The clinical deployment of AI diagnostic models demands more than benchmark accuracy - it demands robustness across the full spectrum of disease presentations. However, publicly available chest radiographic datasets systematically underrepresent critical clinical feature combinations, leaving models under-trained precisely where clinical stakes are highest. We present CARS, a clinically aware and anatomically grounded framework that addresses this gap through principled synthetic image generation. CARS applies targeted perturbations to clinical feature vectors, enabling controlled insertion and deletion of pathological findings while explicitly preserving anatomical structure. We evaluate CARS across seven backbone architectures by fine-tuning models on synthetic subsets and testing on a held-out MIMIC-CXR benchmark. Compared to prior feature perturbation approaches, fine-tuning on CARS-generated images consistently improves precision-recall performance, reduces predictive uncertainty, and improves model calibration. Structural and semantic analyses demonstrate high anatomical fidelity, strong feature alignment, and low semantic uncertainty. Independent evaluation by two expert radiologists further confirms realism and clinical agreement. As the field moves toward regulated clinical AI, CARS demonstrates that anatomically faithful synthetic data generation for better feature space coverage is a viable and effective strategy for improving both the performance and trustworthiness of chest X-ray classification systems - without compromising clinical integrity.
ABSTRACT Objective To clarify the link between environmental pollution and diabetes risk by focusing on pancreatic β‐cells as key targets of environmental insults, with emphasis on the role of endocrine‐disrupting chemicals (EDCs) in pancreatic dysfunction and diabetes pathogenesis. Methods This narrative review synthesises recent research on EDCs, focusing on their effects on β‐cells. The literature search included studies in English on EDCs, diabetes, and β‐cell function, utilising Boolean operators to refine the search. Results EDCs impair β‐cell function through mechanisms such as oxidative stress, mitochondrial damage, and epigenetic changes. These pollutants disrupt insulin synthesis, secretion, and β‐cell survival, which is distinct from their general metabolic effects. Additionally, EDCs may interact synergistically with traditional diabetes risk factors, such as high‐fat diets, amplifying the risk of diabetes. Conclusion Environmental pollutants play a significant role in β‐cell dysfunction and diabetes, offering new directions for research and prevention.
Oriel Perets, Ofir Ben Shoham, Nir Grinberg
et al.
Medical benchmark datasets significantly contribute to developing Large Language Models (LLMs) for medical knowledge extraction, diagnosis, summarization, and other uses. Yet, current benchmarks are mainly derived from exam questions given to medical students or cases described in the medical literature, lacking the complexity of real-world patient cases that deviate from classic textbook abstractions. These include rare diseases, uncommon presentations of common diseases, and unexpected treatment responses. Here, we construct Clinically Uncommon Patient Cases and Diagnosis Dataset (CUPCase) based on 3,562 real-world case reports from BMC, including diagnoses in open-ended textual format and as multiple-choice options with distractors. Using this dataset, we evaluate the ability of state-of-the-art LLMs, including both general-purpose and Clinical LLMs, to identify and correctly diagnose a patient case, and test models' performance when only partial information about cases is available. Our findings show that general-purpose GPT-4o attains the best performance in both the multiple-choice task (average accuracy of 87.9%) and the open-ended task (BERTScore F1 of 0.764), outperforming several LLMs with a focus on the medical domain such as Meditron-70B and MedLM-Large. Moreover, GPT-4o was able to maintain 87% and 88% of its performance with only the first 20% of tokens of the case presentation in multiple-choice and free text, respectively, highlighting the potential of LLMs to aid in early diagnosis in real-world cases. CUPCase expands our ability to evaluate LLMs for clinical decision support in an open and reproducible manner.
