King Wai Chiu Lai

Older adults frequently face difficulties in activities of daily living (ADLs) due to age-related declines in strength, coordination, and perception. Myoelectric control provides an intuitive human–robot interface by translating muscle activity into assistive commands. However, its practical application is still challenged by signal annotation, multijoint coordination, and cross-task generalization. This study proposes a 3-level intelligent framework for multijoint upper-limb assistance based on electromyography (EMG) to support the daily living activities of older adults. At the physiological level, situation-aware labeling protocols matched to different EMG conditions are proposed to reduce annotation ambiguity and improve robustness to signal changes. At the functional level, focusing on elemental joint activities, a deep backbone model is designed to infer both single-joint movements and coordinated multijoint patterns with an accuracy of 95.34%. At the behavioral level, the model is further distilled to support complex ADL tasks with human–robot interactions while continually incorporating new knowledge without catastrophic forgetting. The framework is implemented in real time on an EMG-controlled multijoint robotic system, providing smooth and coordinated assistance in daily activities. Overall, the proposed framework provides a systematic solution for EMG-based multijoint coordination, encompassing the entire pathway from physiological signal processing to functional intent decoding and behavioral adaptation during daily activities. It offers a technical approach to coordinated upper-limb assistance and lays a broader foundation for the design of practical and adaptive assistive systems, contributing to improved autonomy for older adults and supporting the broader societal goal of healthy aging.

Sarcopenia, characterized by progressive loss of muscle mass and function, significantly impacts the quality of life in aging populations. Early detection and personalized intervention are crucial yet challenging due to the limited accessibility and scalability of traditional diagnostic methods. Building upon our previous work on gait-based assessment, this study presents a novel framework for early sarcopenia-prone risk detection using the five-time sit-to-stand (5TSTS) test, embodying Healthcare Industry 5.0’s vision of mass personalization with human-centered technology. Utilizing the Internet of Things (IoT)-enabled wearable inertial measurement units (IMUs) and advanced analytics, our system segments 5TSTS into four biomechanically significant submotions [standing up (StU), standing transition (StT), sitting down (SiD), and sitting transition (SiT)]. This granular segmentation allows mass personalization in diagnostic evaluations by capturing individual-specific biomechanical profiles via wavelet-based feature extraction and machine learning (ML) techniques. Our framework employs big data analytics tools, including the extreme gradient boosting (XGBoost)-based feature selection and support vector machine synthetic minority oversampling technique (SVMSMOTE), to handle class imbalance and optimize individualized predictive accuracy. Tested on data from 52 elderly participants (aged 65–84 years), the system achieves outstanding personalized classification accuracy—up to 97.97% for multiclass risk stratification and 99.28% for binary (healthy versus sarcopenia-prone) classification—highlighting its potential for precise, patient-specific clinical decision-making. Furthermore, the wireless capability of our IoT-enabled wearable IMUs, coupled with minimal setup requirements, facilitates seamless data integration into cloud-based healthcare systems. This integration supports real-time remote monitoring and personalized health management. By leveraging advanced sensing, analytics, and connectivity technologies, our approach significantly advances personalized, accessible, and scalable sarcopenia-prone risk assessment, thereby contributing directly to the vision of Healthcare Industry 5.0.

Cell spreading is a fundamental process in physiological and pathological contexts, including tissue formation, wound healing, and cancer cell extravasation. Previous studies have examined biophysical mechanisms governing early spreading (around 1-10 min) while biomolecular processes also begin to emerge, yet initial spreading in an even earlier stage (<1 min) remains largely unexplored. Here, we present a deterministic model based on interfacial energy balance─integrating strain energy, surface adhesion energy, and viscous dissipation─to quantitatively describe initial spreading dynamics. Using interference reflection microscopy (IRM), we characterize spreading behaviors of three breast cell lines (MCF-10A, MCF-7, and MDA-MB-231) on extracellular matrix-coated substrates. Model predictions, incorporating biomechanical and biochemical parameters measured through IRM and atomic force microscopy (AFM), show strong agreements with experimental observations. This work provides a universal framework for understanding initial spreading and offers insights into strategies to regulate initial cell spreading, with potential applications in cancer treatment and tissue engineering.

Muscle function and composition are important indicators of age-related health. However, current assessment methods are often complex and expensive, making the early detection of related problems difficult. Therefore, developing a cost-effective and easily accessible daily based detection method is an essential research focus. This study introduces a novel portable porous-structured (i.e., CNT/PDMS nanocomposite) and flexible piezoresistive sensor for evaluating muscle function and relative skeletal muscle mass index, offering advantages of cost-effectiveness, safety, and user-friendliness. The porous architecture significantly enhances sensitivity, while the flexible design ensures excellent conformability to the skin and adaptability to complex body movements. The prototype sensor demonstrates a linear detection range of 0-39 kPa with dual-stage sensitivities of 0.03398 kPa^-1 (0-7 kPa) and 0.000922 kPa^-1 (7-39 kPa). The sensor maintains stable performance for over a week and exhibits reliable operation unaffected by body temperature or perspiration, and the material cost does not exceed 10 HKD. The gait data can be easily collected by wearing the sensor on the left gastrocnemius muscle. Our study encompassed 23 participants from both the elderly and young age groups. The supervised learning achieved a maximum accuracy of 93.48% in distinguishing between the elderly and the young subjects. Unsupervised learning analysis further validated the efficacy of our flexible sensor in muscle function assessment. Additionally, an Adaboost regression model was employed to predict the relative skeletal muscle mass index, achieving a mean error of 2.8%. This flexible sensor demonstrates significant potential for the daily monitoring of muscle function and mass, enabling early detection and prevention of sarcopenia and other muscle-related disorders. Its wearable and noninvasive characteristics make it an attractive solution for muscle assessment in clinical, sports, and home environments.

