Large language models (LLMs) have demonstrated strong performance on medical benchmarks, including question answering and diagnosis. To enable their use in clinical settings, LLMs are typically further adapted through continued pretraining or post-training using clinical data. However, most medical LLMs are trained on data from a single institution, which faces limitations in generalizability and safety in heterogeneous systems. Federated learning (FL) is a promising solution for enabling collaborative model development across healthcare institutions. Yet applying FL to LLMs in medicine remains fundamentally limited. First, conventional FL requires transmitting the full model during each communication round, which becomes impractical for multi-billion-parameter LLMs given the limited computational resources. Second, many FL algorithms implicitly assume data homogeneity, whereas real-world clinical data are highly heterogeneous across patients, diseases, and institutional practices. We introduce the model-agnostic and parameter-efficient federated learning framework for adapting LLMs to medical applications. Fed-MedLoRA transmits only low-rank adapter parameters, reducing communication and computation overhead, while Fed-MedLoRA+ further incorporates adaptive, data-aware aggregation to improve convergence under cross-site heterogeneity. We apply the framework to clinical information extraction (IE), which transforms patient narratives into structured medical entities and relations. Accuracy was assessed across five patient cohorts through comparisons with BERT models, and LLaMA-3 and DeepSeek-R1, GPT-4o models. Evaluation settings included (1) in-domain training and testing, (2) external validation on independent cohorts, and (3) a low-resource new-site adaptation scenario using real-world clinical notes from the Yale New Haven Health System.
Catherine Krus, Ian Zander, Tyler J. Sherman
et al.
Bluetongue virus (BTV) and epizootic hemorrhagic disease virus (EHDV) are two viruses belonging to the genus Orbivirus that are transmitted via insect vector, the Culicoides biting midge, causing disease in domestic and wild ruminants. These infections can lead to significant morbidity, mortality, and production losses in livestock, with economic consequences for cattle and sheep industries. Despite their growing impact due to environmental and anthropogenic changes, little is known of the prevalence of these viruses in North American bison (Bison bison). We present the first cross-sectional survey of BTV and EHDV in North American bison, with samples collected from 287 animals across 9 herds in 7 U.S. states from September to November 2023. Using competitive enzyme-linked immunosorbent assays (cELISA), we detected seroprevalence rates of 56.5% for BTV and 57.5% for EHDV. We found higher seroprevalence in North American bison compared to reports in European bison populations, suggesting that bison could potentially serve as incidental hosts of orbiviruses during key transmission periods; however, their role in virus transmission remains uncertain and warrants further investigation, particularly regarding the duration of viremia, potential amplification capacity, and year-to-year variability in PCR positivity. Logistic regression analysis revealed age as a significant predictor for both BTV (OR: 1.15, CI: 1.05–1.26, p: 0.006) and EHDV (OR: 1.16, CI: 1.06–1.28, p: 0.0014) seropositivity. PCR amplification identified circulating BTV serotypes 6, 11, 13, 17. Additionally, age was negatively associated with PCR positivity for both BTV (OR: 0.70, CI: 0.53–0.93, p: 0.014) and EHDV (OR: 0.56, CI: 0.33–0.93, p: 0.024), suggesting a decline in detectable viremia with increasing age. Although complex environmental and epidemiological factors likely play a role, this trend may be due to older animals having experienced more vector seasons, thereby increasing their cumulative exposure and subsequent immunity to these viruses over time. The significant age-associated dynamics reveal the importance of considering life stage in disease surveillance and management. Our study also highlights the importance of integrating bison into future vector-borne disease research and control strategies to mitigate risks to livestock, wildlife, and ecosystem health.