Patients with rare neurological diseases report cognitive symptoms -"brain fog"- invisible to traditional tests. We propose continuous neurocognitive monitoring via smartphone speech analysis integrated with Relational Graph Transformer (RELGT) architectures. Proof-of-concept in phenylketonuria (PKU) shows speech-derived "Proficiency in Verbal Discourse" correlates with blood phenylalanine (p = -0.50, p < 0.005) but not standard cognitive tests (all |r| < 0.35). RELGT could overcome information bottlenecks in heterogeneous medical data (speech, labs, assessments), enabling predictive alerts weeks before decompensation. Key challenges: multi-disease validation, clinical workflow integration, equitable multilingual deployment. Success would transform episodic neurology into continuous personalized monitoring for millions globally.
Diagnosing and managing oral diseases necessitate advanced visual interpretation across diverse imaging modalities and integrated information synthesis. While current AI models excel at isolated tasks, they often fall short in addressing the complex, multimodal requirements of comprehensive clinical dental practice. Here we introduce DentVLM, a multimodal vision-language model engineered for expert-level oral disease diagnosis. DentVLM was developed using a comprehensive, large-scale, bilingual dataset of 110,447 images and 2.46 million visual question-answering (VQA) pairs. The model is capable of interpreting seven 2D oral imaging modalities across 36 diagnostic tasks, significantly outperforming leading proprietary and open-source models by 19.6% higher accuracy for oral diseases and 27.9% for malocclusions. In a clinical study involving 25 dentists, evaluating 1,946 patients and encompassing 3,105 QA pairs, DentVLM surpassed the diagnostic performance of 13 junior dentists on 21 of 36 tasks and exceeded that of 12 senior dentists on 12 of 36 tasks. When integrated into a collaborative workflow, DentVLM elevated junior dentists' performance to senior levels and reduced diagnostic time for all practitioners by 15-22%. Furthermore, DentVLM exhibited promising performance across three practical utility scenarios, including home-based dental health management, hospital-based intelligent diagnosis and multi-agent collaborative interaction. These findings establish DentVLM as a robust clinical decision support tool, poised to enhance primary dental care, mitigate provider-patient imbalances, and democratize access to specialized medical expertise within the field of dentistry.
ObjectivesTo investigate various supplements that improve insulin resistance, hormonal status, and oxidative stress in overweight or obese women with polycystic ovarian syndrome (PCOS).MethodsA literature search was conducted on four different databases, which led to the discovery of twenty - five randomized controlled trials (RCTs). These RCTs evaluated the efficacy of various supplements in improving insulin resistance (IR), hormonal status, and oxidative stress among overweight or obese women diagnosed with PCOS. Subsequently, data extraction and analysis were carried out to determine the quality of the study’s methodological design and the potential for bias. Moreover, a meta-analysis was performed using the data from the RCTs.ResultsA total of 25 RCTs were carried out, and 1636 women were enrolled. All participants were overweight or obese. The standardized mean differences (SMD) were as follows: For fasting plasma glucose (FPG), it was -0.34 (95% confidence interval [CI], -0.49 to -0.19, p = 0.123, I2 = 30.8%); for insulin, it was -0.67 (95% CI, -0.83 to -0.52, p = 0.208, I2 = 24%); for fasting insulin (FI), it was -0.26 (95% CI, -0.52 to -0.00, p = 0.269, I2 = 21.9%); for homeostatic model assessment-insulin resistance index (HOMA-IR), it was -0.59 (95% CI, -0.73 to -0.45, p = 0.015, I2 = 48.7%); for homoeostatic model assessment beta - cell function (HOMA-B), it was -0.51 (95% CI, -0.75 to -0.27, p = 0.547, I2 = 0%); for quantitative insulin sensitivity check index (QUICKI), it was 0.94 (95% CI, 0.76 to -1.12, p = 0.191, I2 = 27.5%); for total testosterone, it was -0.61 (95% CI, -1.14 to -0.09, p = 0.00, I2 = 78.5%); for testosterone, it was -0.38 (95% CI, -0.86 to 0.10, p = 0.03, I2 = 71.5%); for follicle - stimulating hormone (FSH), it was 0.16 (95% CI, -0.08 to 0.40, p = 0.470, I2 = 0%); for luteinizing hormone (LH), it was -0.56 (95% CI, -1.32 to 0.20, p = 0.000, I2 = 91.1%); for sex hormone - binding globulin (SHBG), it was 0.35 (95% CI, 0.02 to 0.69, p = 0.000, I2 = 78%); for dehydroepiandrosterone (DHEAS), it was -0.27 (95% CI, -0.76 to 0.21, p = 0.001, I2 = 78.7%); for plasma total antioxidant capacity (TAC), it was 0.87 (95% CI, 0.45 to 1.30, p = 0.004, I2 = 71.3%); for plasma malondialdehyde (MDA), it was -0.57 (95% CI, -0.79 to -0.36, p = 0.992, I2 = 0.0%).ConclusionThis study’s findings indicate that, in comparison with a placebo, supplements have a favorable effect on IR, hormonal functions, and oxidative stress in PCOS. Nevertheless, it is crucial to note that the above-drawn conclusions need to be verified by more high-quality studies, given the limitations regarding the number and quality of the included studies.