Liquid metals (LMs) have emerged as prominent materials for flexible pressure sensing owing to their exceptional conductivity and fluidity. Typically, external loads induce changes in the shape and volume of conductive LM pathways to achieve pressure detection. To optimize sensor's pressure sensitivity, theoretical modeling and finite element simulations are employed to investigate the effects of microchannel thickness and patterns. Results revealed that symmetrical patterns and thinner microchannels significantly enhanced sensor's pressure sensitivity. Furthermore, a novel polyvinyl alcohol (PVA) sacrificial template method is proposed that enables the flexible fabrication of microchannels with various shapes and thicknesses, achieving a minimum channel thickness of 25 µm. The LM sensor demonstrates excellent performance metrics, including a maximum sensitivity of 0.01212 kPa⁻¹, a wide detection range from 0 to 60 kPa, and remarkable cyclic stability up to 3000 cycles. In practical applications, the sensor enables high‐precision monitoring of various human movements, whereas sensor arrays can effectively detect force distributions across different objects. This paper presents a straightforward and efficient approach for regulating and designing conductive microchannel paths. Additionally, the integration of finite element simulations facilitates optimal sensor pattern design, and the fabricated sensors show tremendous potential for applications in pressure recognition and motion detection.

Sarcopenia is a progressive condition characterized by age-related losses in muscle mass and strength, and irreversible in its advanced stages. While sarcopenia negatively impacts daily living, accurately, quickly and economically assessing its effects can be challenging due to individual variability in activity levels. This study introduced a novel approach for assessing the risk of sarcopenia-prone using machine learning and wearable IoT (Internet of Things) sensors. A total of 53 community-dwelling older adults aged 65+ underwent gait analysis using dual sensors. Nineteen gait features were extracted from each cycle and used to train classification algorithms to categorize participants as healthy, risk level 1, risk level 2, or risk level 3 for sarcopenia. Binary classification of healthy vs. sarcopenic-prone achieved 97.41% accuracy on average, while four-class classification averaged 94.67%. Notably, the research discovered worsening gait symmetry with increasing sarcopenia-prone severity. These results indicate IoT sensor-assessed gait may serve as a sensitive indicator for daily sarcopenia-prone screening. Accurate assessment of sarcopenia-prone individuals can be achieved through only a 4-meter walking test, significantly reducing the burden for older adults. This approach offers a cost-effective, convenient, and accurate method for early sarcopenia risk detection and intervention, potentially improving quality of life for older adults. This system could also aid in creating widely applicable monitoring products for assessing sarcopenia risk, supporting IoT, and thereby enabling early identification and intervention for individuals at risk of this condition.

Liquid metals are highly conductive like metallic materials and have excellent deformability due to their liquid state, making them rather promising for flexible and stretchable wearable sensors. However, patterning liquid metals on soft substrates has been a challenge due to high surface tension. In this paper, a new method is proposed to overcome the difficulties in fabricating liquid-state strain sensors. The method involves adding nickel powder particles to the liquid metal to maintain the liquid metal's fluidity while lowering the surface tension so that the liquid metal can be easily patterned on a soft substrate using magnets. With the addition of 12 wt % nickel powder (40 μm) to the liquid metal, a gauge factor of 5.17 can be achieved at 300% strain. In addition, by the integration of multiple strain sensors in a smart glove to monitor 14 joints of the human hand, 10 sign language gestures can be recognized by comparing the results of five different machine learning models, among which the quadratic discriminative analysis model can be accomplished with an accuracy rate of 100%. The magnetically patterned nickel-containing liquid-metal strain sensors proposed in this study have a wide range of applications in intelligent soft robots and human-machine interfaces.

The patch clamp technique is a fundamental tool for investigating ion channel dynamics and electrophysiological properties. This study proposes the first artificial intelligence framework for characterizing multiple ion channel kinetics of whole‐cell recordings. The framework integrates machine learning for anomaly detection and deep learning for multi‐class classification. The anomaly detection excludes recordings that are incompatible with ion channel behavior. The multi‐class classification combined a 1D convolutional neural network, bidirectional long short‐term memory, and an attention mechanism to capture the spatiotemporal patterns of the recordings. The framework achieves an accuracy of 97.58% in classifying 124 test datasets into six categories based on ion channel kinetics. The utility of the novel framework is demonstrated in two applications: Alzheimer's disease drug screening and nanomatrix‐induced neuronal differentiation. In drug screening, the framework illustrates the inhibitory effects of memantine on endogenous channels, and antagonistic interactions among potassium, magnesium, and calcium ion channels. For nanomatrix‐induced differentiation, the classifier indicates the effects of differentiation conditions on sodium and potassium channels associated with action potentials, validating the functional properties of differentiated neurons for Parkinson's disease treatment. The proposed framework is promising for enhancing the efficiency and accuracy of ion channel kinetics analysis in electrophysiological research.