herein, a supervised machine-learning regressor is developed for transforming AFM force-distance curves into nanorheological behavior. The method reduces the computational time required to process a force volume of a cell made of 2.62 × 105 curves from several hours to minutes. In fact, the regressor increases the throughput by 50-fold. The training and the validation of the regressor are performed by using theoretical curves derived from a contact mechanics model that combined power-law rheology with bottom effect corrections and functional data analysis. The regressor predicts the modulus and the fluidity coefficient of mammalian cells with a relative error below 4%.
Sweat rate measures key physiological states such as hydration levels and heat tolerance. Incorporating wearable technology with sweat rate sensors allows individuals to conveniently monitor their health, optimize workouts, and enhance occupational safety. However, challenges persist in such integration techniques, including intricate manufacturing, non-linear responses to changes in sweat rates, and errors from the intermediate measurement of the distance sweat travels in the sensor. To address these issues, we present a comprehensive wearable platform that includes a fully printed, flexible sensor patch, readout electronics, and a mobile app for continuous, real-time monitoring of sweat rate. Utilizing direct 3D printing and scalable microfluidic fabrication, we produced a sensor patch measuring 700 mm 2 and weighing 380 mg. The microfluidic channels are 850 μm wide, and 164 μm thick with serpentine electrodes measure sweat rate using capacitance. Our custom readout electronics capture these changes in capacitance to accurately measure sweat rate, achieving a sensitivity of 0.01 μLmin -1. We validated the sensor's performance against analytical models, simulations, and on-body trials with commercial sensors. This cost-effective, flexible, and fully integrated sweat-sensing solution has significant potential in precision health.
Early detection of bowel cancer is crucial for substantially improving patient outcomes, highlighting the need for less invasive diagnostic methods. Herein, an innovative diagnostic application of vibrating microrobots combined with laser speckle contrast imaging (LSCI) to minimally invasively estimate the elasticity of potential tumors and surrounding healthy tissue is proposed. By positioning a vibrating microrobot on tissue surfaces, the resonant frequencies of the resulting vibrations are analyzed to create detailed elasticity maps. These maps reveal tumor margins and provide critical information about the tumor's properties spatially. This approach leverages the high spatio-temporal resolution and noncontact nature of LSCI to offer a minimally invasive elastography method for potential use in colonoscopy, providing an alternative to complex and time-consuming biopsy analysis and advancing future cancer diagnostics.
Large vision language models (LVLMs) have achieved superior performance on natural image and text tasks, inspiring extensive fine-tuning research. However, their robustness against hallucination in clinical contexts remains understudied. We propose the Medical Visual Hallucination Test (MedVH), a novel evaluation framework assessing hallucination tendencies in both medical-specific and general-purpose LVLMs. MedVH encompasses six tasks targeting medical hallucinations, including two traditional tasks and four novel tasks formatted as multi-choice visual question answering and long response generation. Our extensive experiments with six evaluation metrics reveal that medical LVLMs, despite promising performance on standard medical tasks, are particularly susceptible to hallucinations-often more so than general models. This raises significant concerns about domain-specific model reliability. For real-world applications, medical LVLMs must accurately integrate medical knowledge while maintaining robust reasoning to prevent hallucination. We explore mitigation methods without model-specific fine-tuning, including prompt engineering and collaboration between general and domain-specific models. Our work provides a foundation for future evaluation studies. The dataset is available at PhysioNet: https://physionet.org/content/medvh.
Inspired by bacteria flagella, miniature robots often use a helical shape to propel themselves in fluids at low Reynolds numbers. The helical microstructures in the robots are often rigid and are made by advanced 3D micro-/nanofabrication techniques. However, it remains challenging to fabricate these 3D helical structures without complicated machinery. Herein, for the first time, a magnetically transformable millirobot-TwistBot-with a flexible body that can transform from a simple flat ribbon to a helical shape under an applied magnetic field is reported, enabling its propulsion in viscous fluids. The robot's twisting is modeled using numerical simulation and its geometry is optimized to maximize the twist angle. The unique shape transformation not only allows the propulsion through narrow lumens but also facilitates TwistBot in carrying and delivering solid cargo successfully to the target. The concept of the TwistBot opens new opportunities in designing soft transformable minirobots for targeted cargo delivery.
