To describe the incidence density of intussusception and its distribution characteristics across different ages, genders, and time periods among resident children aged 0-3 years in Jiangsu Province from 2018 to 2023, providing a scientific basis for health administrative departments in the region to formulate prevention and control strategies and measures for intussusception in children aged 0-3 years. A retrospective cohort study was conducted using data from the Jiangsu Regional Health Information Platform (RHIP). Resident children aged 0-3 years born between January 1, 2018, and December 31, 2023, were included. Suspected cases were identified by retrieving the International Statistical Classification of Diseases and Related Health Problems, 10 revision (ICD-10) code "K56.1" and semantically related keywords for "intussusception" from the original text using a regular expression matching algorithm. An incident case was defined as the first occurrence of the disease during the study period. Follow-up person-years were calculated using the exact person-time method. The Poisson distribution was applied to estimate the overall and subgroup incidence densities and their 95% confidence intervals (CI). Poisson regression models were constructed to calculate incidence rate ratio (IRR) and examine the effects of age (in months) and seasonal differences on incidence density. A birth cohort comprising 2 252 691 children aged 0-3 years was established, accumulating a total follow-up of 5 316 389.55 person-years. During the study period, 2 650 incident cases of intussusception were identified, including 1 600 boys (60.4%) and 1 050 girls (39.6%). The overall incidence density was 49.8 (95%CI: 47.9-51.7) per 100 000 person-years. The incidence density was 57.8 (95%CI: 54.9-60.6) per 100 000 person-years for boys and 41.2 (95%CI: 38.7-43.7) per 100 000 person-years for girls, with a statistically significant difference between the sexes. Seasonal variations were observed, with peaks occurring in summer and winter. The incidence density exhibited a unimodal pattern, peaking at 8 months (98.5 per 100 000 person-years). The median age of onset was 19.2 months [interquartile range (IQR): 11.4-26.6 months]. This large retrospective cohort study based on the Jiangsu RHIP clarified the epidemiological characteristics of intussusception among local children aged 0-3 years from 2018 to 2023. Boys and children aged 8 months to 3 years were identified as high-risk populations for intussusception, with peak incidence occurring in summer and winter. These findings provide evidence-based support for health authorities to formulate targeted strategies for early surveillance, prevention, and healthcare resource allocation.

The theoretical connotation and operation of Miu Ci and Ju Ci remain highly controversial at present.Traditional viewpoints mostly hold that Miu Ci targets collaterals, while Ju Ci works on meridians. Based on the analysis of the original texts and textual exegesis from Huangdi Neijing (The Yellow Emperor's Inner Classic), it is believed that both Miu Ci and Ju Ci are the techniques to puncture acupoints on the contralateral side of the body, and these two techniques were formed during the medical practice of blood-letting at early stage of stone acupuncture and developed under the theoretical framework of qi regulation of filiform needle. At the time when Suwen·Miuci Lun (Chapter 63 of Huangdi Neijing)was completed, "Miu Ci needling collaterals" and "Ju Ci needling meridians" were factitiously distinguished. Analyzing the essence of Miu Ci and Ju Ci is conducive to restoring their true representations in Huangdi Neijing, thereby liberating them from the constraints of traditional thinking and promoting their significant value of clinical application.

Early diagnosis is crucial in improving oral cancer outcomes. Patient education materials support timely recognition and management. However, these resources are often written above recommended reading levels, beyond patients' health literacy and limiting accessibility. To assess the readability of available patient information on oral cancer by the NHS, to evaluate three large language models (LLMs; ChatGPT, Claude and Gemini) in simplifying texts while preserving their content, and to propose an improved leaflet based on UK materials, expert review and LLM adjustment to match average UK reading levels. Materials were collected from NHS-affiliated websites. Original and LLM-simplified texts were assessed using validated readability tools (FRES, FKGL, GFI, CLI and SMOG). Content fidelity was assessed using character 3-5-g cosine, sentence-content retention and latent semantic analysis (LSA). An expert review was applied to the proposed leaflet. LLM-revisions significantly improved readability across all five indices (p < 0.0001). Mean FRES of original texts was 66.4 ± 7.7, while Claude (81.6 ± 6.2) was the only model to surpass the 80 benchmark. Semantic similarity to source text remained high (LSA means 0.97 ± 0.04, 0.94 ± 0.09 and 0.96 ± 0.08; character 3-5-g cosine 0.85 ± 0.05, 0.80 ± 0.08 and 0.82 ± 0.08 for respective models). Baseline readability of the proposed leaflet was comparable to NHS materials (FRES 65.7); Claude increased this to 81.2. LLM-based simplification enhanced readability while preserving content fidelity. This approach can help enhance accessibility, particularly for populations disproportionately affected by oral cancer. With human oversight, it could be adopted at the policy level to standardise patient education and reduce health literacy disparities.

