Objective: Research on eye lens dosimetry for radiation workers has increased after the 2012 ICRP118 update on eye lens dose limits. However, corneal dosimetry remains underexplored due to historical focus and measurement challenges. This study uses a high-resolution digital eye phantom in Monte Carlo simulations to estimate corneal and lens doses for nuclear medicine staff, with and without lead glasses. Method: The Monte Carlo code GATE (version 9.0) based on GEANT4 (version 10.6) was used to estimate and compare doses in a digital eye phantom, accounting for primary and scattered radiation from common radionuclides (F18, I131, Tc99m) with varying lead glass shielding (0 to 0.75 mm). Results: Across all radionuclides, the dose to the cornea was consistently higher than the dose to the lens. Notably, the ratio of corneal to lens dose increased with thicker lead glasses, indicating a greater dose reduction to the lens compared to the cornea. Conclusion: The findings show that corneal doses from all studied radionuclides exceeded lens doses. Although increasing lead glass thickness reduced both, the reduction was more significant for the lens, raising the cornea-to-lens dose ratio. This trend suggests that while thicker lead glasses enhance lens protection, their practicality may be limited due to diminishing returns and potential discomfort. Keywords: Corneal Dosimetry, Lens Dosimetry, Monte Carlo, GATE, Nuclear Medicine, Simulation
Traditional Chinese Medicine (TCM) is a holistic medical system with millennia of accumulated clinical experience, playing a vital role in global healthcare-particularly across East Asia. However, the implicit reasoning, diverse textual forms, and lack of standardization in TCM pose major challenges for computational modeling and evaluation. Large Language Models (LLMs) have demonstrated remarkable potential in processing natural language across diverse domains, including general medicine. Yet, their systematic evaluation in the TCM domain remains underdeveloped. Existing benchmarks either focus narrowly on factual question answering or lack domain-specific tasks and clinical realism. To fill this gap, we introduce MTCMB-a Multi-Task Benchmark for Evaluating LLMs on TCM Knowledge, Reasoning, and Safety. Developed in collaboration with certified TCM experts, MTCMB comprises 12 sub-datasets spanning five major categories: knowledge QA, language understanding, diagnostic reasoning, prescription generation, and safety evaluation. The benchmark integrates real-world case records, national licensing exams, and classical texts, providing an authentic and comprehensive testbed for TCM-capable models. Preliminary results indicate that current LLMs perform well on foundational knowledge but fall short in clinical reasoning, prescription planning, and safety compliance. These findings highlight the urgent need for domain-aligned benchmarks like MTCMB to guide the development of more competent and trustworthy medical AI systems. All datasets, code, and evaluation tools are publicly available at: https://github.com/Wayyuanyuan/MTCMB.
Sanchayan Santra, Vishal Chudasama, Pankaj Wasnik
et al.
Precise Event Spotting (PES) aims to identify events and their class from long, untrimmed videos, particularly in sports. The main objective of PES is to detect the event at the exact moment it occurs. Existing methods mainly rely on features from a large pre-trained network, which may not be ideal for the task. Furthermore, these methods overlook the issue of imbalanced event class distribution present in the data, negatively impacting performance in challenging scenarios. This paper demonstrates that an appropriately designed network, trained end-to-end, can outperform state-of-the-art (SOTA) methods. Particularly, we propose a network with a convolutional spatial-temporal feature extractor enhanced with our proposed Adaptive Spatio-Temporal Refinement Module (ASTRM) and a long-range temporal module. The ASTRM enhances the features with spatio-temporal information. Meanwhile, the long-range temporal module helps extract global context from the data by modeling long-range dependencies. To address the class imbalance issue, we introduce the Soft Instance Contrastive (SoftIC) loss that promotes feature compactness and class separation. Extensive experiments show that the proposed method is efficient and outperforms the SOTA methods, specifically in more challenging settings.