Diseases of the endocrine glands. Clinical endocrinology
Abstract Introduction Polycystic ovary syndrome (PCOS) is a common endocrine pathology in women. In addition to infertility, women with PCOS have metabolic dysregulation which predisposes them to Type 2 diabetes, cardiovascular disease and non‐alcoholic fatty liver disease. Moreover, women with PCOS have changes in their gut microbial community that may be indicative of dysbiosis. While hyperandrogenism is associated with both the development of metabolic dysfunction and gut dysbiosis in females, the mechanisms involved are not well understood. Methods We used dihydrotestosterone (DHT) and ovariectomy (OVX) mouse models coupled with metabolic assessments and 16S rRNA gene sequencing to explore the contributions of hyperandrogenism and oestrogen deficiency to the development of insulin resistance and gut microbial dysbiosis in pubertal female mice. Results We demonstrated that, while DHT treatment or OVX alone were insufficient to induce insulin resistance during the pubertal‐to‐adult transition, combining OVX with DHT resulted in insulin resistance similar to that observed in letrozole‐treated mice with elevated testosterone and decreased oestrogen levels. In addition, our results showed that OVX and DHT in combination resulted in a distinct shift in the gut microbiome compared to DHT or OVX alone, suggesting that the substantial metabolic dysregulation occurring in the OVX + DHT model was accompanied by unique changes in the abundances of gut bacteria including S24‐7, Rikenellaceae and Mucispirillum schaedleri. Conclusions While hyperandrogenism plays an important role in the development of metabolic dysregulation in female mice, our results indicate that investigation into additional factors influencing insulin resistance and the gut microbiome during the pubertal‐to‐adult transition could provide additional insight into the pathophysiology of PCOS.
Diseases of the endocrine glands. Clinical endocrinology
The rising prevalence of vision-threatening retinal diseases poses a significant burden on the global healthcare systems. Deep learning (DL) offers a promising solution for automatic disease screening but demands substantial data. Collecting and labeling large volumes of ophthalmic images across various modalities encounters several real-world challenges, especially for rare diseases. Here, we introduce EyeDiff, a text-to-image model designed to generate multimodal ophthalmic images from natural language prompts and evaluate its applicability in diagnosing common and rare diseases. EyeDiff is trained on eight large-scale datasets using the advanced latent diffusion model, covering 14 ophthalmic image modalities and over 80 ocular diseases, and is adapted to ten multi-country external datasets. The generated images accurately capture essential lesional characteristics, achieving high alignment with text prompts as evaluated by objective metrics and human experts. Furthermore, integrating generated images significantly enhances the accuracy of detecting minority classes and rare eye diseases, surpassing traditional oversampling methods in addressing data imbalance. EyeDiff effectively tackles the issue of data imbalance and insufficiency typically encountered in rare diseases and addresses the challenges of collecting large-scale annotated images, offering a transformative solution to enhance the development of expert-level diseases diagnosis models in ophthalmic field.