This article was designed to investigate the bursa of Fabricius (BF) morphometric and effects of glutathione peroxidase enzyme (GPX-1) in infectious bursal disease vaccinated (IBDV) broiler supplemented with selenium-nanomaterials (Se-NPs). Ninety-six one-day-old of the Ross 308 broilers were assigned into E1, E2, E3, and E4 of 24 chicks each; those of E1 served as control, E2 vaccinated with an intermediate strain of IBD at 12 and 20 days old through the eye drop, E3 received a mix of Se-NPs and the vaccine, and E4 was supplemented with Se-NPs (0.3mg/kg). At 19, 26, and 42 days of age, the diameters of the bursa of Fabricius were measured, and liver tissue was sampled to determine the reverse transcription polymerase chain reaction (RT-PCR) of the Glutathione peroxidase enzyme. Results indicated that group E4 (selenium-Nanoparticles group) significantly increased the bursal morphometric at 26 days old, and E3 increased the morphometric of the bursa Fabricius at 42 days old, as well as supplementation of Se-NPs significantly up-regulated RT-PCR of GPX-1 at 26 and 42 days old. The highest gene expression was in the selenium-nanoparticle group at 26 and 42 days old, in contrast to the other groups. Based on this finding, it can be concluded that nanomaterials improved the morphology of immune organs and enzyme peroxidase activity in vaccinated broilers with the Gumboro vaccine.
Traditional Chinese Medicine (TCM) theory is built on imagistic thinking, in which medical principles and diagnostic and therapeutic logic are structured through metaphor and metonymy. However, existing English translations largely rely on literal rendering, making it difficult for target-language readers to reconstruct the underlying conceptual networks and apply them in clinical practice. This study adopted a human-in-the-loop (HITL) framework and selected four passages from the medical canon Huangdi Neijing that are fundamental in theory. Through prompt-based cognitive scaffolding, DeepSeek V3.1 was guided to identify metaphor and metonymy in the source text and convey the theory in translation. In the evaluation stage, ChatGPT 5 Pro and Gemini 2.5 Pro were instructed by prompts to simulate three types of real-world readers. Human translations, baseline model translations, and prompt-adjusted translations were scored by the simulated readers across five cognitive dimensions, followed by structured interviews and Interpretative Phenomenological Analysis (IPA). Results show that the prompt-adjusted LLM translations perform best across all five dimensions, with high cross-model and cross-role consistency. The interview themes reveal differences between human and machine translation, effective strategies for metaphor and metonymy transfer, and readers' cognitive preferences. This study provides a cognitive, efficient, and replicable HITL methodological pathway for the translation of ancient, concept-dense texts such as TCM.
Evidence-based medicine (EBM) holds a crucial role in clinical application. Given suitable medical articles, doctors effectively reduce the incidence of misdiagnoses. Researchers find it efficient to use large language models (LLMs) techniques like RAG for EBM tasks. However, the EBM maintains stringent requirements for evidence, and RAG applications in EBM struggle to efficiently distinguish high-quality evidence. Therefore, inspired by the meta-analysis used in EBM, we provide a new method to re-rank and filter the medical evidence. This method presents multiple principles to filter the best evidence for LLMs to diagnose. We employ a combination of several EBM methods to emulate the meta-analysis, which includes reliability analysis, heterogeneity analysis, and extrapolation analysis. These processes allow the users to retrieve the best medical evidence for the LLMs. Ultimately, we evaluate these high-quality articles and show an accuracy improvement of up to 11.4% in our experiments and results. Our method successfully enables RAG to extract higher-quality and more reliable evidence from the PubMed dataset. This work can reduce the infusion of incorrect knowledge into responses and help users receive more effective replies.
Traditional Chinese Medicine (TCM) presents a rich and structurally unique knowledge system that challenges conventional applications of large language models (LLMs). Although previous TCM-specific LLMs have shown progress through supervised fine-tuning, they often face limitations in alignment, data quality, and evaluation consistency. In this study, we introduce Ladder-base, the first TCM-focused LLM trained with Group Relative Policy Optimization (GRPO), a reinforcement learning method that improves reasoning and factual consistency by optimizing response selection based on intra-group comparisons. Ladder-base is built upon the Qwen2.5-7B-Instruct foundation model and trained exclusively on the textual subset of the TCM-Ladder benchmark, using 80 percent of the data for training and the remaining 20 percent split evenly between validation and test sets. Through standardized evaluation, Ladder-base demonstrates superior performance across multiple reasoning metrics when compared to both state-of-the-art general-purpose LLMs such as GPT-4, Gemini 2.5, Claude 3, and Qwen3 and domain-specific TCM models including BenTsao, HuatuoGPT2, and Zhongjing. These findings suggest that GRPO provides an effective and efficient strategy for aligning LLMs with expert-level reasoning in traditional medical domains and supports the development of trustworthy and clinically grounded TCM artificial intelligence systems.
