Search for: All records

Abstract Recent advancements in wearable sensor technologies have enabled real-time monitoring of physiological and biochemical signals, opening new opportunities for personalized healthcare applications. However, conventional wearable devices often depend on rigid electronics components for signal transduction, processing, and wireless communications, leading to compromised signal quality due to the mechanical mismatches with the soft, flexible nature of human skin. Additionally, current computing technologies face substantial challenges in efficiently processing these vast datasets, with limitations in scalability, high power consumption, and a heavy reliance on external internet resources, which also poses security risks. To address these challenges, we have developed a miniaturized, standalone, chip-less wearable neuromorphic system capable of simultaneously monitoring, processing, and analyzing multimodal physicochemical biomarker data (i.e., metabolites, cardiac activities, and core body temperature). By leveraging scalable printing technology, we fabricated artificial synapses that function as both sensors and analog processing units, integrating them alongside printed synaptic nodes into a compact wearable system embedded with a medical diagnostic algorithm for multimodal data processing and decision making. The feasibility of this flexible wearable neuromorphic system was demonstrated in sepsis diagnosis and patient data classification, highlighting the potential of this wearable technology for real-time medical diagnostics. 

Text-to-video (T2V) generation has been recently enabled by transformer-based diffusion models, but current T2V models lack capabilities in adhering to the real-world common knowledge and physical rules, due to their limited understanding of physical realism and deficiency in temporal modeling. Existing solutions are either data-driven or require extra model inputs, but cannot be generalizable to out-of-distribution domains. In this paper, we present PhyT2V, a new data-independent T2V technique that expands the current T2V model’s capability of video generation to out-of-distribution domains, by enabling chain-of-thought and step-back reasoning in T2V prompting. Our experiments show that PhyT2V improves existing T2V models’ adherence to real-world physical rules by 2.3x, and achieves 35% improvement compared to T2V prompt enhancers. 

Federated Learning (FL) can be affected by data and device heterogeneities, caused by clients' different local data distributions and latencies in uploading model updates (i.e., staleness). Traditional schemes consider these heterogeneities as two separate and independent aspects, but this assumption is unrealistic in practical FL scenarios where these heterogeneities are intertwined. In these cases, traditional FL schemes are ineffective, and a better approach is to convert a stale model update into a unstale one. In this paper, we present a new FL framework that ensures the accuracy and computational efficiency of this conversion, hence effectively tackling the intertwined heterogeneities that may cause unlimited staleness in model updates. Our basic idea is to estimate the distributions of clients' local training data from their uploaded stale model updates, and use these estimations to compute unstale client model updates. In this way, our approach does not require any auxiliary dataset nor the clients' local models to be fully trained, and does not incur any additional computation or communication overhead at client devices. We compared our approach with the existing FL strategies on mainstream datasets and models, and showed that our approach can improve the trained model accuracy by up to 25% and reduce the number of required training epochs by up to 35%. 

Two‐photon polymerization (TPP) enables the fabrication of intricate 3D microstructures with submicron precision, offering significant potential in biomedical applications like tissue engineering. In such applications, to print materials and structures with defined mechanics, it is crucial to understand how TPP printing parameters impact the material properties in a physiologically relevant liquid environment. Herein, an experimental approach utilizing microscale tensile testing (μTT) for the systematic measurement of TPP‐fabricated microfibers submerged in liquid as a function of printing parameters is introduced. Using a diurethane dimethacrylate‐based resin, the influence of printing parameters on microfiber geometry is first explored, demonstrating cross‐sectional areas ranging from 1 to 36 μm². Tensile testing reveals Young's moduli between 0.5 and 1.5 GPa and yield strengths from 10 to 60 MPa. The experimental data show an excellent fit with the Ogden hyperelastic polymer model, which enables a detailed analysis of how variations in writing speed, laser power, and printing path influence the mechanical properties of TPP microfibers. The μTT method is also showcased for evaluating multiple commercial resins and for performing cyclic loading experiments. Collectively, this study builds a foundation toward a standardized microscale tensile testing framework to characterize the mechanical properties of TPP printed structures. 

Managing stress is essential for mental and physical health, yet current methods rely on subjective self-assessments or indirect physiological measurements, often lacking accuracy. Existing wearable sensors primarily target a single stress hormone, cortisol, using single-point measurements that fail to capture real-time changes and distinguish between acute and chronic stress. To address this, we present Stressomic, a wearable multiplexed microfluidic biosensor for noninvasive monitoring of cortisol, epinephrine, and norepinephrine in sweat. Stressomic integrates iontophoresis-driven sweat extraction with bursting valve-regulated microfluidic channels for continuous sampling and analysis. Gold nanodendrite–decorated laser-engraved graphene electrodes achieve picomolar-level sensitivity, enabling simultaneous detection of multiple stress hormones. Electrochemical assays and human studies demonstrate that Stressomic reliably tracks hormone fluctuations in response to physical, psychological, and pharmacological stressors. Distinct temporal patterns reveal the dynamic interplay between the hypothalamic-pituitary-adrenal axis and the sympathetic nervous system. This platform enables continuous, multiplexed stress profiling, offering opportunities for early detection of maladaptive responses, personalized stress management, and deeper insights into stress biology. 

Wearable wireless sensing systems capable of monitoring critical biomarkers are increasingly valued for health monitoring and disease management. However, current systems often compromise on sensitivity, multi-parameter sensing capacity, compactness, power consumption, and wearing comfort, limiting their skin-interfaced applications. Here, we present a wireless, battery-free, multiplexed sensing system based on a high-order exceptional point (EP)—a singularity in non-Hermitian realm. This system features multiple distinct resonance frequencies that sensitively respond to physiological variations, exhibiting over 10-fold sensitivity improvement and 50% size reduction compared to conventional resonator-based systems. Moreover, the sensor employs a strain-resilient liquid metal composite and a soft substrate, ensuring wearing comfort, mechanical reliability, and antimicrobial properties. The effectiveness of this system is demonstrated through in situ monitoring of skin temperature and sweat electrolytes during exercise and prolonged monitoring of sweat glucose and ammonium in different subjects. The proposed system is highly transformative, offering benefits for various wearable applications.