Deep Learning-Based Link Quality Estimation for RIS-Assisted UAV-Enabled Wireless Communications System.

In recent years, unmanned aerial vehicles (UAVs) have become a valuable platform for many applications, including communication networks. UAV-enabled wireless communication faces challenges in complex urban and dynamic environments. UAVs can suffer from power limitations and path losses caused by non-line-of-sight connections, which may hamper communication performance. To address these issues, reconfigurable intelligent surfaces (RIS) have been proposed as helpful technologies to enhance UAV communication networks. However, due to the high mobility of UAVs, complex channel environments, and dynamic RIS configurations, it is challenging to estimate the link quality of ground users. In this paper, we propose a link quality estimation model using a gated recurrent unit (GRU) to assess the link quality of ground users for a multi-user RIS-assisted UAV-enabled wireless communication system. Our proposed framework uses a time series of user channel data and RIS phase shift information to estimate the quality of the link for each ground user. The simulation results showed that the proposed GRU model can effectively and accurately estimate the link quality of ground users in the RIS-assisted UAV-enabled wireless communication network.

Glove-based sensors for multimodal monitoring of natural sweat.

Sweat sensors targeting exercise or chemically induced sweat have shown promise for noninvasive health monitoring. Natural thermoregulatory sweat is an attractive alternative as it can be accessed during routine and sedentary activity without impeding user lifestyles and potentially preserves correlations between sweat and blood biomarkers. We present simple glove-based sensors to accumulate natural sweat with minimal evaporation, capitalizing on high sweat gland densities to collect hundreds of microliters in just 30 min without active sweat stimulation. Sensing electrodes are patterned on nitrile gloves and finger cots for in situ detection of diverse biomarkers, including electrolytes and xenobiotics, and multiple gloves or cots are worn in sequence to track overarching analyte dynamics. Direct integration of sensors into gloves represents a simple and low-overhead scheme for natural sweat analysis, enabling sweat-based physiological monitoring to become practical and routine without requiring highly complex or miniaturized components for analyte collection and signal transduction.

Nicotine Monitoring with a Wearable Sweat Band.

The tobacco epidemic is a public health threat that has taken a heavy toll of lives around the globe each year. Smoking affects both the smokers and those who are exposed to secondhand smoke, and careful tracking of exposure can be key to mitigating the potential hazards. For smokers, the variation of chemical compositions between commercial cigarettes has led to ambiguity in estimating the health risks, both for active smokers and others involuntarily exposed to tobacco smoke and byproducts. In this regard, sweat possesses an attractive opportunity to monitor smoke exposure due to sweat's abundance in biomolecules and its great accessibility. Here, we present a wearable sweat band to monitor nicotine, a prominent ingredient in cigarettes, as a viable way to quantitatively assess a wearer's exposure to smoking. Both smokers and normal subjects are tested to demonstrate the use of this device for smoke-related health monitoring. Our results exhibit confirmable and elevated nicotine levels in sweat for subjects inhaling cigarette smoke. This continuous and personalized sweat sensing device is leverage to monitor smoke pollution for a potentially broad population.

Regional and correlative sweat analysis using high-throughput microfluidic sensing patches toward decoding sweat.

Recent technological advancements in wearable sensors have made it easier to detect sweat components, but our limited understanding of sweat restricts its application. A critical bottleneck for temporal and regional sweat analysis is achieving uniform, high-throughput fabrication of sweat sensor components, including microfluidic chip and sensing electrodes. To overcome this challenge, we introduce microfluidic sensing patches mass fabricated via roll-to-roll (R2R) processes. The patch allows sweat capture within a spiral microfluidic for real-time measurement of sweat parameters including [Na+], [K+], [glucose], and sweat rate in exercise and chemically induced sweat. The patch is demonstrated for investigating regional sweat composition, predicting whole-body fluid/electrolyte loss during exercise, uncovering relationships between sweat metrics, and tracking glucose dynamics to explore sweat-to-blood correlations in healthy and diabetic individuals. By enabling a comprehensive sweat analysis, the presented device is a crucial tool for advancing sweat testing beyond the research stage for point-of-care medical and athletic applications.