Large Language Models (LLMs) have demonstrated significant potential in medicine, with many studies adapting them through continued pre-training or fine-tuning on medical data to enhance domain-specific accuracy and safety. However, a key open question remains: to what extent do LLMs memorize medical training data. Memorization can be beneficial when it enables LLMs to retain valuable medical knowledge during domain adaptation. Yet, it also raises concerns. LLMs may inadvertently reproduce sensitive clinical content (e.g., patient-specific details), and excessive memorization may reduce model generalizability, increasing risks of misdiagnosis and making unwarranted recommendations. These risks are further amplified by the generative nature of LLMs, which can not only surface memorized content but also produce overconfident, misleading outputs that may hinder clinical adoption. In this work, we present a study on memorization of LLMs in medicine, assessing its prevalence (how frequently it occurs), characteristics (what is memorized), volume (how much content is memorized), and potential downstream impacts (how memorization may affect medical applications). We systematically analyze common adaptation scenarios: (1) continued pretraining on medical corpora, (2) fine-tuning on standard medical benchmarks, and (3) fine-tuning on real-world clinical data, including over 13,000 unique inpatient records from Yale New Haven Health System. The results demonstrate that memorization is prevalent across all adaptation scenarios and significantly higher than that reported in the general domain. Moreover, memorization has distinct characteristics during continued pre-training and fine-tuning, and it is persistent: up to 87% of content memorized during continued pre-training remains after fine-tuning on new medical tasks.
Zheng Hui, Yijiang River Dong, Ehsan Shareghi
et al.
As large language models (LLMs) are increasingly deployed in high-risk domains such as law, finance, and medicine, systematically evaluating their domain-specific safety and compliance becomes critical. While prior work has largely focused on improving LLM performance in these domains, it has often neglected the evaluation of domain-specific safety risks. To bridge this gap, we first define domain-specific safety principles for LLMs based on the AMA Principles of Medical Ethics, the ABA Model Rules of Professional Conduct, and the CFA Institute Code of Ethics. Building on this foundation, we introduce Trident-Bench, a benchmark specifically targeting LLM safety in the legal, financial, and medical domains. We evaluated 19 general-purpose and domain-specialized models on Trident-Bench and show that it effectively reveals key safety gaps -- strong generalist models (e.g., GPT, Gemini) can meet basic expectations, whereas domain-specialized models often struggle with subtle ethical nuances. This highlights an urgent need for finer-grained domain-specific safety improvements. By introducing Trident-Bench, our work provides one of the first systematic resources for studying LLM safety in law and finance, and lays the groundwork for future research aimed at reducing the safety risks of deploying LLMs in professionally regulated fields. Code and benchmark will be released at: https://github.com/zackhuiiiii/TRIDENT
BackgroundLower limb strength and power are critical for martial arts athletes to perform complex movements such as aerial outward swings. French Contrast Training (FCT), which integrates heavy compound exercises, plyometrics, and assisted plyometric movements into a single session, has been proposed to elicit superior neuromuscular adaptations. However, the effectiveness of FCT in elite martial arts athletes remains unclear.ObjectiveThis study aimed to investigate the effects of an 8-week FCT program on lower limb strength and power in elite martial arts athletes.MethodsIn total, 24 elite male martial arts athletes were randomly assigned to the FCT group (n = 12) or the control group (n = 12). Both groups completed an 8-week (twice a week) training program. The FCT protocol included the following four sequential exercises per session: 85% one-repetition maximum (1RM) back squats, countermovement jumps (CMJs), 30% 1RM squat jumps (SJs), and band-assisted jumps. The control group performed traditional resistance training for the same muscle groups. The pre- and post-intervention assessments included isometric mid-thigh pulls [IMTPs; maximal force output (MFO), relative MFO, and rate of force development (RFD)], CMJs, SJs with jump height, peak power output (PPO), and mean power output (MPO), elasticity index (EI), and dynamic strength index (DSI). All the data were analyzed using the linear mixed model. Effect sizes were calculated using Cohen's d. Statistical significance was set at p < 0.05.ResultsThe FCT group showed significantly greater improvements than the control group in IMTP (MFO: p < 0.001, d = 1.66; relative MFO: p = 0.001, d = 1.51; RFD: p = 0.001, d = 1.52), CMJ (jump height: p = 0.011, d = 1.14; PPO: p < 0.001, d = 1.61; MPO: p = 0.013, d = 1.11), SJ (jump height: p = 0.019, d = 1.03; PPO: p = 0.043, d = 0.88; MPO: p < 0.001, d = 1.63), EI (p = 0.521, d = −0.27), and DSI (p < 0.001, d = 2.07). No adverse events were reported.ConclusionsThis study provides preliminary evidence that French Contrast Training effectively enhances lower body strength and power in elite martial arts athletes.