Epileptic seizures arise from abnormally synchronised neural activity and remain a major global health challenge, affecting more than 50 million people worldwide. Despite advances in pharmacological interventions, a significant proportion of patients continue to experience uncontrolled seizures, underscoring the need for alternative neuromodulation strategies. Rhythmic neural entrainment has recently emerged as a promising mechanism for disrupting pathological synchrony, but most existing systems rely on complex analogue electronics or high-power stimulation hardware. This study investigates a minimal digital custom-designed chip that generates a stable 6 Hz oscillation capable of entraining epileptic seizure activity. Using a publicly available EEG seizure dataset, we extracted and averaged analogue seizure waveforms, digitised them to emulate neural front-ends, and directly interfaced the digitised signals with digital output recordings acquired from the chip using a Saleae Logic analyser. The chip pulse train was resampled and low-pass-reconstructed to produce an analogue 6 Hz waveform, allowing direct comparison between seizure morphology, its digitised representation, and the entrained output. Frequency-domain and time-domain analyses demonstrate that the chip imposes a narrow-band 6 Hz rhythm that overrides the broadband spectral profile of seizure activity. These results provide a proof-of-concept for low-power digital custom-designed entrainment as a potential pathway toward simplified, wearable seizure-interruption devices for precision medicine and future healthcare devices.
The rise in chronic diseases over the last century presents a significant health and economic burden globally. Here we apply evolutionary medicine and life history theory to better understand their development. We highlight an imbalanced metabolic axis of growth and proliferation (anabolic) versus maintenance and dormancy (catabolic), focusing on major mechanisms including IGF-1, mTOR, AMPK, and Klotho. We also relate this axis to the hyperfunction theory of aging, which similarly implicates anabolic mechanisms like mTOR in aging and disease. Next, we highlight the Brain-Body Energy Conservation model, which connects the hyperfunction theory with energetic trade-offs that induce hypofunction and catabolic health risks like impaired immunity. Finally, we discuss how modern environmental mismatches exacerbate this process. Following our review, we discuss future research directions to better understand health risk. This includes studying IGF-1, mTOR, AMPK, and Klotho and how they relate to health and aging in human subsistence populations, including with lifestyle shifts. It also includes understanding their role in the developmental origins of health and disease as well as the social determinants of health disparities. Further, we discuss the need for future studies on exceptionally long-lived species to understand potentially underappreciated trade-offs and costs that come with their longevity. We close with considering possible implications for therapeutics, including (1) compensatory pathways counteracting treatments, (2) a Goldilocks zone, in which suppressing anabolic metabolism too far introduces catabolic health risks, and (3) species constraints, in which therapeutics tested in shorter lived species with greater anabolic imbalance will be less effective in humans.
