Severe acute respiratory syndrome coronavirus 2 (SARSCoV-2) infection, which causes coronavirus disease 2019 (COVID-19), first emerged in China in December 2019 and has now spread worldwide, with a reported 351,731 confirmed cases and 15,374 deaths as of 23 March 2020 according to John Hopkins University. The infection is typically characterized by respiratory symptoms, which indicates droplet transmission. However, several case studies have reported gastrointestinal symptoms and/or evidence that some patients with SARSCoV-2 infection have viral RNA or live infectious virus present in faeces, which suggests that another possible route might be faecal–oral transmission. In a clinical characterization of ten paediatric patients with SARSCoV-2 infection in China, none of whom required respiratory support or intensive care and all of whom lacked signs of pneumonia, eight tested positive on rectal swabs, even after nasopharyngeal testing was negative. The details were published as a Brief Communication in Nature Medicine. The patients, whose ages ranged from 2 months to 15 years, initially tested positive after being screened by nasopharyngeal swab realtime reverse transcription PCR (RT–PCR). Next, the researchers conducted a series of nasopharyngeal and rectal swabs to investigate the pattern of viral excretion. Eight patients had realtime RT–PCRpositive rectal swabs. In addition, these eight patients had persistently positive rectal swabs even after their nasopharyngeal tests were negative. Four patients were discharged after two consecutive negative rectal swabs, but the rectal swabs of two of these patients later became positive again, despite nasopharyngeal tests remaining negative. Finally, the researchers used the viral RNA measurements to determine that viral shedding from the digestive system might be longerlasting than that from the respiratory tract. The findings suggest that we also need to use rectal swabs to confirm diagnosis of COVID-19, says Kang Zhang, a corresponding author of the study. There had been earlier reports, particularly in adults, of gastrointestinal symptoms and of the possibility of a faecal–oral route of transmission. In a cohort of 1,099 patients with COVID-19 from 552 hospitals in China, published in the New England Journal of Medicine, 5.0% of patients presented with nausea or vomiting and 3.8% presented with diarrhoea. Also, preliminary findings published in the American Journal of Gastroenterology found that of 204 patients with COVID-19 (mean age 54.9 years) who presented to three hospitals in China, 99 (48.5%) patients presented with digestive symptoms as their chief complaint. 60% of patients without digestive symptoms were cured and discharged, compared with 34.3% of patients with digestive symptoms. In a short Research Letter published in the Journal of the American Medical Association, different tissues of patients with COVID-19 (n = 1,070 specimens from 205 patients of mean age 44 years) were tested by RT–PCR. 32% of pharyngeal swabs (126 of 398) and 29% of faecal samples (44 of 153) tested positive. Electron microscopy of four SARSCoV-2positive faecal specimens detected live virus in stool samples from two patients who did not have diarrhoea. The precise mechanisms by which SARSCoV-2 interacts with the gastrointestinal tract remain unknown. SARSCoV-2 is thought to use ACE2 as a viral receptor, and ACE2 mRNA is highly expressed in the gastrointestinal system. In preliminary findings published in Gastroenterology, researchers examined clinical specimens from 73 hospitalized patients with SARSCoV-2 infection. 39 patients tested positive for SARSCoV-2 RNA in stool samples. In addition, 17 patients remained positive for SARSCoV-2 in stool after becoming negative in respiratory samples. Viral host receptor ACE2 stained positive mostly in gastrointestinal epithelial cells. Together, these findings have implications for our understanding of SARSCoV-2 transmission. “Asymptomatic children and adults may be shedding infectious virus and they could transmit it. This is another reason to emphasize good personal hygiene,” says Mary Estes at Baylor College of Medicine, Texas, who was not involved in these studies. “Physicians and caretakers of potentiallyinfected children need to be aware that stools might be infectious,” adds Estes. The results are preliminary and further research is needed. “We are now assembling a much larger cohort to confirm our results and will test more patients to confirm faecal–oral transmission,” says Zhang.
Plant disease diagnosis is essential to farmers' management choices because plant diseases frequently lower crop yield and product quality. For harvests to flourish and agricultural productivity to boost, grape leaf disease detection is important. The plant disease dataset contains grape leaf diseases total of 9,032 images of four classes, among them three classes are leaf diseases, and the other one is healthy leaves. After rigorous pre-processing dataset was split (70% training, 20% validation, 10% testing), and two pre-trained models were deployed: InceptionV3 and Xception. Xception shows a promising result of 96.23% accuracy, which is remarkable than InceptionV3. Adversarial Training is used for robustness, along with more transparency. Grad-CAM is integrated to confirm the leaf disease. Finally deployed a web application using Streamlit with a heatmap visualization and prediction with confidence level for robust grape leaf disease classification.