Most in vitro studies regarding new anticancer treatments are performed on 2D cultures, despite this approach imposes several limitations in recapitulating the real tumor behavior and in predicting the effects of therapy on both cancer and healthy tissues. Herein, advanced in vitro models based on scaffolds that support the 3D growth of glioma cells, further allowing the cocultures with healthy brain cells, are presented. These scaffolds, doped with superparamagnetic iron oxide nanoparticles and obtained through 2-photon polymerization, can be remotely manipulated thanks to an external magnet, thus obtaining biomimetic 3D organization recapitulating the brain cancer microenvironment. From a geometric point of view, the structure is functional to both cell culture on individual unit scaffolds and to tailored cocultures fostered by magnetic-driven unit assembly, also allowing for cell migration thanks to passages/fenestrations on adjacent structures. Leveraging magnetic dragging, for which a mathematical model is introduced, multiple cocultures are achieved, highlighting the high versatility and the user-friendly character of the proposed platform that can help overcome the current challenges in 3D cocultures handling, and open the way to the construction of increasingly biomimetic artificial systems.
Developing small-scale soft continuum robots with large-angle steering capacity and high-precision manipulation offers broad opportunities in various biomedical settings. However, existing continuum robots reach the bottleneck in actuation on account of the contradiction among small size, compliance actuation, large tender range, high precision, and small dynamic error. Herein, a 3D-printed millimeter-scale soft continuum robot with an ultrathin hollow skeleton wall (300 μm) and a large inner-to-outer ratio (0.8) is reported. After coating a thin ferromagnetic elastomer layer (≈100-150 μm), the proposed soft continuum robot equipped with hybrid actuation (tendon- and magnetic-driven mode) achieves large-angle (up to 100°) steering and high-precision (low to 2 μm for static positioning) micromanipulation simultaneously. Specifically, the robot implements an ultralow dynamic tracking error of ≈10 μm, which is ≈30-fold improved than the state of art. Combined with a microneedle/knife or nasopharyngeal swab, the robot reveals the potential for versatile biomedical applications, such as drug injection on the target tissue, diseased tissue ablation, and COVID-19 nasopharyngeal sampling. The proposed millimeter-scale soft continuum robot presents remarkable advances in large-range and high-precise actuation, which provides a new method for miniature continuum robot design and finds broad applications in biomedical engineering.
Wearable sensors are transforming our capacity to monitor a broad range of activities for recreation and health purposes. Developing low-cost personalized sensors on a range of materials could enable broad applicability irrespective of the material substrate. Here, the methods of fabrication, characterization, and application of kirigami graphene strain sensors are described. The dynamic range of these sensors is characterized, showing that the kirigami structure enhances application-specific device performance, demonstrating that this strategy is applicable to a range of materials. We apply this strategy to develop personalized sensors for a variety of measurement frequencies and biological phenomena including evaluation of abdominal distention and respiration in a pig model, as well as human heart rate measurement, limb actuation, and hand gesture interpretation in human volunteers. Through these experiments, we show that this low-cost strategy for customized graphene sensors can be broadly applied across a range of consumer and health applications.
Controlled/living radical polymerization (CLRP) techniques are widely utilized to synthesize advanced and controlled synthetic polymers for chemical and biological applications. While automation has long stood as a high-throughput (HTP) research tool to increase productivity as well as synthetic/analytical reliability and precision, oxygen intolerance of CLRP has limited the widespread adoption of these systems. Recently, however, oxygen-tolerant CLRP techniques, such as oxygen-tolerant photoinduced electron/energy transfer-reversible addition-fragmentation chain transfer (PET-RAFT), enzyme degassing of RAFT (Enz-RAFT), and atom-transfer radical polymerization (ATRP), have emerged. Herein, the use of a Hamilton MLSTARlet liquid handling robot for automating CLRP reactions is demonstrated. Synthesis processes are developed using Python and used to automate reagent handling, dispensing sequences, and synthesis steps required to create homopolymers, random heteropolymers, and block copolymers in 96-well plates, as well as postpolymerization modifications. Using this approach, the synergy between highly customizable liquid handling robotics and oxygen-tolerant CLRP to automate advanced polymer synthesis for HTP and combinatorial polymer research is demonstrated.
Artificial intelligence algorithms are being adopted to analyze medical data, promising faster interpretation to support doctors' diagnostics. The next frontier is to bring these powerful algorithms to implantable medical devices. Herein, a closed-loop solution is proposed, where a cellular neural network is used to detect abnormal wavefronts and wavebrakes in cardiac signals recorded in human tissue is trained to achieve >96% accuracy, >92% precision, >99% specificity, and >93% sensitivity, when floating point precision weights are assumed. Unfortunately, the current hardware technologies for floating point precision are too bulky or energy intensive for compact standalone applications in medical implants. Emerging device technologies, such as memristors, can provide the compact and energy-efficient hardware fabric to support these efforts and can be reliably embedded with existing sensor and actuator platforms in implantable devices. A distributed design that considers the hardware limitations in terms of overhead and limited bit precision is also discussed. The proposed distributed solution can be easily adapted to other medical technologies that require compact and efficient computing, like wearable devices and lab-on-chip platforms.