Traditional Chinese Medicine (TCM) formulas play a significant role in treating epidemics and complex diseases. Existing models for TCM utilize traditional algorithms or deep learning techniques to analyze formula relationships, yet lack comprehensive results, such as complete formula compositions and detailed explanations. Although recent efforts have used TCM instruction datasets to fine-tune Large Language Models (LLMs) for explainable formula generation, existing datasets lack sufficient details, such as the roles of the formula's sovereign, minister, assistant, courier; efficacy; contraindications; tongue and pulse diagnosis-limiting the depth of model outputs. To address these challenges, we propose ZhiFangDanTai, a framework combining Graph-based Retrieval-Augmented Generation (GraphRAG) with LLM fine-tuning. ZhiFangDanTai uses GraphRAG to retrieve and synthesize structured TCM knowledge into concise summaries, while also constructing an enhanced instruction dataset to improve LLMs' ability to integrate retrieved information. Furthermore, we provide novel theoretical proofs demonstrating that integrating GraphRAG with fine-tuning techniques can reduce generalization error and hallucination rates in the TCM formula task. Experimental results on both collected and clinical datasets demonstrate that ZhiFangDanTai achieves significant improvements over state-of-the-art models. Our model is open-sourced at https://huggingface.co/tczzx6/ZhiFangDanTai1.0.
Traditional Chinese Medicine (TCM) represents a rich repository of ancient medical knowledge that continues to play an important role in modern healthcare. Due to the complexity and breadth of the TCM literature, the integration of AI technologies is critical for its modernization and broader accessibility. However, this integration poses considerable challenges, including the interpretation of obscure classical Chinese texts and the modeling of intricate semantic relationships among TCM concepts. In this paper, we develop OpenTCM, an LLM-based system that combines a domain-specific TCM knowledge graph and Graph-based Retrieval-Augmented Generation (GraphRAG). First, we extract more than 3.73 million classical Chinese characters from 68 gynecological books in the Chinese Medical Classics Database, with the help of TCM and gynecology experts. Second, we construct a comprehensive multi-relational knowledge graph comprising more than 48,000 entities and 152,000 interrelationships, using customized prompts and Chinese-oriented LLMs such as DeepSeek and Kimi to ensure high-fidelity semantic understanding. Last, we empower OpenTCM with GraphRAG, enabling high-fidelity ingredient knowledge retrieval and diagnostic question-answering without model fine-tuning. Experimental evaluations demonstrate that OpenTCM achieves mean expert scores (MES) of 4.378 in ingredient information retrieval and 4.045 in diagnostic question-answering tasks, outperforming state-of-the-art solutions in real-world TCM use cases.
Lukas Buess, Matthias Keicher, Nassir Navab
et al.
Generative artificial intelligence (AI) models, such as diffusion models and OpenAI's ChatGPT, are transforming medicine by enhancing diagnostic accuracy and automating clinical workflows. The field has advanced rapidly, evolving from text-only large language models for tasks such as clinical documentation and decision support to multimodal AI systems capable of integrating diverse data modalities, including imaging, text, and structured data, within a single model. The diverse landscape of these technologies, along with rising interest, highlights the need for a comprehensive review of their applications and potential. This scoping review explores the evolution of multimodal AI, highlighting its methods, applications, datasets, and evaluation in clinical settings. Adhering to PRISMA-ScR guidelines, we systematically queried PubMed, IEEE Xplore, and Web of Science, prioritizing recent studies published up to the end of 2024. After rigorous screening, 144 papers were included, revealing key trends and challenges in this dynamic field. Our findings underscore a shift from unimodal to multimodal approaches, driving innovations in diagnostic support, medical report generation, drug discovery, and conversational AI. However, critical challenges remain, including the integration of heterogeneous data types, improving model interpretability, addressing ethical concerns, and validating AI systems in real-world clinical settings. This review summarizes the current state of the art, identifies critical gaps, and provides insights to guide the development of scalable, trustworthy, and clinically impactful multimodal AI solutions in healthcare.