Does early specialization facilitate later athletic excellence, or is early diversification better? This is a longstanding theoretical controversy in sports science and medicine. Evidence from studies investigating athletes’ starting age, childhood/adolescent progress, and amounts of coach-led practice and peer-led play in their main sport and in other sports has been mixed. Each participation variable was positively correlated with performance in some studies but uncorrelated or negatively correlated with performance in others. However, samples were heterogeneous in age, sports, and performance levels. This study aimed to establish robust, generalizable findings through a systematic review and meta-analysis. We investigated three questions: (1) did higher- and lower-performing athletes differ in childhood/adolescent progress, starting age, or amounts of main-sport or other-sports practice or play; (2) do effects differ between junior and adult athletes, compared performance levels, or types of sports; and (3) are effect sizes from different predictors associated with one another? We conducted a systematic literature search in SPORTDiscus, ERIC, ProQuest, PsycINFO, PubMed, Scopus, WorldCat, and Google Scholar until 28 February 2021. Selection criteria included original research studies comparing higher- versus lower-performing athletes regarding one or more of our predictor variables within defined age categories, sports, and sex, and reporting effect sizes or data needed to compute effects sizes. Mean meta-analytic Cohen’s d was calculated for each predictor. Quality of evidence was evaluated using GRADE. In total, 71 study reports met all eligibility criteria and included 262 international athlete samples, 685 effect sizes, and a total sample size of 9241 athletes from local to Olympic competition level and from diverse sports. The following findings emerged. (1) Compared with their national-class counterparts, adult world-class athletes had more childhood/adolescent multi-sport coach-led practice, a later main-sport start, less main-sport practice, and slower initial progress (|0.23| 0.05). (4) Results were robust across types of sports. (5) Effect sizes from different predictors were associated with one another (|0.64|< r <|0.79|). A GRADE assessment revealed a low quality of evidence for peer-led play but a moderate to high quality of evidence for all other predictors. Excess childhood/adolescent specialized practice may hinder athletes’ long-term development through overuse injury, burnout, suboptimal athlete–sport match, and limiting long-term learning capital. By contrast, adult world-class athletes’ childhood/adolescent multi-sport practice with reduced main-sport practice implied a relatively resource-preserving, cost-reducing, and risk-buffering pattern that yielded greater long-term sustainability and practice efficiency.
Chun Yin Kong, Picasso Vasquez, Makan Farhoodimoghadam
et al.
In the rapidly evolving landscape of veterinary healthcare, integrating machine learning (ML) clinical decision-making tools with electronic health records (EHRs) promises to improve diagnostic accuracy and patient care. However, the seamless integration of ML classifiers into existing EHRs in veterinary medicine is frequently hindered by the rigidity of EHR systems or the limited availability of IT resources. To address this shortcoming, we present Anna, a freely-available software solution that provides ML classifier results for EHR laboratory data in real-time.
The analysis of high-intensity runs (or sprints) in soccer has long been a topic of interest for sports science researchers and practitioners. In particular, recent studies suggested contextualizing sprints based on their tactical purposes to better understand the physical-tactical requirements of modern match-play. However, they have a limitation in scalability, as human experts have to manually classify hundreds of sprints for every match. To address this challenge, this paper proposes a deep learning framework for automatically classifying sprints in soccer into contextual categories. The proposed model covers the permutation-invariant and sequential nature of multi-agent trajectories in soccer by deploying Set Transformers and a bidirectional GRU. We train the model with category labels made through the collaboration of human annotators and a rule-based classifier. Experimental results show that our model classifies sprints in the test dataset into 15 categories with the accuracy of 77.65%, implying the potential of the proposed framework for facilitating the integrated analysis of soccer sprints at scale.
Problem definition: Professional sports leagues may be suspended due to various reasons such as the recent COVID-19 pandemic. A critical question the league must address when re-opening is how to appropriately select a subset of the remaining games to conclude the season in a shortened time frame. Academic/practical relevance: Despite the rich literature on scheduling an entire season starting from a blank slate, concluding an existing season is quite different. Our approach attempts to achieve team rankings similar to that which would have resulted had the season been played out in full. Methodology: We propose a data-driven model which exploits predictive and prescriptive analytics to produce a schedule for the remainder of the season comprised of a subset of originally-scheduled games. Our model introduces novel rankings-based objectives within a stochastic optimization model, whose parameters are first estimated using a predictive model. We introduce a deterministic equivalent reformulation along with a tailored Frank-Wolfe algorithm to efficiently solve our problem, as well as a robust counterpart based on min-max regret. Results: We present simulation-based numerical experiments from previous National Basketball Association (NBA) seasons 2004--2019, and show that our models are computationally efficient, outperform a greedy benchmark that approximates a non-rankings-based scheduling policy, and produce interpretable results. Managerial implications: Our data-driven decision-making framework may be used to produce a shortened season with 25-50\% fewer games while still producing an end-of-season ranking similar to that of the full season, had it been played.