A multi-modal sweat sensing patch for cross-verification of sweat rate, total ionic charge, and Na+ concentration.

Sweat sensors introduced in recent years have targeted a variety of sweat features and biomarkers for non-invasive health monitoring. Amongst these targets, reliable monitoring of sweat rate is crucial due to its modulation of sweat analyte concentrations and its intrinsic significance to numerous medical and physiological health conditions. Here we present a sweat rate sensor structure comprising of electrodes with interdigitated fingers in a microfluidic channel. Each time the accumulating sweat impinges on an electrode finger, the sensor reports a jump in admittance that can be simply and efficiently counted to estimate sweat rate, overcoming selectivity limitations of previously reported sweat rate sensors. We further integrate an impedimetric sensor for measuring total ionic charge concentration and an electrochemical Na+ sensor, together creating a multi-modal system for analyzing fluid and electrolyte secretion. We demonstrate how low analyte diffusion rates through this microfluidic device allow for multi-purpose sensor function, including utilizing the sweat rate sensor signal to corroborate total ionic sensor measurements. This cross-verification capability ensures data integrity in real time, satisfying a vital consideration for personalized healthcare technologies. We use the presented patch for continuous analysis of sweat rate, total ionic charge concentration, and Na+ concentration during exercise, while demonstrating how multi-modal cross-verification brings new trust to sensor readings.

Wearable Sweat Band for Noninvasive Levodopa Monitoring.

Levodopa is the standard medication clinically prescribed to patients afflicted with Parkinson's disease. In particular, the monitoring and optimization of levodopa dosage are critical to mitigate the onset of undesired fluctuations in the patients' physical and emotional conditions such as speech function, motor behavior, and mood stability. The traditional approach to optimize levodopa dosage involves evaluating the subjects' motor function, which has many shortcomings due to its subjective and limited quantifiable nature. Here, we present a wearable sweat band on a nanodendritic platform that quantitatively monitors levodopa dynamics in the body. Both stationary iontophoretic induction and physical exercise are utilized as our methods of sweat extraction. The sweat band measures real-time pharmacokinetic profiles of levodopa to track the dynamic response of the drug metabolism. We demonstrated the sweat band's functionalities on multiple subjects with implications toward the systematic administering of levodopa and routine management of Parkinson's disease.

A Fully Integrated and Self-Powered Smartwatch for Continuous Sweat Glucose Monitoring.

Wearable devices for health monitoring and fitness management have foreseen a rapidly expanding market, especially those for noninvasive and continuous measurements with real-time display that provide practical convenience and eliminated safety/infection risks. Herein, a self-powered and fully integrated smartwatch that consists of flexible photovoltaic cells and rechargeable batteries in the forms of a "watch strap", electrochemical glucose sensors, customized circuits, and display units integrated into a "dial" platform is successfully fabricated for real-time and continuous monitoring of sweat glucose levels. The functionality of the smartwatch, including sweat glucose sensing, signal processing, and display, can be supported with the harvested/converted solar energy without external charging devices. The Zn-MnO2 batteries serve as intermediate energy storage units and the utilization of aqueous electrolytes eliminated safety concerns for batteries, which is critical for wearable devices. Such a wearable system in a smartwatch fashion realizes integration of energy modules with self-powered capability, electrochemical sensors for noninvasive glucose monitoring, and in situ and real-time signal processing/display in a single platform for the first time. The as-fabricated fully integrated and self-powered smartwatch also provides a promising protocol for statistical study and clinical investigation to reveal correlations between sweat compositions and human body dynamics.

Roll-to-Roll Gravure Printed Electrochemical Sensors for Wearable and Medical Devices.