Clinical data governance is the cornerstone of reliable intelligent healthcare systems. However, real-world clinical records frequently suffer from complex data quality issues that demand high semantic fidelity and processing efficiency to resolve. Existing section identification and fragmented standardization methods either fail to address these intricate anomalies or inadvertently sacrifice semantic integrity. Meanwhile, directly deploying Large Language Models (LLMs) for this task as free-form text generators introduces hallucinations and computational bottlenecks. To bridge these gaps, we propose GovernAgent, an LLM-driven framework that overcomes these limitations through two core designs. First, inspired by the intrinsic structure of clinical records, our approach introduces a hierarchical governance mechanism. By employing cascading Note- and Section-Level agents, it constrains the governance space in a top-down manner, systematically disentangling these anomalies into resolvable, multi-level quality issues. Second, the framework employs a Constrained Action Planning mechanism. By restricting the LLM to a hybrid "Copy-Generate" action space rather than free-text generation, it maximizes original text reuse, thereby mitigating hallucinations, guaranteeing medical provenance, and ensuring high efficiency. Evaluations on real-world hospital datasets demonstrate that GovernAgent improves governance accuracy and efficiency, minimizes hallucinations, demonstrates high practical adaptability, and empowers downstream clinical applications. Code: https://github.com/kaiyinzhou/GovernAgent.

Large language models (LLMs) are increasingly explored for clinical use. However, the extent to which such models can reliably support physicians in reporting, staging, and the assessment of classification remains an active area of research. This study aimed to evaluate and compare multiple LLMs for automated PROMISE V2 classification for prostate cancer. A total of 126 unambiguous German-language PSMA PET/CT text reports were retrospectively analyzed, with reference standards established by expert consensus based on image interpretation and the original report text. Five LLMs (GPT-5.4, DeepSeek-V3.2, Claude Sonnet 4.6, Gemini 3 Flash and Grok 4) were assessed using two English-language prompting strategies of varying complexity. Agreement with the reference standard served as the primary endpoint. Performance varied in the short-prompt setting (36.5-79.4%) but improved consistently with the long prompt (74.6-86.5%), with Gemini 3 Flash achieving the highest agreement. Across PROMISE V2 subcategories, agreement rates were high (miT: 81.0-92.1%, miN: 92.9-96.0%, miM: 92.9-95.2%), despite inter-model differences. In conclusion, contemporary LLMs demonstrate promising performance in deriving PROMISE V2 scores from unambiguous original report texts, particularly when guided by detailed prompts.

ABSTRACT We have refined the original text: Vehicle trajectory prediction (VTP) is a pivotal technique for autonomous driving, enabling safe and efficient navigation in complex traffic scenarios. Although recent deep learning-based methods have advanced VTP by modeling spatiotemporal interactions, persistent challenges remain in capturing inter-vehicle and vehicle-lane dependencies without relying on high-definition (HD) maps, limiting the robustness and generalization of existing models. To address these limitations, we propose the Spatiotemporal Interaction Aware Network (SIA-Net), a novel and scalable framework for robust VTP. SIA-Net comprises two core modules: the Temporal Interaction-aware Module (TIAM), which employs bidirectional attention and multiscale temporal convolutions to capture dynamic vehicle interactions and motion patterns; and the Spatial Interaction-aware Module (SIAM), which integrates lane-aware attention and graph convolutional networks to model spatial topological relationships. Experiments on the NGSIM and HighD datasets demonstrate that SIA-Net outperforms state-of-the-art baselines, achieving a 32.1% reduction in 1-second RMSE while offering strong generalizability across unmapped and complex traffic environments.