The surge of large language models (LLMs) has driven significant progress in medical applications, including traditional Chinese medicine (TCM). However, current medical LLMs struggle with TCM diagnosis and syndrome differentiation due to substantial differences between TCM and modern medical theory, and the scarcity of specialized, high-quality corpora. To this end, in this paper we propose BianCang, a TCM-specific LLM, using a two-stage training process that first injects domain-specific knowledge and then aligns it through targeted stimulation to enhance diagnostic and differentiation capabilities. Specifically, we constructed pre-training corpora, instruction-aligned datasets based on real hospital records, and the ChP-TCM dataset derived from the Pharmacopoeia of the People's Republic of China. We compiled extensive TCM and medical corpora for continual pre-training and supervised fine-tuning, building a comprehensive dataset to refine the model's understanding of TCM. Evaluations across 11 test sets involving 31 models and 4 tasks demonstrate the effectiveness of BianCang, offering valuable insights for future research. Code, datasets, and models are available on https://github.com/QLU-NLP/BianCang.
The objective of this research is to introduce a network specialized in predicting drugs that can be repurposed by investigating real-world evidence sources, such as clinical trials and biomedical literature. Specifically, it aims to generate drug combination therapies for complex diseases (e.g., cancer, Alzheimer's). We present a multilayered network medicine approach, empowered by a highly configured ChatGPT prompt engineering system, which is constructed on the fly to extract drug mentions in clinical trials. Additionally, we introduce a novel algorithm that connects real-world evidence with disease-specific signaling pathways (e.g., KEGG database). This sheds light on the repurposability of drugs if they are found to bind with one or more protein constituents of a signaling pathway. To demonstrate, we instantiated the framework for breast cancer and found that, out of 46 breast cancer signaling pathways, the framework identified 38 pathways that were covered by at least two drugs. This evidence signals the potential for combining those drugs. Specifically, the most covered signaling pathway, ID hsa:2064, was covered by 108 drugs, some of which can be combined. Conversely, the signaling pathway ID hsa:1499 was covered by only two drugs, indicating a significant gap for further research. Our network medicine framework, empowered by GenAI, shows promise in identifying drug combinations with a high degree of specificity, knowing the exact signaling pathways and proteins that serve as targets. It is noteworthy that ChatGPT successfully accelerated the process of identifying drug mentions in clinical trials, though further investigations are required to determine the relationships among the drug mentions.
In this study, we present a method for estimating the viscoelasticity of a leaf-spring sports prosthesis using advanced energy minimizing inverse kinematics based on the Piece-wise Constant Strain (PCS) model to reconstruct the three-dimensional dynamic behavior. Dynamic motion analysis of the athlete and prosthesis is important to clarify the effect of prosthesis characteristics on foot function. However, three-dimensional deformation calculations of the prosthesis and viscoelasticity have rarely been investigated. In this letter, we apply the PCS model to a prosthesis deformation, which can calculate flexible deformation with low computational cost and handle kinematics and dynamics. In addition, we propose an inverse kinematics calculation method that is consistent with the material properties of the prosthesis by considering the minimization of elastic energy. Furthermore, we propose a method to estimate the viscoelasticity by solving a quadratic programming based on the measured motion capture data. The calculated strains are more reasonable than the results obtained by conventional inverse kinematics calculation. From the result of the viscoelasticity estimation, we simulate the prosthetic motion by forward dynamics calculation and confirm that this result corresponds to the measured motion. These results indicate that our approach adequately models the dynamic phenomena, including the viscoelasticity of the prosthesis.
Alice L Kennard, Suzanne Rainsford, Kelly L Hamilton
et al.