As recent developments in noninvasive biosensors spearhead the thrust toward personalized health and fitness monitoring, there is a need for high throughput, cost-effective fabrication of flexible sensing components. Toward this goal, we present roll-to-roll (R2R) gravure printed electrodes that are robust under a range of electrochemical sensing applications. We use inks and electrode morphologies designed for electrochemical and mechanical stability, achieving devices with uniform redox kinetics printed on 150 m flexible substrate rolls. We show that these electrodes can be functionalized into consistently high performing sensors for detecting ions, metabolites, heavy metals, and other small molecules in noninvasively accessed biofluids, including sensors for real-time, in situ perspiration monitoring during exercise. This development of robust and versatile R2R gravure printed electrodes represents a key translational step in enabling large-scale, low-cost fabrication of disposable wearable sensors for personalized health monitoring applications.

A Wearable Microfluidic Sensing Patch for Dynamic Sweat Secretion Analysis.

Wearable sweat sensing is a rapidly rising research area driven by its promising potential in health, fitness, and diagnostic applications. Despite the growth in the field, major challenges in relation to sweat metrics remain to be addressed. These challenges include sweat rate monitoring for its complex relation with sweat compositions and sweat sampling for sweat dynamics studies. In this work, we present a flexible microfluidic sweat sensing patch that enhances real-time electrochemical sensing and sweat rate analysis via sweat sampling. The device contains a spiral-patterned microfluidic component that is embedded with ion-selective sensors and an electrical impedance-based sweat rate sensor on a flexible plastic substrate. The patch is enabled to autonomously perform sweat analysis by interfacing the sensing component with a printed circuit board that is capable of on-site signal conditioning, analysis, and transmission. Progressive sweat flow in the microfluidic device, governed by the pressure induced by the secreted sweat, enhances sweat sampling and electrochemical detection via a defined sweat collection chamber and a directed sweat route. The characteristic of the sweat rate sensor is validated through a theoretical simulation, and the precision and accuracy of the flow rate is verified with a commercial syringe pump and a Macroduct sweat collector. On-body simultaneous monitoring of ion (H+, Na+, K+, Cl-) concentration and sweat rate is also demonstrated for sensor functionality. This sweat sensing patch provides an integrated platform for a comprehensive sweat secretion analysis and facilitates physiological and clinical investigations by closely monitoring interrelated sweat parameters.

Methylxanthine Drug Monitoring with Wearable Sweat Sensors.

Drug monitoring plays crucial roles in doping control and precision medicine. It helps physicians tailor drug dosage for optimal benefits, track patients' compliance to prescriptions, and understand the complex pharmacokinetics of drugs. Conventional drug tests rely on invasive blood draws. While urine and sweat are attractive alternative biofluids, the state-of-the-art methods require separate sample collection and processing steps and fail to provide real-time information. Here, a wearable platform equipped with an electrochemical differential pulse voltammetry sensing module for drug monitoring is presented. A methylxanthine drug, caffeine, is selected to demonstrate the platform's functionalities. Sweat caffeine levels are monitored under various conditions, such as drug doses and measurement time after drug intake. Elevated sweat caffeine levels upon increasing dosage and confirmable caffeine physiological trends are observed. This work leverages a wearable sweat sensing platform toward noninvasive and continuous point-of-care drug monitoring and management.

Wearable Microfluidic Diaphragm Pressure Sensor for Health and Tactile Touch Monitoring.

Flexible pressure sensors have many potential applications in wearable electronics, robotics, health monitoring, and more. In particular, liquid-metal-based sensors are especially promising as they can undergo strains of over 200% without failure. However, current liquid-metal-based strain sensors are incapable of resolving small pressure changes in the few kPa range, making them unsuitable for applications such as heart-rate monitoring, which require a much lower pressure detection resolution. In this paper, a microfluidic tactile diaphragm pressure sensor based on embedded Galinstan microchannels (70 µm width × 70 µm height) capable of resolving sub-50 Pa changes in pressure with sub-100 Pa detection limits and a response time of 90 ms is demonstrated. An embedded equivalent Wheatstone bridge circuit makes the most of tangential and radial strain fields, leading to high sensitivities of a 0.0835 kPa-1 change in output voltage. The Wheatstone bridge also provides temperature self-compensation, allowing for operation in the range of 20-50 °C. As examples of potential applications, a polydimethylsiloxane (PDMS) wristband with an embedded microfluidic diaphragm pressure sensor capable of real-time pulse monitoring and a PDMS glove with multiple embedded sensors to provide comprehensive tactile feedback of a human hand when touching or holding objects are demonstrated.

