Accurate registration of multi-stained whole-slide images (WSIs) enables the integration of complementary morphological and molecular information, improving both clinical diagnostic and prognostic analysis. It also enables efficient transfer of annotations between consecutive or re-stained slides, significantly reducing annotation time and cost, as well as three-dimensional tissue reconstruction. Despite its importance, WSI registration remains challenging due to variations in slide preparation across stains and complex tissue deformations. Existing methods are often tailored to specific staining modalities and fail to generalize across diverse settings, highlighting the need for a robust and generalizable cell-level multi-stain registration method. In this work, we introduce CORE, a novel coarse-to-fine framework for accurate cell-centric registration across six multimodal WSI datasets, encompassing 30 distinct staining types, including Hematoxylin & Eosin (H&E), periodic acid–Schiff (PAS), multiplex immunohistochemistry (mIHC), multiplex immunofluorescence (mIF) and Cyclic immunofluorescence (Cyc-IF). CORE first performs coarse global registration by extracting tissue masks through prompt-based segmentation to remove artifacts and non-tissue regions, followed by rigid alignment using tissue morphology and a pre-trained feature extractor. The resulting alignment is refined using a shape-aware point-set registration model applied to automatically detected nuclei centroids, enabling fine-grained rigid alignment. Finally, Coherent Point Drift (CPD) is used to estimate a non-linear displacement field for non-rigid cellular alignment. Using nuclei correspondences throughout the pipeline, CORE achieves accurate and robust cell-level alignment across modalities.Experimental results show that CORE consistently outperforms state-of-the-art methods in accuracy, robustness, and generalization across both bright-field and immunofluorescence WSIs.
Cloud computing has become a dominant platform for large-scale data storage and processing; however, outsourcing sensitive information to untrusted cloud servers introduces major privacy and security concerns. Homomorphic-based searchable encryption enables secure search and computation over encrypted data without revealing plaintext contents or query information. This paper presents a systematic survey of homomorphic-based SE schemes for multimedia data including text, audio, and image representations. Cloud computing is now a major data storage and processing platform, but sharing sensitive data with external organizations on semi-trusted cloud servers poses significant privacy and security issues. Homomorphic SE allows for secure search via computation on encrypted data without disclosure of the plaintext data or the information of the query. This paper provides a systematic survey of SE schemes for multimedia data such as text, audio and image representations that are based on homomorphic operations. The study offers a comprehensive analysis of homomorphic encryption techniques employed in cloud-assisted retrieval systems. Special focus is given to the security of search patterns, access pattern protection, probabilistic trapdoor generation, privacy preservation, and formal security models such as IND-CPA, IND-CKA, semantic security, forward as well as backward privacy and ORAM-assisted retrieval. The research also covers emerging architectures like MPC-assisted search, FHE-ORAM hybrid architectures, as well as recent developments in FHE acceleration using GPU-based implementations and hardware-assisted optimizations. This work provides a consolidated reference framework and future research roadmap toward secure, scalable, and privacy preserving homomorphic SE systems for cloud-assisted multimedia domains.
This study examines plasma-induced bubble behaviour and mass-transfer intensification using a plasma bubble reactor. The reactor consists of eight 200 μm holes to generate millimeter-scale bubbles in glycerol water or sodium dodecyl sulphate (SDS) solutions. This small bubble column enables the identification of unique plasma-induced reactions inside the bubbles for subsequent gas–liquid absorption. High-speed imaging shows that plasma activation reduces bubble diameter, narrows size distribution, shortens attachment time, and increases post-detachment path length, effectively extending the reaction window. This behaviour is strongly governed by liquid properties: low-viscosity, high-surface-tension media (water, SDS) support stable plasma bubbles, while higher glycerol concentrations decrease the volumetric mass-transfer coefficient substantially. OES indicates a gliding-arc-like discharge dominated by N2 SPS/FNS, OH, and O emissions, suitable for NO/NO2 chemistry but not for low-temperature pathways such as ozone formation. Plasma oscillations generate ultrasonication-like interfacial renewal, while plasma enhance microturbulence throughout the 150 mL bath, yielding a higher absorption rate of reactive species than that achieved in conventional small bubble columns. Force-balance analysis (buoyancy, drag, Basset, surface tension, gas momentum) clarifies how plasma modifies interfacial stress to bubble size, detachment time to improve gas–liquid absorption. Overall, plasma bubbles operate as mobile microreactors, coupling intensified hydrodynamics with reactive plasma chemistry to achieve superior mass-transfer effect.
Predictive language comprehension allows children to use prior context to anticipate upcoming information before it becomes available in speech. Although young children use sentence meaning to guide real-time comprehension, less is known about when they generate more specific expectations about upcoming words without direct cues. Across two visual-world eye-tracking experiments, we used a grammatical contrast in Spanish as a test case to examine whether children aged 24, 30, and 36 months could use sentence context to guide attention before the expected word was heard. Highly predictable sentence contexts were followed by a silent interval before the final noun, allowing us to distinguish anticipatory looking from later word recognition. In Experiment 1, visual displays included a gender-matched object that shared grammatical gender with the expected noun and also carried transparent surface-form cues. In Experiment 2, those cues were removed, providing a stricter test of whether children could use internally generated expectations before word onset. Results showed anticipatory looks to the expected target across age groups, indicating context-based prediction. However, looks to the gender-matched object followed a more selective developmental pattern. With surface cues, the Gender-matched > Unrelated contrast was reliable at 30 and 36 months and larger at 36 than at 24 months. Without surface cues, it was reliable only at 36 months. These findings suggest that children’s predictions develop from broad expectations about upcoming words toward more specific, linguistically structured expectations between 24 and 36 months. More generally, early predictive processing becomes increasingly precise as language representations become more stable and accessible.
The transition from Child and Adolescent Mental Health Services (CAMHS) to Adult Mental Health Services (AMHS) presents significant challenges, underscoring the need for improved transitional care procedures. Few European countries have implemented national transition-related guidance, despite the potential of clinical guidance to support appropriate care decisions and practices. We conducted a literature review to inform the development of the European Society of Child and Adolescent Psychiatry (ESCAP) transition guidance for clinicians. Following systematic principles, four databases (Medline, Embase, PsychInfo, Web of Science) were searched to identify relevant international research published from January 1, 2019 to April 10, 2025, to build on existing evidence. Titles and abstracts were reviewed by two independent reviewers. We screened 12,595 records and included 149 reports published since 2019. Illness severity was the primary predictor of AMHS transition, with only 20–25