Abstract Background Understanding the patient perspective of frailty is critical to offering holistic patient-centred care. Rehabilitation strategies for patients with advanced chronic kidney disease (CKD) and frailty are limited in their ability to overcome patient-perceived barriers to participation, resulting in high rates of drop-out and non-adherence. The aim of this study was to explore patient perspectives and preferences regarding experiences with rehabilitation to inform a CKD/Frailty rehabilitation model. Methods This qualitative study involved two focus groups, six individual semi-structured interviews and three caregiver semi-structured interviews with lived experience of advanced kidney disease and frailty. Interviews were recorded, transcribed, and coded for meaningful concepts and analysed using inductive thematic analysis using constant comparative method of data analysis employing Social Cognitive Theory. Results Six major themes emerged including accommodating frailty is an act of resilience, exercise is endorsed for rehabilitation but existing programs have failed to meet end-users’ needs. Rehabilitation goals were framed around return to normative behaviours and rehabilitation should have a social dimension, offering understanding for “people like us”. Participants reported on barriers and disruptors to frailty rehabilitation in the CKD context. Participants valued peer-to-peer education, the camaraderie of socialisation and the benefit of feedback for maintaining motivation. Patients undertaking dialysis described the commodity of time and the burden of unresolved symptoms as barriers to participation. Participants reported difficulty envisioning strategies for frailty rehabilitation, maintaining a focus on the immediate and avoidance of future uncertainty. Conclusions Frailty rehabilitation efforts in CKD should leverage shared experiences, address comorbidity and symptom burden and focus on goals with normative value.
Gabriela Mąsior, Szymon Dowgiert, Zuzanna Ambroziewicz
et al.
The increasing prevalence of recurrent infections and immune-related disorders has intensified the search for effective immune-modulating interventions. Bacterial lysates, derived from inactivated bacterial components, have emerged as promising immune-boosting supplements. These lysates activate both innate and adaptive immune responses by interacting with pattern recognition receptors such as Toll-like receptors (TLRs), triggering signaling pathways that enhance immune surveillance and pathogen recognition. Among the most extensively studied is OM-85, which has demonstrated efficacy in reducing the frequency and severity of respiratory tract infections (RTIs) and modulating immune pathways involved in asthma and allergic rhinitis. This review evaluates the mechanisms of action, clinical efficacy, and safety profiles of bacterial lysates, particularly focusing on their role in preventing RTIs and managing chronic inflammatory diseases. Evidence supports their potential to reduce antibiotic use and improve overall clinical outcomes in both pediatric and adult populations. Despite their general safety, certain contraindications exist, especially in patients with hypersensitivity or autoimmune conditions. Further research is required to optimize their use, particularly in immunocompromised populations and for broader applications, such as viral infections. Bacterial lysates represent a novel, cost-effective approach to enhancing immune health and reducing the global reliance on antibiotics.
Miguel A. Diaz, M.S., Eric A. Branch, M.D., Jacob G. Dunn, D.O.
et al.
Purpose: To assess the biomechanical performance of different stitching methods using a suturing device by comparing the elongation, stiffness, failure load, and time to stitch completion in cadaveric semitendinosus tendons (STs) and quadriceps tendons (QTs). Methods: A total of 24 STs and 16 QTs were harvested from cadaveric knee specimens (N = 40). Samples were randomly divided into 2 groups: whipstitch (WS) and whip-lock (WL) stitch. Both tendon ends were clamped to a graft preparation stand, and a 2-part needle was used to place 5 stitches, each 0.5 cm apart. Stitching time was recorded. Samples were preconditioned and then underwent cyclic loading from 50 to 200 N at 1 Hz for 500 cycles, followed by load-to-failure testing at 20 mm/min. Stiffness (in newtons per millimeter), ultimate failure load (in newtons), peak-to-peak displacement (in millimeters), elongation (in millimeters), and failure displacement (in millimeters) were recorded. Results: Completion of the WS was significantly faster than the WL stitch in the ST (P < .001) and QT (P = .004). For the ST, the WL stitch exhibited higher ultimate failure loads and construct stiffness than the WS. Regarding the QT, the WL stitch showed higher stiffness and displacement than the WS; however, the ultimate failure load was higher for the WS in the QT. The ultimate failure load in the QT was higher than that in the ST for both stitches. In the ST, only 25% of WSs and 100% of WL stitches failed due to suture breakage. In the QT, suture breakage led to the failure of 100% of both the WL stitches and WSs. Conclusions: In the ST, the WL stitch resulted in improved biomechanical performance through higher ultimate load and fewer failures from tissue damage compared with the WS. In the QT, both the WS and the WL stitch showed similar biomechanical performance with ultimate failure loads above established clinical failure thresholds. Clinical Relevance: Various types of ligament and tendon injuries require suturing to enable repair or reconstruction. The success of ligament or tendon surgery often relies on soft-tissue quality. It is important to investigate the biomechanical properties of stitching techniques that help preserve soft-tissue quality as a step to determining their clinical suitability.