Background and aims: Noninvasive preoperative radiologic prediction of histologic grade—a key prognostic factor—is invaluable. We aim to compare the diagnostic values of 3D magnetic resonance elastography (MRE), intravoxel incoherent motion (IVIM), and conventional contrast-enhanced magnetic resonance imaging (cMRI) in predicting the histologic grade of hepatocellular carcinoma (HCC). Methods: This institutional review board-approved retrospective study included patients who underwent MRI between December 2014 and October 2021. Sixty-eight patients with pathologically confirmed HCCs who underwent MRE, IVIM, and cMRI imaging were included in the analysis. Two radiologists measured HCC stiffness volumetrically and over a single slice, and also measured apparent diffusion coefficient (ADC), IVIM-derived parameters, and enhancement ratio (ER) on arterial phase images via cMRI. Student’s t-test or the Mann–Whitney U test was used for group comparisons. Receiver operating characteristic (ROC) curve analyses were performed to evaluate the diagnostic performance. Results: Histologically, fifty-three (78%) patients had well-differentiated or moderately differentiated HCCs, and fifteen (22%) patients had poorly differentiated HCCs. Both the volumetric stiffness and single-ROI tumor stiffness were significantly elevated in the poorly differentiated HCC group (P < 0.001, P = 0.001), and the volumetric stiffness was a better measurement of stiffness because it had a higher ROC curve value (0.816). However, the ADC, the true diffusion coefficient (D), the pseudodiffusion coefficient (D∗), the pseudodiffusion fraction (f), and ER during the arterial phases on cMRI were not significantly different between the two groups (P = 0.309, 0.187, 0.440, 0.350, and 0.714, respectively). Conclusions: Stiffness measured with 3D MRE may be useful for noninvasively predicting HCC histologic grade, and the volumetric measuring method achieved the highest ROC curve value, outperforming single-ROI HCC stiffness, IVIM parameters, and arterial-phase ER on cMRI.
Diseases of the digestive system. Gastroenterology
Rare diseases represent the long tail of medical imaging, where AI models often fail due to the scarcity of representative training data. In clinical workflows, radiologists frequently consult case reports and literature when confronted with unfamiliar findings. Following this line of reasoning, we introduce RADAR, Retrieval Augmented Diagnostic Reasoning Agents, an agentic system for rare disease detection in brain MRI. Our approach uses AI agents with access to external medical knowledge by embedding both case reports and literature using sentence transformers and indexing them with FAISS to enable efficient similarity search. The agent retrieves clinically relevant evidence to guide diagnostic decision making on unseen diseases, without the need of additional training. Designed as a model-agnostic reasoning module, RADAR can be seamlessly integrated with diverse large language models, consistently improving their rare pathology recognition and interpretability. On the NOVA dataset comprising 280 distinct rare diseases, RADAR achieves up to a 10.2% performance gain, with the strongest improvements observed for open source models such as DeepSeek. Beyond accuracy, the retrieved examples provide interpretable, literature grounded explanations, highlighting retrieval-augmented reasoning as a powerful paradigm for low-prevalence conditions in medical imaging.
Recent advances in artificial intelligence (AI) and multimodal data collection are revolutionizing dermatology. Generative AI and machine learning approaches offer opportunities to enhance the diagnosis and treatment of inflammatory skin diseases, including atopic dermatitis, psoriasis, hidradenitis suppurativa, and autoimmune connective tissue disease. This review examines the current landscape of AI applications for inflammatory skin diseases and explores how generative AI and machine learning methods can advance the field through deep phenotyping, disease heterogeneity characterization, drug development, personalized medicine, and clinical care. We discuss the promises and challenges of these technologies and present a vision for their integration into clinical practice.
Anthony Hevia, Sanjana Chintalapati, Veronica Ka Wai Lai
et al.
We present ROBOTO2, an open-source, web-based platform for large language model (LLM)-assisted risk of bias (ROB) assessment of clinical trials. ROBOTO2 streamlines the traditionally labor-intensive ROB v2 (ROB2) annotation process via an interactive interface that combines PDF parsing, retrieval-augmented LLM prompting, and human-in-the-loop review. Users can upload clinical trial reports, receive preliminary answers and supporting evidence for ROB2 signaling questions, and provide real-time feedback or corrections to system suggestions. ROBOTO2 is publicly available at https://roboto2.vercel.app/, with code and data released to foster reproducibility and adoption. We construct and release a dataset of 521 pediatric clinical trial reports (8954 signaling questions with 1202 evidence passages), annotated using both manually and LLM-assisted methods, serving as a benchmark and enabling future research. Using this dataset, we benchmark ROB2 performance for 4 LLMs and provide an analysis into current model capabilities and ongoing challenges in automating this critical aspect of systematic review.