Eghosa Osas UWAGBOE, Adejoke Adebusola ADELUSI, Francis Adetunji ADESIDA
A comprehensive Integrated Pest Management (IPM) is an ecosystem approach to control insect and disease pests to minimize the use of pesticides. Cross River State was purposively selected based on training done by International Institute of Tropical Agriculture (IITA) through Sustainable Tree Crop Programme. The study explored the determinants of adoption behaviour. A systematic sampling technique was used to select a total of 108 IPM trained respondents. Structured questionnaire was used to gather information on farmers’socio-economic factors affecting IPM Adoption, determine the IPM adoption behaviour of farmers and ascertain the constraints experienced from IPM adoption. Male respondents were 81.5% while the females were 18.5%. Majority (65.1%), were between the age range of 41 and 60 years which is an indication that they were still in their prime age. Majority (94.4%), of the respondents were educated and majority (97.2%) of the farmers own small farms between 1 and 5 ha. Most of the respondents rated inaccessibility to market information as the highest constraint affecting IPM adoption with a Weighted Mean Score of 0.8. Majority (79.6%) had high level of intensity of IPM adoption while most of the respondents rated both pest monitoring and planting resistant varieties as the highest rate of adoption with a score of 2.9. A significant relationship exists between sex (X2 =42.815, p<0.05), age (X2 =65.148, p<0.05), education (X2 = 40.426, p<0.05), years of experience (X2 =110.333, p<0.05), and adoption behaviour. The contingency coefficient (CC) shows very strong relationship of sex 0.5328, age 0.6134, marital status 0.7758, education 0.5218 and years of experience 0.7108 with adoption behaviour. Farmers need to be encouraged in adopting IPM through marketing information that would reduce extortion of the farmers by local buying agents.
Hanadi B. A. Baghdadi, Mohamed Abdelsalam, Marwa M. Attia
Abstract Strongylus vulgaris, a devastating parasitic nematode in equids, causes life-threatening verminous aneurysms that are challenging to diagnose early. This study pioneered integrating nanotechnology into an indirect enzyme-linked immunosorbent assay (i-ELISA) system to enhance the sensitivity and specificity for detecting S. vulgaris larval antigens in equine serum samples, with PCR confirmation of the species. A conventional i-ELISA and an innovative nano-based ELISA were developed using excretory-secretory antigens from adult S. vulgaris worms. The nano-ELISA incorporated gold nanoparticles (17.4–41.4 nm) conjugated with detection antibodies, enabling remarkable signal amplification. Of the 120 examined equines, 100 (83.33%) were positive for S. vulgaris infection. A conventional i-ELISA and an innovative nano-ELISA incorporating 17.4–41.4 nm gold nanoparticles were optimized using S. vulgaris excretory-secretory antigens. Both assays demonstrated high specificity, with no cross-reactivity against sera from animals infected with other helminth parasites. Remarkably, optical density (OD) readings from both i-ELISAs exhibited a positive quantitative correlation with infection intensity. The i-ELISA OD ranged from 0.45–0.74 (G3), 0.75–0.94 (G2), to 0.95–2.5 (G1). The nano-ELISA showed enhanced signal amplification, with OD ranging from 0.40–0.84 (G3), 0.85–0.99 (G2), to 1.0–3.5 (G1). This nanotechnology-amplified ELISA opens new, highly sensitive, and specific techniques for parasitic diagnosis in equine medicine. Its superior performance, facilitated by signal-amplifying gold nanoparticles, illuminates nanotechnology's potential in revolutionizing parasitological diagnostics for enhanced animal health and welfare management.
The emergent abilities of large language models (LLMs) have demonstrated great potential in solving medical questions. They can possess considerable medical knowledge, but may still hallucinate and are inflexible in the knowledge updates. While Retrieval-Augmented Generation (RAG) has been proposed to enhance the medical question-answering capabilities of LLMs with external knowledge bases, it may still fail in complex cases where multiple rounds of information-seeking are required. To address such an issue, we propose iterative RAG for medicine (i-MedRAG), where LLMs can iteratively ask follow-up queries based on previous information-seeking attempts. In each iteration of i-MedRAG, the follow-up queries will be answered by a conventional RAG system and they will be further used to guide the query generation in the next iteration. Our experiments show the improved performance of various LLMs brought by i-MedRAG compared with conventional RAG on complex questions from clinical vignettes in the United States Medical Licensing Examination (USMLE), as well as various knowledge tests in the Massive Multitask Language Understanding (MMLU) dataset. Notably, our zero-shot i-MedRAG outperforms all existing prompt engineering and fine-tuning methods on GPT-3.5, achieving an accuracy of 69.68% on the MedQA dataset. In addition, we characterize the scaling properties of i-MedRAG with different iterations of follow-up queries and different numbers of queries per iteration. Our case studies show that i-MedRAG can flexibly ask follow-up queries to form reasoning chains, providing an in-depth analysis of medical questions. To the best of our knowledge, this is the first-of-its-kind study on incorporating follow-up queries into medical RAG. The implementation of i-MedRAG is available at https://github.com/Teddy-XiongGZ/MedRAG.
