Hong Kong University of Science and Technology
Hong Kong University of Science and Technology Institutional RepositoryNot a member yet
162821 research outputs found
Sort by
Leakage-free and flexible liquid metal/elastomer composite thermal pads with enhanced electromagnetic interference shielding
No full textLiquid metal/elastomer composites (LMECs) hold great potential for multifunctional applications in electronics and power systems, owing to their unique combination of high electrical conductivity, superior thermal conductivity, and mechanical flexibility. However, critical challenges remain which hinder the practical application, including liquid metal (LM) leakage and limited ability to form interconnected pathways. Herein, we report on leakage-free and multifunctional LMEC thermal pads via controlled oxidation and MXene immobilization. Free LM droplets were confined by oxidized LM clusters generated through accelerated oxidation and further anchored by the Ti3C2Tx MXene nanosheets pre-dispersed in the LM. Thermodynamic analysis revealed a preferential localization of MXene at the LM/elastomer interface, effectively bridging LM domains, promoting continuous pathway formation, and enhancing filler/matrix interfacial compatibility. As a result of the strong interactions among MXene, LM, and its oxide, the leakage of LM was completely suppressed under both static conditions and external stimulation (e.g., 10 MPa compression). The LMEC thermal pads demonstrated a thermal conductivity of 6.34 W/(m·K) which is among the highest in the isotropic polymer/filler system. Moreover, the thermal pads showed superior electromagnetic interference (EMI) shielding effectiveness with a value of 69.8 dB/mm at 0.5 mm thickness. This work provides a facile and effective approach for fabricating leakage-free LMECs as multifunctional thermal management materials for next-generation electronics.</p
From thermal stress to health risk: Energy and housing injustice in Hong Kong's subdivided homes
No full textEnergy insecurity poses a critical threat to the well-being of Hong Kong households, particularly those residing in subdivided units—small spaces created by partitioning larger apartments—especially during periods of extreme weather. This mixed-methods study integrates survey and interview data to examine how energy insecurity and housing precarity intersect and influence self-reported health outcomes. Using an energy justice framework, the analysis highlights persistent energy injustices in Hong Kong. Most respondents reported heavy reliance on air conditioning, resulting in significant cost burdens, while over half experienced mould growth during the wet season; consequently, about one-third reported at least occasional physical health impacts. Ventilation deficiency, including unusable windows and dependence on appliance-based systems, emerged as a strong predictor of ill health. Overall, intensive coping strategies during temperature extremes, coupled with inadequate housing quality, exacerbate domestic thermal stress and health vulnerability. Institutional barriers further compound these injustices, such as landlord-controlled metering, subsidy schemes that exclude renters, and complex application procedures. Although subsidies and respite shelters exist, awareness remains limited. Accordingly, we recommend policy reforms grounded in energy justice principles, including recipient-centred subsidies, metering regulation, and integrated housing and public health interventions
Enhancing LLM-based building data query with chain-of-thought, retrieval-augmented generation, and fine-tuning
No full textTo enhance energy efficiency and occupant satisfaction, modern buildings have collected rich streams of operational and sensor data. Semantic models for buildings, such as the Brick schema expressed in the Resource Description Framework (RDF) and Web Ontology Language (OWL), have standardized the representation of devices, points, and systems. However, non-expert users still faced barriers to accessing such data, because effective use required proficiency in the SPARQL Protocol and RDF Query Language (SPARQL) and navigation of thousands of interconnected nodes and relations. This paper presents BuildingGPT2, a framework that combined large language model fine-tuning, vector-graph retrieval-augmented generation, and chain-of-thought prompting to enable natural-language querying of Brick-based models. The framework was trained on semantic models from 40 real buildings and evaluated in a zero-shot setting on 5 held-out buildings. Using LLaMA 3.1–70B, SPARQL query generation accuracy improved from 49.25 % to 97.11 %, substantially lowering the barrier to interacting with building semantic models.</p
Janus-faced role of oxygen vacancies inα-Ga<sub>2</sub>O<sub>3 </sub>: a catalyst design optimization for photocatalytic water splitting
No full textα-Ga2O3 has garnered increasing attention due to its unique properties in the field of photocatalysis. However, the mechanistic aspects of photocatalytic water-splitting on α-Ga2O3 surfaces containing oxygen vacancies are still not well understood, hindering the development of high-performance catalysts. Here, we present a comprehensive and systematic investigation into the effects of oxygen vacancies on the photocatalytic water splitting in α-Ga2O3 from structural and electronic perspectives, based on functional theory calculations. In the presence of oxygen vacancies, vacancies at particular sites can lead to structural distortions that significantly improve hydrogen evolution reaction activity. In contrast, oxygen evolution reaction activity is inhibited on both the (001) and (012) surfaces with oxygen vacancies, attributed to the enhanced adsorption strength of intermediates resulting from the electrons associated with the vacancies. This study elucidates the Janus-faced role of oxygen vacancies in α-Ga2O3 during photocatalytic water splitting, offering theoretical guidance for developing novel Ga2O3 catalysts with superior catalytic activity.</p