Medical Entity Recognition (MedER) systems are needed to enhance the use and accessibility of Natural Language Processing (NLP) methods in the medical field. Since medical entity recognition in Bangla is a relatively new field, no such datasets are currently available in any repository. Unlike AI-generated data, which may contain biases or errors from auto- mated algorithms, the Bangla-MedER dataset is a manually curated resource for multi-type medical entity recognition in Bangla-language drug-indication text. A total of 2980 records were collected from publicly available medicine-related websites, pharmaceutical articles, and other sources of drug information. Each record contains the original Bangla medical text, along with expert-verified annotations across six entity types: medicine/chemical name, organ, disease, hormone, pharmacological class, and common medical terms. The raw transcribed text is provided to support reproducible research. All annotations were performed manually under the guidance of a certified medical expert, and proprietary brand names or personally identifiable information were removed to ensure privacy. The Bangla-MedER dataset enables a variety of applications, including medical named entity recognition in Bengali medical text, biomedical information retrieval, and clinical decision-support systems in a low-resource language environment. The complete raw dataset, along with documentation, is publicly available, offering a benchmark for medical entity recognition, healthcare NLP, biomedical informatics, and another related research.

The widespread usage of generative AI (genAI) challenges traditional notions of authorship, raising questions about how readers perceive and attribute textual origin. As this sociolinguistic issue is recognized by people, the study presented in this paper was initiated from a collaboration between professional researchers and citizens engaging in encounters discussing language-related social issues. Following a citizen sociolinguistic approach, we engaged 576 participants in a questionnaire study assessing their ability to distinguish AI-generated (GPT-3.5) from human-authored narrative texts in Hungarian. The results of the study show that respondents were significantly more successful than chance (66%) in identifying the author of the text regardless of their age, gender, occupation, or self-reported AI skills. However, the performance score was dependent on the author of the text, as AI-authored texts were identified more successfully than those written by humans. We also found that respondents preferred texts they believed to be human-authored over those they assumed to be generated by AI, demonstrating an anthropocentric bias. Beyond these results, the study highlights how the growing invisibilization of genAI technologies challenges long-standing ideas about authorship, even in the case of narrative texts. • Respondents identified the author with greater accuracy than random chance, but did not reach perfect accuracy. • Performance scores were not affected by respondents' occupation, gender, age, self-reported AI skills and AI literacy. • Performance scores were significantly higher for AI-generated texts compared to those authored by humans. • Respondents preferred texts they believed were written by a human over texts assumed to be written by AI.

The research on translation of Emily Dickinson’s poetry has gradually attracted attention from scholars at home and abroad in recent years. But so far, comparison and evaluation of different translations are mostly based on subjective impression and less on objective empirical study. A statistical analysis of formal factors such as stanza, line, punctuation, and vocabulary use in 200 Poems of Emily Dickinson translated by Wang Hongyin and I Know He Exists translated by Tu An is performed with the help of programs written in Python language to objectively compare and summarize translation style of the two translations, and to reveal some of the features of current Chinese translation of Emily Dickinson’s poetry. It is found that the two translations deviate slightly from the original in terms of stanza and line number. On the whole, Tu’s translation deviates less from original text, its use of translation language and punctuation are very close to original text, with only occasional adjustment to original line and word sequences, its overall style is plain and smooth. Wang’s translation deviates relatively greater from original text. Lots of alterations are made to original punctuation, its diction incorporates more translator’s own understanding of original poems, many modal words, onomatopoeia and er-hua words with characteristics of northern dialect, etc. are added, and omits original sentence elements frequently, all of which resulting in a translation with distinctive translator’s personality and diverse language styles.

This paper takes the relevance theory as the research perspective, selects the 1957 English translation of Shapiro 's The Outlaws of the Marsh as the core object, conducts a specific analysis using the explicit reasoning model, sorts out the relevant strategies of cognitive context reconstruction by comparing the original text and the translation, and focuses on how translators handle semantic fidelity and communicative validity in cross-cultural communication. A closer look reveals that Shapiro's translation is highly in line with the cognitive environment of the target readers. It not only retains the narrative features of the original work but also does a good job in the metonymy and reconstruction of cultural images. This provides a very valuable case for the cognitive pragmatics study of the translation of Chinese classical literature into foreign languages.This paper takes the relevance theory as the basic perspective of research, selects the English translation of The Outlaws of the Marsh published by Shapiro in 1957 as the core research text, conducts a comparative analysis of the original text and the translation, sorts out various specific strategies of the translator to complete the reconstruction of cognitive context, and also focuses on analyzing the translator in the practice of cross-cultural communication. How can a balance be achieved between semantic fidelity and communicative effectiveness? After research, it can be concluded that in the process of translation, Shapiro fully considered the cognitive environment characteristics of the target readers. He not only completely retained the original narrative features of the original work but also made proper metonymy and reconstruction of the cultural images in the text. Such translation practice has also provided support for the cognitive pragmatics research in the direction of translating Chinese classical literature into foreign languages. It left behind typical cases of reference value.