The precise quantification of nucleic acids is pivotal in molecular biology, underscored by the rising prominence of nucleic acid amplification tests (NAAT) in diagnosing infectious diseases and conducting genomic studies. This review examines recent advancements in digital Polymerase Chain Reaction (dPCR) and digital Loop-mediated Isothermal Amplification (dLAMP), which surpass the limitations of traditional NAAT by offering absolute quantification and enhanced sensitivity. In this review, we summarize the compelling advancements of dNNAT in addressing pressing public health issues, especially during the COVID-19 pandemic. Further, we explore the transformative role of artificial intelligence (AI) in enhancing dNAAT image analysis, which not only improves efficiency and accuracy but also addresses traditional constraints related to cost, complexity, and data interpretation. In encompassing the state-of-the-art (SOTA) development and potential of both software and hardware, the all-encompassing Point-of-Care Testing (POCT) systems cast new light on benefits including higher throughput, label-free detection, and expanded multiplex analyses. While acknowledging the enhancement of AI-enhanced dNAAT technology, this review aims to both fill critical gaps in the existing technologies through comparative assessments and offer a balanced perspective on the current trajectory, including attendant challenges and future directions. Leveraging AI, next-generation dPCR and dLAMP technologies promises integration into clinical practice, improving personalized medicine, real-time epidemic surveillance, and global diagnostic accessibility.
Accurate interpretation of lab results is crucial in clinical medicine, yet most patient portals use universal normal ranges, ignoring conditional factors like age and gender. This study introduces Lab-AI, an interactive system that offers personalized normal ranges using retrieval-augmented generation (RAG) from credible health sources. Lab-AI has two modules: factor retrieval and normal range retrieval. We tested these on 122 lab tests: 40 with conditional factors and 82 without. For tests with factors, normal ranges depend on patient-specific information. Our results show GPT-4-turbo with RAG achieved a 0.948 F1 score for factor retrieval and 0.995 accuracy for normal range retrieval. GPT-4-turbo with RAG outperformed the best non-RAG system by 33.5% in factor retrieval and showed 132% and 100% improvements in question-level and lab-level performance, respectively, for normal range retrieval. These findings highlight Lab-AI's potential to enhance patient understanding of lab results.
The emerging trend of advancing generalist artificial intelligence, such as GPTv4 and Gemini, has reshaped the landscape of research (academia and industry) in machine learning and many other research areas. However, domain-specific applications of such foundation models (e.g., in medicine) remain untouched or often at their very early stages. It will require an individual set of transfer learning and model adaptation techniques by further expanding and injecting these models with domain knowledge and data. The development of such technologies could be largely accelerated if the bundle of data, algorithms, and pre-trained foundation models were gathered together and open-sourced in an organized manner. In this work, we present OpenMEDLab, an open-source platform for multi-modality foundation models. It encapsulates not only solutions of pioneering attempts in prompting and fine-tuning large language and vision models for frontline clinical and bioinformatic applications but also building domain-specific foundation models with large-scale multi-modal medical data. Importantly, it opens access to a group of pre-trained foundation models for various medical image modalities, clinical text, protein engineering, etc. Inspiring and competitive results are also demonstrated for each collected approach and model in a variety of benchmarks for downstream tasks. We welcome researchers in the field of medical artificial intelligence to continuously contribute cutting-edge methods and models to OpenMEDLab, which can be accessed via https://github.com/openmedlab.