Alzheimer's disease is not the outcome of a single cause but the convergence of many. This review reframes dementia as a systemic failure, where amyloid plaques and tau tangles are not root causes but late-stage byproducts of the underlying metabolic collapse. We begin by tracing the historical merger of early- and late-onset Alzheimer's into a single disease category, a conceptual error that may have misdirected decades of research. We then synthesize evidence pointing to metabolic dysfunction - especially mitochondrial damage - as a more likely initiating event. Through this lens, we examine diverse contributing factors including type 2 diabetes, hyperglycemia-induced oxidative stress, infections and neuroinflammation. Finally, we assess current treatment limitations and argue that prevention, grounded in early metabolic and vascular interventions, holds the most promise for altering the course of this complex disease.
The nationwide claims data lake for sleep apnoea (ALASKA)—real-life data for understanding and increasing obstructive sleep apnea (OSA) quality of care study—investigated long-term continuous positive airway pressure (CPAP) termination rates, focusing on the contribution of comorbidities. The French national health insurance reimbursement system data for new CPAP users aged ≥18 years were analyzed. Innovative algorithms were used to determine the presence of specific comorbidities (hypertension, diabetes and chronic obstructive pulmonary disease (COPD)). Therapy termination was defined as cessation of CPAP reimbursements. A total of 480,000 patients were included (mean age 59.3 ± 13.6 years, 65.4% male). An amount of 50.7, 24.4 and 4.3% of patients, respectively, had hypertension, diabetes and COPD. Overall CPAP termination rates after 1, 2 and 3 years were 23.1, 37.1 and 47.7%, respectively. On multivariable analysis, age categories, female sex (1.09 (1.08–1.10) and COPD (1.12 (1.10–1.13)) and diabetes (1.18 (1.16–1.19)) were significantly associated with higher CPAP termination risk; patients with hypertension were more likely to continue using CPAP (hazard ratio 0.96 (95% confidence interval 0.95–0.97)). Therapy termination rates were highest in younger or older patients with ≥1 comorbidity. Comorbidities have an important influence on long-term CPAP continuation in patients with OSA.
Romilda Cardin, Debora Bizzaro, Francesco Paolo Russo
et al.
Background and Objective: Drug-induced liver injury (DILI) is increasingly becoming a cause of acute hepatitis. The study evaluated the role of liver-specific microRNAs (miRNAs) and keratin-18 (K-18) markers M30 (apoptosis) and M65 (necrosis) as biomarkers of acute hepatitis. Methods: Sixty-eight patients were sub-grouped as DILI, HBV- and alcohol-related acute hepatitis. Five healthy controls were included. The expression of plasma miR-21-5p, miR-34a-5p and miR-122-5p was evaluated by RT-qPCR analysis using healthy volunteers as reference. M30 and M65 were determined with ELISA kits. Results: All markers were significantly higher in the acute liver disease patients compared to controls. In DILI, miRNA levels positively correlated with M30, M65 and ALT. miR-122-5p had the highest AUC of 0.73, sensitivity of 76.2 and specificity of 72.2 in identifying DILI from other groups. Patients with hepatocellular-pattern DILI showed higher miR-122-5p and miR-21-5p compared to patients with cholestatic or mixed pattern. A new score to discriminate DILI versus other causes of acute hepatitis was developed using the identified risk factors as follows: 0.012 × miR-34a-5p + 0.012 × miR-122-5p − 0.001 × M30 + 2.642 × 1 (if mixed pattern) + 0.014 × 1 (if hepatocellular pattern) + 1.887. The AUC of the score was 0.86, with a sensitivity and specificity of 81%, better than the values of the single markers. Conclusions: Liver-specific miRNAs and K-18 could be promising serum biomarkers of DILI, especially when used in combination.
Kiyotaka Hosoda, Akira Shimizu, Koji Kubota
et al.
Abstract Aim Methods of predicting severe postoperative complications after anatomical resection for hepatocellular carcinoma are yet to be established. We aimed to clarify the relationship between inflammation‐based prognostic scores and liver fibrosis markers and the incidence of postoperative complications after anatomical resection for hepatocellular carcinoma as well as the usefulness of these markers in surgical procedure selection. Methods We included 374 patients with hepatocellular carcinoma who had undergone initial hepatectomy between January 2007 and December 2021. The association between inflammation‐based prognostic scores or liver fibrosis markers and postoperative complications was evaluated, and severe postoperative complication rates in the high‐risk group defined by these markers were compared in terms of surgical procedure. Results The advanced lung cancer inflammation index and nonalcoholic fatty liver disease fibrosis score correlated significantly with severe postoperative complications after anatomical resection, with areas under the curve of 0.67 and 0.61, respectively. The combined advanced lung cancer inflammation index and nonalcoholic fatty liver disease fibrosis score resulted in a larger area under the curve (0.69). Furthermore, in the high‐risk group determined by the combined score, the anatomical resection group had a significantly higher incidence of severe complications than the partial resection group (P < 0.01). There were no significant differences in prognosis among the surgical procedures in the high‐risk group. Conclusion The combined advanced lung cancer inflammation index and nonalcoholic fatty liver disease fibrosis score serves as a predictive marker for severe postoperative complications after anatomical resection. This combined marker may contribute to appropriate surgical procedure selection.
