Duo Wai-Chi Wong

Finite element (FE) analysis has become a cornerstone of computational orthopaedics, yet foot and ankle modelling faces major translation barriers in bridging in silico predictions to real-world clinical and industrial applications. This narrative umbrella review synthesizes published review literature to map the evolving application landscape across disease-specific pathomechanics, surgical interventions, and footwear or orthotic design. We examine prevailing FE workflows and highlight pervasive methodological vulnerabilities that impede clinical adoption, including widespread reliance on single-subject designs, oversimplifications, and validation constraints. Furthermore, we present meta-perspectives on translational barriers, such as computational and labour bottlenecks, input parameter identifiability, and the disconnect between continuum mechanical stress metrics and dynamic biological tissue failure thresholds; alongside critical viewpoints on the Validating-Simulation Paradox, data quality versus quantity, and epistemological questions surrounding virtual cohorts and cross-matched data. We explore how artificial intelligence (AI) addresses these challenges by expanding sample sizes through generative models, delivering rapid feedback via data-driven or physics-aware surrogates, and minimizing manual inputs using agentic AI. However, current AI architectures cannot replace numerical solvers due to domain boundaries and physical anchoring requirements. Ultimately, we highlight how AI can help resolve translational barriers by enhancing uncertainty quantification and fostering clinical trust to bridge the code-to-clinic divide.

Accurate, unobtrusive in-bed occupancy-posture Internet of Thing (IoT) monitoring is critical both for timely detection of “missing patient” events, reducing false alarms in nocturnal surveillance, and informing sleep posture-related health risk management. This article presents a novel two-stage, multitask, multiple ultrawideband (UWB) radar IoT system for joint in-bed occupancy detection and fine-grained sleep posture classification, using an array of eight UWB radars. The technical novelty centers on three mechanisms: a view-weighted multiradar network (VWM-Net) that adaptively fuses single-radar echo maps into a spatial echo feature map (SEFM); an integration of DenseNet and ConvNeXt2 (DCNX2); and a cross-modal generative model that translates SEFMs into human-interpretable depth posture images. Specifically, a UCYC-GAN is proposed by integrating a U-Net structure with a CycleGAN generator, and its visual utility is compared against NICE-GAN and bidirectional diffusion bridge model (BDBM). The system is trained and validated on a cohort of 100 participants across ten fine-grained postures and four blanket conditions, supplemented with a life-sized dummy and cluttered objects to model nonhuman occupancy and environmental distractors. In the first stage, a dedicated classifier distinguishes humans from nonhuman entities with 99.09% accuracy. In the second stage, the multihead DCNX2 backbone performed multitask classification, achieving 93.75% accuracy for ten fine-grained postures, 98% for four coarse-grained postures, 88.25% for blanket-coverage types, 98.13% for blanket presence, and 95.13% for participant gender. The overall can effectively mitigate false positives from environmental clutter, advances noncontact in-bed occupancy-posture surveillance, and, through a cloud-enabled web dashboard and APIs, supports real-time visualization and retrospective data analysis.

Introduction: Frailty in chronic kidney disease (CKD) patients is associated with increased risk of adverse health outcomes. Understanding the contributing factors to frailty in this population is crucial for developing targeted interventions and improving patient care. The objective of this study is to identify and quantify potential risk factors associated with frailty in chronic kidney disease patients. Methods: We conducted a cross‐sectional study using data from the China Health and Retirement Longitudinal Study (CHARLS) cohorts of 2011 and 2015. A Bayesian mixed‐effects logistic regression model was utilized to analyze the relationship between selected features and frailty in chronic kidney disease. Results: Of 1,924 participants, about one‐third ( n = 634) were classified as frail. Depression emerged as the strongest predictor of frailty (adjusted OR 289.49, 95%CrI 47.24–2284.82). Other significant factors included stroke (adjusted OR 18.04, 95%CrI 1.94–183.99), chronic lung disease (adjusted OR 7.96, 95%CrI 1.84–39.44), and rheumatism (adjusted OR 3.84, 95%CrI 1.12–15.14). Poor vision (adjusted OR 5.84, 95% CrI 1.68–25.25) and poor sleep quality (adjusted OR 5.12, 95% CrI 1.53–21.13), though the latter showed some instability after data imputation, were also associated with higher frailty risk. Elevated cystatin C levels showed a strong positive association (adjusted OR 52.46, 95%CrI 9.51–356.12), while higher creatinine levels were associated with lower frailty risk (adjusted OR 0.24, 95%CrI 0.07–0.81). Conclusion: This study identified several potential contributing factors to frailty in CKD patients, with depression emerging as the strongest predictor. The counterintuitive relationship between creatinine levels and frailty underscores the complex interplay between muscle mass/quality and kidney function in frailty development, warranting further investigation.