Dissolved organic matter in surface sediments along a river-to-ocean continuum: molecular characteristics and sediment–water exchange dynamics
No full textDissolved organic matter in estuarine sediments (SDOM) plays a crucial role in coastal carbon cycles, mediating the transformation, exchange, and sequestration of carbon between sediments and overlying waters. However, the mechanisms that govern the transformation and fate of SDOM in estuarine environments remain poorly understood. In this study, we conducted a molecular composition analysis of SDOM in the surface sediments of the Changjiang Estuary–East China Sea continuum, using ultrahigh-resolution mass spectrometry. Building on previous characterizations of stable and radioactive carbon isotopes in dissolved organic carbon and sedimentary organic carbon from surface sediments and bottom waters, this study aimed to track the provenance and age of SDOM at the molecular level. Additionally, by comparing with bottom-water DOM, we sought to elucidate the exchange dynamics of SDOM at the sediment-water interface. Our results indicate that SDOM is more biologically labile than bottom-water DOM, characterized by an enrichment in aliphatic, low-molecular-weight, nitrogen-containing compounds that enhance its susceptibility to microbial degradation. Frequent hydrodynamic disturbances and sediment resuspension in the dynamic mobile mud zone of the inner shelf create a hotspot for DOM exchange between sediments and overlying waters. Active vertical mixing facilitates the downward transport of fresh surface-derived DOM, while riverine particles repeatedly adsorb and release DOM, allowing for the concurrent sequestration and remobilization of both marine and terrestrial DOM. These findings highlight the essential role of SDOM as a highly reactive carbon pool that accelerates OM turnover in large river-dominated ocean margins. Furthermore, this research elucidates the rapid biogeochemical mechanisms driving carbon cycling in estuaries and reveals how SDOM either transforms or preserves OM in sediments, with significant implications for refining global coastal carbon budgets.</p
IAENet: an importance-aware ensemble model for 3D point cloud-based anomaly detection
No full textSurface anomaly detection is pivotal for ensuring product quality in industrial manufacturing. While 2D image-based methods have achieved remarkable success, 3D point cloud-based detection remains underexplored despite its richer geometric cues. We argue that the key bottleneck is the absence of powerful pretrained foundation backbones in 3D comparable to those in 2D. To bridge this gap, we propose Importance-Aware Ensemble Network (IAENet), an ensemble framework that synergizes 2D pretrained expert with 3D expert models. However, naively fusing predictions from disparate sources is non-trivial: existing strategies can be affected by a poorly performing modality and thus degrade overall accuracy. To address this challenge, We introduce a novel Importance-Aware Fusion (IAF) module that dynamically assesses the contribution of each source and reweights their anomaly scores. Furthermore, we devise critical loss functions that explicitly guide the optimization of IAF, enabling it to combine the collective knowledge of the source experts but also preserve their unique strengths, thereby enhancing the overall performance of anomaly detection. Extensive experiments show that IAENet achieves a new state-of-the-art for point-level localization and ranks second at object-level on MVTec 3D-AD dataset. On the Eyecandies dataset, it achieves the best performance in both levels. Additionally, it substantially reduces false positive rates, underscoring its practical value for industrial deployment.</p
An optimization study considering MNP temporal evolution improves therapeutic efficacy in hyperthermia treatment
No full textThis work investigates optimal strategies for achieving the most effective tumor ablation outcomes in magnetic hyperthermia by incorporating the thermal exposure time, magnetic nanoparticle (MNP) dose, injection sites, and waiting time before alternating magnetic field (AMF) application. The optimization framework highlights the significance of thermal exposure time, as this treatment duration substantially influences the temporary distribution of the heat source, i.e., MNPs, and consequently affects the thermal dose for efficacy evaluation. Multi-site MNP injections are involved in both circular and elliptical tumor configurations, and a transversal blood vessel introduces asymmetric cooling effects. This optimization framework can achieve efficient convergence, demonstrating its effectiveness in identifying the optimal strategy. Without the influence of the blood vessel, optimal injections exhibit a centrosymmetric distribution in the circular tumor model; comparatively, a linear distribution along the major axis with approximately halved treatment duration is observed in the elliptical model. When the blood vessel is nearby, the notable asymmetric cooling effects complicate treatment, where the random search method is more effective. Increasing the tumor–vessel distance enhances tumor ablation, reduces MNP dosage and treatment time, and decreases the average injection site deviation; however, the impact becomes marginal at larger distances. This optimization study facilitates the efficacy of practical treatment.</p