Clinical text cannot be shared freely due to data protection regulations. This paper introduces SURROGATOR, an open-source framework that replaces personal identifiable information with fictitious, high-fidelity synthetic surrogates to maintain text utility for human and artificial intelligence. We focus on the automated replacement of names, dates, and locations. Evaluation on a synthetic clinical corpus demonstrated high data utility, with named entity recognition performance remaining en par with the original text (F1-score of 0.70 vs. 0.73). Furthermore, a re-identification attack using a large language model resulted in an accuracy of 50.8%, which is equivalent to random chance. We conclude that SURROGATOR effectively balances patient privacy and data utility, providing a robust solution for the secure sharing of clinical documents in medical research.

Federated Learning (FL) allows institutions to train shared models without exchanging raw data, making it a promising approach for healthcare applications that involve sensitive electronic health records (EHRs). However, despite this distributed design, the gradients exchanged during training can still reveal private information. In this study, we analyze how vulnerable transformer-based language models are to gradient inversion attacks, focusing on the Decepticons method, which can reconstruct original training text from shared gradients. We simulate a cross-silo FL setup with three types of French clinical reports (genetic, anesthesia, and birth records) to evaluate how batch size and sequence length affect reconstruction quality. Our experiments show that a malicious server can recover clinical text with high accuracy: token-level recovery exceeded 95% when training with batch size 1 and remained above 60% for sequences of up to 512 tokens. Reconstructed examples contained identifying elements (names, dates, genetic markers), revealing serious privacy risks for real-world use. These results emphasize that FL alone is insufficient for sensitive clinical text and that privacy-preserving defenses must be integrated before real-world deployment.

ABSTRACT This article examines the martyrology composed at St Augustine’s Abbey in the late eleventh/early twelfth century, coinciding with Goscelin’s time there as precentor. By recording the variants present in the martyrology text (compared to the original martyrology text by Usuard) connections can be made between Goscelin’s scholarly output and the feast day additions. Furthermore, through close textual analysis of the added entries, patterns can be identified that correspond with Goscelin’s distinctive writing style. By identifying these links between the martyrology and Goscelin, more light can be shed on his time at St Augustine’s as precentor, and how his hagiographical output may have influenced the liturgy there. The article also discusses the institutional differences between St Augustine’s and Christ Church after the Norman Conquest. The additions suggest that through the text, St Augustine’s strongly asserted its own historical traditions and therefore bears witness to its independence from Christ Church.

Reading English biomedical literature is a core competency for medical students, yet many rely heavily on full-text translation tools. Although translation appears to reduce language barriers, its impact on cognitive load and durable learning remains unclear. This study compared three reading modalities to evaluate their effects on cognitive load, immediate comprehension, and delayed vocabulary retention in medical students. In this three-arm randomized controlled trial, 160 Chinese medical students were randomized to one of three reading conditions: Original Text Group (OTG), AI-assisted Bilingual Group (ABG), or Full Translation Group (FTG). Immediate post-reading outcomes included intrinsic, extraneous, and germane cognitive load, as well as comprehension performance. One-week delayed productive vocabulary retention was assessed using a prespecified scoring rubric. Group differences were analyzed using analysis of covariance (ANCOVA) adjusted for academic year and CET-6 level. A total of 148 participants completed the immediate post-reading assessment, and 148 completed the one-week follow-up according to the available dataset. Compared with the OTG and ABG, the FTG showed significantly higher extraneous cognitive load (both p < 0.001). Immediate comprehension scores were high across groups and did not differ significantly (p = 0.157). Delayed productive vocabulary retention differed significantly among groups, with the highest mean score in the OTG (6.38 ± 2.13), followed by the ABG (5.58 ± 1.89) and FTG (4.90 ± 2.00) (overall p < 0.001). Post-hoc analyses showed that the OTG outperformed both the ABG (p = 0.011) and FTG (p < 0.001). Among Chinese medical students reading an English biomedical article, full translation was associated with higher perceived extraneous cognitive load and weaker one-week retention of methodological English vocabulary than direct reading of the original English text. A bilingual parallel-text format showed intermediate performance. These findings suggest that translation may be more appropriately used as a selective scaffold rather than a default substitute when the educational goal includes durable acquisition of English medical terminology.