Early diagnosis of Alzheimer’s disease (AD) remains challenging due to the extremely low concentration of relevant biomarkers and the limited sensitivity of conventional detection techniques. In this study, we present a carbon nanotube field-effect transistor (CNT-FET) immunosensor for label-free detection of phosphorylated tau at threonine 217 (p-tau217). The device employs a Y₂O₃/HfO₂ dielectric layer and gold nanoparticles (AuNPs) to improve biofunctionalization, with anti-p-tau217 antibodies immobilized on the CNT channels. In phosphate-buffered saline (PBS), the sensor exhibited a linear response over a concentration range of 3 fM to 30 pM (R² = 0.973) and achieved a limit of detection (LOD) of 1.66 fM. The device demonstrated high selectivity, with a normalized signal response (NSR) for p-tau217 that was 5–6 times higher than for human serum albumin (HSA) and p-tau231, even at 1000-fold higher concentrations of these interferents. The sensor exhibited reproducibility with a relative standard deviation (RSD) of 4.8% (n = 9) and storage stability with only a 10% decrease in signal after 7 days at 4 °C. Mechanistic analysis indicated that the net positive charge and structural flexibility of the p-tau217 peptide led to a reduction in drain current upon binding, consistent with electrostatic gating effects in p-type CNT-FETs. Current limitations include the absence of standardized p-tau217 reference materials. Future work will focus on validation with clinical samples. This CNT-FET platform enables rapid, minimally invasive detection of p-tau217 and holds strong potential for integration into clinical workflows to facilitate early AD diagnosis.

Background: Swallowing is a complex biomechanical process, and its impairment (dysphagia) poses major health risks for older adults. Current diagnostic methods such as videofluoroscopic swallowing (VFSS) and fiberoptic endoscopic evaluation of swallowing (FEES) are effective but invasive, resource-intensive, and unsuitable for continuous monitoring. This study proposes a novel end-to-end RGB–D framework for automated swallowing event localization in continuous video streams. Methods: The framework enhances the AdaTAD backbone through three key innovations: (i) finding the optimal strategy to integrate depth information to capture subtle neck movements, (ii) examining the best adapter design for efficient temporal feature adaptation, and (iii) introducing a Kolmogorov–Arnold Network (KAN) decoder that leverages Chebyshev polynomials for non-linear temporal modeling. Evaluation on a proprietary swallowing dataset comprising 641 clips and 3153 annotated events demonstrated the effectiveness of the proposed framework. We analysed and compared the modification strategy across designs of adapters, decoders, input channel combinations, regression methods, and patch embedding techniques. Results: The optimized configuration (VideoMAE + GRNConvNeXtAdapter + KAN + RGD + boundary regression + sinusoidal embedding) achieved an average mAP of 83.25%, significantly surpassing the baseline I3D + RGB + MLP model (61.55%). Ablation studies further confirmed that each architectural component contributed incrementally to the overall improvement. Conclusions: These results establish the feasibility of accurate, non-invasive, and automated swallowing event localization using depth-augmented video. The proposed framework paves the way for practical dysphagia screening and long-term monitoring in clinical and home-care environments.