Electronic Medical Records (EMRs), while integral to modern healthcare, present challenges for clinical reasoning and diagnosis due to their complexity and information redundancy. To address this, we proposed medIKAL (Integrating Knowledge Graphs as Assistants of LLMs), a framework that combines Large Language Models (LLMs) with knowledge graphs (KGs) to enhance diagnostic capabilities. medIKAL assigns weighted importance to entities in medical records based on their type, enabling precise localization of candidate diseases within KGs. It innovatively employs a residual network-like approach, allowing initial diagnosis by the LLM to be merged into KG search results. Through a path-based reranking algorithm and a fill-in-the-blank style prompt template, it further refined the diagnostic process. We validated medIKAL's effectiveness through extensive experiments on a newly introduced open-sourced Chinese EMR dataset, demonstrating its potential to improve clinical diagnosis in real-world settings.
This study compares the efficacy of GPT-4 and clinalytix Medical AI in predicting the clinical risk of delirium development. Findings indicate that GPT-4 exhibited significant deficiencies in identifying positive cases and struggled to provide reliable probability estimates for delirium risk, while clinalytix Medical AI demonstrated superior accuracy. A thorough analysis of the large language model's (LLM) outputs elucidated potential causes for these discrepancies, consistent with limitations reported in extant literature. These results underscore the challenges LLMs face in accurately diagnosing conditions and interpreting complex clinical data. While LLMs hold substantial potential in healthcare, they are currently unsuitable for independent clinical decision-making. Instead, they should be employed in assistive roles, complementing clinical expertise. Continued human oversight remains essential to ensure optimal outcomes for both patients and healthcare providers.
Parvin Mirmiran, Shabnam Hosseini, Zahra Bahadoran
et al.
Abstract Background We aimed to assess potential associations of habitual dietary pattern scores in relation to the risk of pre-diabetes (Pre-DM) progression to type 2 diabetes mellitus (T2DM) or the chance of returning to normal glycemia. Methods This cohort study included 334 Pre-DM individuals (mean age of 49.4 years, and 51.5% men) who participated in the third phase of the Tehran Lipid and Glucose Study (2006–2008) and followed up for a median of 9 years. A validated food frequency questionnaire at baseline assessed usual intakes of the participants. Major dietary patterns were identified using principal component analysis. The DASH score and Mediterranean diet score (MDS) were also calculated. Multinomial logistic regression analysis was used to estimate the odds ratios (95% confidence intervals (CIs)) of developing T2DM and returning to normal glycemia in relation to dietary pattern scores. Results During the study follow-up, 39.8% progressed to T2DM, and 39.8% returned to normal glycemia. Three following major dietary patterns, including Western-style (with a higher load of red meats, hydrogenated fats, sodium, and total fat intakes), healthy pattern (with a higher load of whole grains, vegetables, and dairy products), and processed-foods pattern (with a higher load of processed-meats, fast-foods, salty snakes, and sweets and candies) were identified. The Western-style dietary pattern increased the risk of progressing to T2DM by 38% (OR = 1.38; 95% CI = 1.00 to 1.89, P = 0.050). Other dietary pattern scores were not related to regression or progression from Pre-DM. Conclusion The Western-style dietary pattern (characterized by higher load of red meats, hydrogenated fats, sodium intake, and high-GI foods) may accelerate the progression of Pre-DM to T2DM.
Diseases of the endocrine glands. Clinical endocrinology
Luiz Phellipe Dell Aquila, Armando Morales, Patricia Moreira
et al.