Harnessing the power of aggregation-induced emission luminogens: From bright probes to theranostic platforms in biomedicine
No full textAggregation-induced luminescence (AIE) is a unique photophysical phenomenon where certain molecules emit almost no light in dilute solutions but significantly increase their luminescence when aggregated or in the solid state. AIE luminogens (AIEgens), as a breakthrough in luminescence research, have shown great application potential in the biomedical field due to their unique molecular mechanisms and excellent optical properties. Compared with traditional fluorescent materials, AIEgens have significant characteristics and advantages: High brightness during polymerization (generating strong fluorescence), excellent light stability (ideal long-term, high-intensity optical imaging and treatment), concentration-dependent response (detecting dynamic changes in the biological environment, such as pH, viscosity and enzyme activity), good biocompatibility (low toxicity and high biodegradability), and multi-functionality (binding imaging). Comprehensive diagnosis and treatment are carried out through treatment and sensing. Based on these advantages, this article systematically discusses the application and progress of AIEgens in biomedical imaging (including cancer cells, microorganisms, and in vivo imaging), microbial detection and treatment (including bacteria, fungi, and viruses), tumor diagnosis and treatment, etc., and explores the future development direction and prospects.</p
Comparative validation of surgical phase recognition, instrument keypoint estimation, and instrument instance segmentation in endoscopy: Results of the PhaKIR 2024 challenge
No full textReliable recognition and localization of surgical instruments in endoscopic video recordings are foundational for a wide range of applications in computer- and robot-assisted minimally invasive surgery (RAMIS), including surgical training, skill assessment, and autonomous assistance. However, robust performance under real-world conditions remains a significant challenge. Incorporating surgical context – such as the current procedural phase – has emerged as a promising strategy to improve robustness and interpretability. To address these challenges, we organized the Surgical Procedure Phase, Keypoint, and Instrument Recognition (PhaKIR) sub-challenge as part of the Endoscopic Vision (EndoVis) challenge at MICCAI 2024. We introduced a novel, multi-center dataset comprising thirteen full-length laparoscopic cholecystectomy videos collected from three distinct medical institutions, with unified annotations for three interrelated tasks: surgical phase recognition, instrument keypoint estimation, and instrument instance segmentation. Unlike existing datasets, ours enables joint investigation of instrument localization and procedural context within the same data while supporting the integration of temporal information across entire procedures. We report results and findings in accordance with the BIAS guidelines for biomedical image analysis challenges. The PhaKIR sub-challenge advances the field by providing a unique benchmark for developing temporally aware, context-driven methods in RAMIS and offers a high-quality resource to support future research in surgical scene understanding.</p
Effects of intravenous ketamine on posttraumatic stress disorder (PTSD): a systematic review
No full textIntroduction: Posttraumatic stress disorder (PTSD) is a mental disorder resulting from exposure to traumatic events. Evidence suggests that ketamine may be efficacious in treating PTSD, however, ketamine's mechanisms in treating PTSD remain unclear. Herein, this review aims to evaluate the clinical outcomes of ketamine treatment in persons with PTSD and investigate the possible neurobiological mechanisms underlying ketamine's therapeutic effect in PTSD. Methods: A systematic search was conducted on PubMed and OVID (MEDLINE, Embase, PsychINFO) from inception until September 2025. Randomized controlled trials reporting on the effects of intravenous ketamine to treat PTSD were included. Results: Seven studies with a total of 323 participants were included in this review. Ketamine administration meaningfully improved PTSD symptoms in two trials as evidenced by significant improvement on the Clinician-Administered PTSD Scale for DSM-5 (CAPS-5) and the Impact of Event Scale-Revised (IES-R) compared to control/placebo. Multi-infusion administration schedules achieved greater clinical outcomes when compared to single-dose administration schedules. Preliminary evidence suggests that repeated lower doses (0.2mg/kg) of ketamine were more efficacious in sustaining treatment effects than standard doses (0.5mg/kg). For persons receiving ketamine, an association was observed between top-down inhibition of the amygdala originating in the ventromedial prefrontal cortex (vmPFC) and symptom improvement. Conclusion: Our results suggest that intravenous ketamine may be efficacious in the treatment of PTSD. Subsequent studies should attempt to evaluate the additive effect of combining ketamine with psychotherapeutic interventions as well as determining mechanistic pathways mediating symptom relief in persons with PTSD.</p