Dataset of authorship attribution from short-text multi topic representative in Bahasa Indonesia

ABSTRACT Despite extensive research on explicitation in translation, the semantic factors influencing explicitation still remain underexplored. This study adopts a corpus-based approach to examine the shifts of causal relations in the explicitation of causal connectives in the English translation of Chinese ecological discourse. The analyses of the parallel corpus reveal a preference for representing causal relations as non-volitional (independent of human intention) in English translations compared with original Chinese texts. Furthermore, the results of the comparable corpus show a significantly higher frequency of non-volitional relations and a significantly lower frequency of epistemic relations (involving human inference) in English translations compared with original English texts. Taken together, these findings suggest a tendency toward a more objective representation of causality in translation, which extends previous research on the role of semantic factors in explicitation.

Abstract This study examines explicature construction in the English translation of the Qur’an, focusing on parenthetical insertions in the Hilali-Khan version. These additions make implicit meanings explicit, guiding readers through culturally and theologically dense material. Applying Huang’s (2014) framework, the analysis categorizes these insertions into five types: disambiguation, reference resolution, saturation, free enrichment, and ad hoc concept construction. Each type reflects a specific inferential process that bridges contextual gaps and enhances interpretability for non-Arabic readers. The findings reveal that explicatures, particularly free enrichment and disambiguation, play a central role in conveying the intended meaning of the original text while maintaining doctrinal and linguistic fidelity. The study also highlights how translators selectively employ these strategies based on contextual, theological, and pragmatic factors. By situating these findings within relevance theory, the research underscores the importance of explicature in sacred text translation, where linguistic precision must be balanced with interpretive accessibility. This work contributes to translation studies and pragmatics by offering a systematic account of how parenthetical explication functions in one of the most influential translations of a religious text.

The article presents a translation-oriented analysis of memoir and autobiographical texts, which are characterized as polygenre. The relevance of the topic is driven by the growing readership interest in memoir and autobiographical texts, as well as by the fact that for an audience not proficient in the original language, translation becomes the only source of information about events described by the participant himself in another language. Thus, unlike biographies, which can be written by domestic researchers, memoirs and autobiographies are created by participants themselves. The article aims to identify how the polygenre nature of these texts (the inclusion of genre-different microtexts within the macrotext) and the author’s mentality influence translation decisions, and to propose a toolkit for analyzing such influence based on cognitive linguistics. The methodological basis is Ch. Fillmore and E.G. Beliaevskaya frame analysis, supplemented by G. Lakoff and M. Johnson theory of cognitive metaphor, and comparative analysis of original texts and their translations. The scientific novelty lies in applying frame analysis to the translation of polygenre memoir and autobiographical texts. The article argues that the boundaries of genre-different microtexts within the memoir and autobiographical macrotext are determined by a change of protagonist. It is shown that the interaction between frames and polygenre is two-fold: a frame can be determined by the author’s professional experience (even if the events described are not directly related to his own life) or by the microtext genre. In the first case, cognitive metaphors may be used, possibly lost in translation. In the second case, the microtext genre activates a specific frame that influences the translation. Frame analysis facilitates verification of translation decisions, including at the editing stage. The results of the study can be applied in translation practice and pedagogy, as well as in editing translations of memoir and autobiographical texts.

The article examines the musical as a polycode genre with a multimedia embodiment, possessing an increased potential to influence on the audience. The author associate the properties of the musical with the strategy of theatricalization – one of the hermeneutic interpretative strategies in modern text analysis, which is also actively used in the practice of teaching Russian as a foreign language (RFL). The educational potential of the genre is presented through the musical "Anna Karenina" based on the eponymous novel by Leo Tolstoy. Using the example of this work, the possibilities of recoding the main meanings of the original text in the verbal, musical and visual codes of the musical are explored. The article demonstrates the potential for using musicals as educational multimedia texts, facilitating a deeper understanding of the source text due to semiotic cross-codes that connect the staging with the original classical source. A conclusion is drawn about the didactic potential of the musical's libretto.

Eco-translatology, with adaptation and selection as its core, offers a holistic view of translation from an ecological perspective. It discusses how to achieve adaptive selection in target language translation from three dimensions: linguistic, cultural, and communicative. Drawing on eco-translatology, this paper takes Ren Aihong’s Chinese translation of Educated: A Memoir as an example to explore English-Chinese translation of autobiographical literary texts. The translation of autobiographical texts is a process of multi-dimensional selection and adaptation, which should not only remain faithful to the original text but also reflect cultural diversity, thereby gaining recognition and acceptance from target language readers. This paper analyzes how the translator makes adaptive choices through ecological adaptation and achieves optimal adaptive transformation in the target language ecology from the three dimensions of language, culture, and communication.