Introduction: In postmenopausal women, vitamin D deficiency has been associated with disability, low muscle mass and fractures. Irisin is an important myokine that may contribute to the maintenance of muscle and bone density. Vitamin D is associated with the growth and function of muscle tissue through interactions between the vitamin D receptor and PGC-1α and activation of p38/MAPK (mitogen-activated protein kinase) in muscle, a mechanism similar to irisin action. The aim of this pilot study was to evaluate the effects of cholecalciferol supplementation on serum irisin levels in sedentary postmenopausal women with hypovitaminosis D (25(OH)D < 20 ng/mL). Material and methods: 80 sedentary postmenopausal women with hypovitaminosis D and low sun exposure were supplemented with cholecalciferol (30,000 IU/month) for 12 months. Calcium, parathyroid hormone, alkaline phosphatase (AP) and irisin levels were measured before and after supplementation. Results: 25(OH) vitamin D increased in all participants. Serum levels of irisin increased (from 0.52 ± 0.27 to 0.80 ± 0.53; p < 0.003), accompanied by a decrease in AP (from 80 ± 24 to 66 ± 23; p < 0.001). Conclusions: Restoration of vitamin D status increased serum irisin levels in sedentary postmenopausal women. Whether increased serum irisin levels may have an impact on clinical outcomes deserves further evaluation.
Diseases of the endocrine glands. Clinical endocrinology
Past studies on the ICD coding problem focus on predicting clinical codes primarily based on the discharge summary. This covers only a small fraction of the notes generated during each hospital stay and leaves potential for improving performance by analysing all the available clinical notes. We propose a hierarchical transformer architecture that uses text across the entire sequence of clinical notes in each hospital stay for ICD coding, and incorporates embeddings for text metadata such as their position, time, and type of note. While using all clinical notes increases the quantity of data substantially, superconvergence can be used to reduce training costs. We evaluate the model on the MIMIC-III dataset. Our model exceeds the prior state-of-the-art when using only discharge summaries as input, and achieves further performance improvements when all clinical notes are used as input.
This dataset represents almost all the harmful diseases for rice in Bangladesh. This dataset consists of 1106 image of five harmful diseases called Brown Spot, Leaf Scaled, Rice Blast, Rice Turngo, Steath Blight in two different background variation named field background picture and white background picture. Two different background variation helps the dataset to perform more accurately so that the user can use this data for field use as well as white background for decision making. The data is collected from rice field of Dhaka Division. This dataset can use for rice leaf diseases classification, diseases detection using Computer Vision and Pattern Recognition for different rice leaf disease.
Propósito. El objetivo de esta revisión es la de conocer en el contexto actual el uso de los numerosos epónimos del hipertiroidismo del siglo XIX, y la de describir la historia correspondiente.
Contenidos. Usando las palabras clave en varias bases de datos, encontramos información entre 2012-2022. Posteriormente narramos la historia de hechos y personajes involucrados en las primeras observaciones sobre el bocio exoftálmico hipertiroideo.
Contribuciones. La utilización de los epónimos del hipertiroidismo se ha reducido notoriamente, y la tendencia actual es la de usar palabras descriptivas y un sistema de códigos que permite explicar más claramente la enfermedad que afecta a un paciente dado, cuando se trata de enviar una cuenta de cobro.
Diseases of the endocrine glands. Clinical endocrinology
Monica Agrawal, Stefan Hegselmann, Hunter Lang
et al.
A long-running goal of the clinical NLP community is the extraction of important variables trapped in clinical notes. However, roadblocks have included dataset shift from the general domain and a lack of public clinical corpora and annotations. In this work, we show that large language models, such as InstructGPT, perform well at zero- and few-shot information extraction from clinical text despite not being trained specifically for the clinical domain. Whereas text classification and generation performance have already been studied extensively in such models, here we additionally demonstrate how to leverage them to tackle a diverse set of NLP tasks which require more structured outputs, including span identification, token-level sequence classification, and relation extraction. Further, due to the dearth of available data to evaluate these systems, we introduce new datasets for benchmarking few-shot clinical information extraction based on a manual re-annotation of the CASI dataset for new tasks. On the clinical extraction tasks we studied, the GPT-3 systems significantly outperform existing zero- and few-shot baselines.