Swallowing is a complex biomechanical process, and its impairment (dysphagia) poses major health risks for older adults. Current diagnostic methods such as videofluoroscopic swallowing (VFSS) and fiberoptic endoscopic evaluation of swallowing (FEES) are effec-tive but invasive, resource-intensive, and unsuitable for continuous monitoring. This study proposes a novel end-to-end RGB–D framework for automated swallowing event localization in continuous video streams. The framework enhances the AdaTAD back-bone through three key innovations: (i) an RGD input strategy that substitutes the noisy blue channel with depth information to capture subtle neck movements, (ii) a GRNCon-vNeXtAdapter that incorporates Global Response Normalization for efficient temporal feature adaptation, and (iii) a Kolmogorov–Arnold Network (KAN) decoder that leverages Chebyshev polynomials for non-linear temporal modeling. (iv) boundary regression head and sinusoidal patch embedding to further optimize model precision. Evaluation on a proprietary swallowing dataset comprising 641 clips and 3,153 annotated events demon-strated the effectiveness of the proposed framework. We analysed and compared modifi-cation strategy across design of adapter, decoder, input channel combination, regression method, and patch embedding technique. The optimized configuration (VideoMAE + GRNConvNeXtAdapter + KAN + RGD + boundary regression + sinusoidal embedding) achieved an average mAP of 83.25%, significantly surpassing the baseline I3D + RGB + MLP model (61.55%). Ablation studies further confirmed that each architectural compo-nent contributed incrementally to the overall improvement. These results establish the feasibility of accurate, non-invasive, and automated swallowing event localization using depth-augmented video. The proposed framework paves the way for practical dysphagia screening and long-term monitoring in clinical and home-care environments.

Background: Hallux valgus is a common foot deformity that may alter plantar pressure distribution. This study aimed to determine whether hallux valgus consistently reduces medial forefoot (hallux and medial metatarsal) load and impulse through a meta‐analysis of plantar pressure data and increases central metatarsal loading.Methods: A systematic review and meta‐analysis were conducted following PRISMA guidelines. PubMed, Web of Science, Scopus, CINAHL, and Embase were searched for studies comparing plantar pressure distribution between hallux valgus and non‐hallux valgus feet. Random‐effects meta‐analyses were performed using standardized mean differences (SMD) for load (force and pressure) and impulse (force and pressure‐time integrals) measures in hallux, medial metatarsal, and central metatarsal regions. Risk of bias was assessed using the JBI critical appraisal checklist for analytical cross‐sectional studies.Results: Twenty‐one studies (3911 participants and 6407 feet) were included. In‐shoe measurements showed significant reductions in hallux loading (force SMD: −0.78 [95% CI: −1.08, −0.49]; force‐time integral SMD: −0.39 [−0.71, −0.07]). Platform‐based measurements yielded inconsistent results. No conclusive evidence was found for reduced medial metatarsal loading or increased central metatarsal loading. Measurement modality significantly influenced results, with platform‐based systems generally showing higher heterogeneity than in‐shoe sensors. Nine (43%) and 12 studies (57%) had a high risk of bias on statistical analysis and confounder management, respectively.Conclusions: Hallux valgus is associated with reduced hallux loading in in‐shoe measurements, but evidence for load redistribution to other forefoot regions is inconclusive. The choice of measurement modality significantly impacts results, highlighting the need for standardized assessment protocols in hallux valgus research.Trial Registration: This study was registered in PROSPERO with the reference number CRD42024574195

Calcaneal fractures are among the most debilitating and costly orthopedic injuries, presenting significant management challenges and poor long-term outcomes. Open reduction and internal fixation (ORIF) remains the gold standard for complex displaced intra-articular fractures, yet high wound complication rates have driven the development of minimally invasive fixation (MIF) techniques. This review systematically evaluates the role of finite element (FE) analysis in calcaneal fracture fixation research, focusing on methodological approaches, model configurations, and reported limitations. A comprehensive search of PubMed, Web of Science, Scopus, and IEEE Xplore identified 23 studies involving 59 computational models derived from 30 subjects. Sanders Type IIB and IIIAB fractures were the most studied due to their clinical relevance and modelling feasibility. FE analyses have progressed from basic stress distribution studies to advanced comparative evaluations and topology-optimized implant designs. Findings indicated that MIF, particularly with supplementary percutaneous screws, offers biomechanical performance comparable to ORIF. However, limitations include reliance on single-subject surrogate models, oversimplified static loading conditions, and limited transparency in model validation. Future research should prioritize multi-patient models, sensitivity analyses of fracture patterns using statistical or generative geometrical approaches, and dynamic loading scenarios (e.g., heel-strike, lunge, fall) to enhance applicability and robustness of findings.