3D Printed "Earable" Smart Devices for Real-Time Detection of Core Body Temperature.

Real-time detection of basic physiological parameters such as blood pressure and heart rate is an important target in wearable smart devices for healthcare. Among these, the core body temperature is one of the most important basic medical indicators of fever, insomnia, fatigue, metabolic functionality, and depression. However, traditional wearable temperature sensors are based upon the measurement of skin temperature, which can vary dramatically from the true core body temperature. Here, we demonstrate a three-dimensional (3D) printed wearable "earable" smart device that is designed to be worn on the ear to track core body temperature from the tympanic membrane (i.e., ear drum) based on an infrared sensor. The device is fully integrated with data processing circuits and a wireless module for standalone functionality. Using this smart earable device, we demonstrate that the core body temperature can be accurately monitored regardless of the environment and activity of the user. In addition, a microphone and actuator are also integrated so that the device can also function as a bone conduction hearing aid. Using 3D printing as the fabrication method enables the device to be customized for the wearer for more personalized healthcare. This smart device provides an important advance in realizing personalized health care by enabling real-time monitoring of one of the most important medical parameters, core body temperature, employed in preliminary medical screening tests.

A Wearable Electrochemical Platform for Noninvasive Simultaneous Monitoring of Ca(2+) and pH.

Homeostasis of ionized calcium in biofluids is critical for human biological functions and organ systems. Measurement of ionized calcium for clinical applications is not easily accessible due to its strict procedures and dependence on pH. pH balance in body fluids greatly affects metabolic reactions and biological transport systems. Here, we demonstrate a wearable electrochemical device for continuous monitoring of ionized calcium and pH of body fluids using a disposable and flexible array of Ca(2+) and pH sensors that interfaces with a flexible printed circuit board. This platform enables real-time quantitative analysis of these sensing elements in body fluids such as sweat, urine, and tears. Accuracy of Ca(2+) concentration and pH measured by the wearable sensors is validated through inductively coupled plasma-mass spectrometry technique and a commercial pH meter, respectively. Our results show that the wearable sensors have high repeatability and selectivity to the target ions. Real-time on-body assessment of sweat is also performed, and our results indicate that calcium concentration increases with decreasing pH. This platform can be used in noninvasive continuous analysis of ionized calcium and pH in body fluids for disease diagnosis such as primary hyperparathyroidism and kidney stones.

Deep Learning in Label-free Cell Classification.

Label-free cell analysis is essential to personalized genomics, cancer diagnostics, and drug development as it avoids adverse effects of staining reagents on cellular viability and cell signaling. However, currently available label-free cell assays mostly rely only on a single feature and lack sufficient differentiation. Also, the sample size analyzed by these assays is limited due to their low throughput. Here, we integrate feature extraction and deep learning with high-throughput quantitative imaging enabled by photonic time stretch, achieving record high accuracy in label-free cell classification. Our system captures quantitative optical phase and intensity images and extracts multiple biophysical features of individual cells. These biophysical measurements form a hyperdimensional feature space in which supervised learning is performed for cell classification. We compare various learning algorithms including artificial neural network, support vector machine, logistic regression, and a novel deep learning pipeline, which adopts global optimization of receiver operating characteristics. As a validation of the enhanced sensitivity and specificity of our system, we show classification of white blood T-cells against colon cancer cells, as well as lipid accumulating algal strains for biofuel production. This system opens up a new path to data-driven phenotypic diagnosis and better understanding of the heterogeneous gene expressions in cells.