Advances in optical coherence tomography (OCT) have enabled noninvasive imaging of substructures of the human retina with high spatial resolution. OCT examinations are now a standard procedure in clinics and an integral part of ophthalmic research. The interpretation of the OCT helps ophthalmologists understand the impact of various retinal and systemic diseases on the structure of the retina in a way not previously possible. In the early stages of retinal diseases, however, the identification and analysis of small and localized substructural changes in the retina remains a challenge. We present an overview of novel visual analytics approaches for the interactive exploration of early retinal changes in single and multiple patients, the comparison of the changes with normative data, and automated quantification and measurement of diagnosis-relevant information. We developed these approaches in close collaboration with ophthalmology researchers and industry experts from a leading OCT device manufacturer. As a result, they not only significantly reduced the time and effort required for OCT data analysis, especially in the context of cross-sectional studies, but have also led to several new discoveries published in biomedical journals.
Eric H. Kim, Brett A. Lidbury, Alice M. Richardson
Background: It is not established whether clinical notes provided on pathology request forms are useful as decision support data when assessing Hepatitis B and C viral infection status. Objective: To determine sensitivity, specificity, and predictive value of clinical notes for identifying infection status of Hepatitis B and C. Methods: The study comprises 179 cases and 166 cases tested for HBsAg and anti-HCV serological markers, respectively, and accompanied by a written description (clinical note) provided on pathology request forms by the clinician on duty. The clinical note sensitivity, specificity, positive (PPV) and negative (NPV) predictive values were calculated using serological HBsAg and anti-HCV tests as gold standards. Results: The sensitivity and specificity of clinical notes for Hepatitis B infection status were 90 percent and 56 percent, respectively. The sensitivity and specificity of clinical notes for Hepatitis C infection status were 86 percent and 21 percent, respectively. Conclusions: Clinical note information identifies moderate-to-high sensitivity with regards to Hepatitis B and C viral infection status, however, given low specificity in both groups, the clinical note is not favourable for ruling disease in, possibly due to high rate of false positives.
Chromatic dispersion is a common problem to degrade the system resolution in optical coherence tomography (OCT). This study is to develop a deep learning network for automated dispersion compensation (ADC-Net) in OCT. The ADC-Net is based on a redesigned UNet architecture which employs an encoder-decoder pipeline. The input section encompasses partially compensated OCT B-scans with individual retinal layers optimized. Corresponding output is a fully compensated OCT B-scans with all retinal layers optimized. Two numeric parameters, i.e., peak signal to noise ratio (PSNR) and structural similarity index metric computed at multiple scales (MS-SSIM), were used for objective assessment of the ADC-Net performance. Comparative analysis of training models, including single, three, five, seven and nine input channels were implemented. The five-input channels implementation was observed as the optimal mode for ADC-Net training to achieve robust dispersion compensation in OCT
Obesity is characterized by low-grade chronic inflammation, which underlies insulin resistance and non-alcoholic fatty liver disease (NAFLD). Swertiamarin is a secoiridoid glycoside that has been reported to ameliorate diabetes and NAFLD in animal models. However, the effects of swertiamarin on obesity-related inflammation and insulin resistance have not been fully elucidated. Thus, this study investigated the effects of swertiamarin on inflammation and insulin resistance in high-fat diet (HFD)-induced obese mice. C57BL/6 mice were fed a HFD or HFD containing swertiamarin for 8 weeks. Obesity-induced insulin resistance and inflammation were assessed in the epididymal white adipose tissue (eWAT) and livers of the mice. Swertiamarin attenuated HFD-induced weight gain, glucose intolerance, oxidative stress, and insulin resistance, and enhanced insulin signalling in mice. Compared to HFD-fed mice, the swertiamarin-treated mice exhibited increased lipolysis and reduced adipocyte hypertrophy and macrophage infiltration in eWAT. Moreover, swertiamarin alleviated HFD-mediated hepatic steatosis and inflammation by suppressing activation of the p38 MAPK and NF-κB pathways within the eWAT and liver of obese mice. In conclusion, supplementation with swertiamarin attenuated weight gain and hepatic steatosis, and alleviated obesity-associated inflammation and insulin resistance, in obese mice.
Diseases of the endocrine glands. Clinical endocrinology, Cytology