Surgery, Diseases of the digestive system. Gastroenterology
Abstract Objective Multiple randomized controlled trials (RCTs) have investigated the efficacy of fecal microbiota transplantation (FMT) for irritable bowel syndrome (IBS), but have yielded inconsistent results. We updated the short-term and long-term efficacy of FMT in treating IBS, and performed a first-of-its-kind exploration of the relationship between gut microbiota and emotions. Methods We conducted a comprehensive search of PubMed, Embase, Web of Science, and the Cochrane Library using various search strategies to identify all eligible studies. The inclusion criteria for data extraction were randomized controlled trials (RCTs) that investigated the efficacy of fecal microbiota transplantation (FMT) compared to placebo in adult patients (≥ 18 years old) with irritable bowel syndrome (IBS). A meta-analysis was then performed to assess the summary relative risk (RR) and corresponding 95% confidence intervals (CIs). Results Out of 3,065 potentially relevant records, a total of 10 randomized controlled trials (RCTs) involving 573 subjects met the eligibility criteria for inclusion in the meta-analysis. The meta-analyses revealed no significant differences in short-term (12 weeks) (RR 0.20, 95% CI -0.04 to 0.44), long-term (52 weeks) global improvement (RR 1.38, 95% CI 0.87 to 2.21), besides short-term (12 weeks) (SMD − 48.16, 95% CI -102.13 to 5.81, I2 = 90%) and long-term (24 weeks) (SMD 2.16, 95% CI -60.52 to 64.83, I2 = 68%) IBS-SSS. There was statistically significant difference in short-term improvement of IBS-QoL (SMD 10.11, 95% CI 0.71 to 19.51, I2 = 82%), although there was a high risk of bias. In terms of long-term improvement (24 weeks and 54 weeks), there were no significant differences between the FMT and placebo groups (SMD 7.56, 95% CI 1.60 to 13.52, I2 = 0%; SMD 6.62, 95% CI -0.85 to 14.08, I2 = 0%). Sensitivity analysis indicated that there were visible significant effects observed when the criteria were based on Rome IV criteria (RR 16.48, 95% CI 7.22 to 37.62) and Gastroscopy (RR 3.25, 95%CI 2.37 to 4.47), Colonoscopy (RR 1.42, 95% CI 0.98 to 2.05). when using mixed stool FMT based on data from two RCTs, no significant difference was observed (RR 0.94, 95% CI 0.66 to -1.34). The remission of depression exhibited no significant difference between the FMT and placebo groups at the 12-week mark (SMD − 0.26, 95% CI -3.09 to 2.58), and at 24 weeks (SMD − 2.26, 95% CI -12.96 to 8.45). Furthermore, major adverse events associated with FMT were transient and self-limiting. Discussion Based on the available randomized controlled trials (RCTs), the current evidence does not support the efficacy of FMT in improving global IBS symptoms in the long term. The differential results observed in subgroup analyses raise questions about the accurate identification of suitable populations for FMT. Further investigation is needed to better understand the reasons behind these inconsistent findings and to determine the true potential of FMT as a treatment for IBS.
Diseases of the digestive system. Gastroenterology
Dirk Nierhoff, Philipp Kasper, Tobias Goeser
et al.
Objectives The management of upper gastrointestinal bleeding (UGIB) has seen rapid advancements with revolutionising innovations. However, insufficient data exist on the necessary number of emergency endoscopies needed to achieve competency in haemostatic interventions.Design We retrospectively analysed all oesophagogastroduodenoscopies with signs of recent haemorrhage performed between 2015 and 2022 at our university hospital. A learning curve was created by plotting the number of previously performed oesophagogastroduodenoscopies with signs of recent haemorrhage against the treatment failure rate, defined as failed haemostasis, rebleeding and necessary surgical or radiological intervention.Results The study population included 787 cases with a median age of 66 years. Active bleeding was detected in 576 cases (73.2%). Treatment failure occurred in 225 (28.6%) cases. The learning curve showed a marked decline in treatment failure rates after nine oesophagogastroduodenoscopies had been performed by the respective endoscopists followed by a first plateau between 20 and 50 procedures. A second decline was observed after 51 emergency procedures followed by a second plateau. Endoscopists with experience of <10 emergency procedures had higher treatment failure rates compared with endoscopists with >51 emergency oesophagogastroduodenoscopies performed (p=0.039) or consultants (p=0.041).Conclusions Our data suggest that a minimum number of 20 oesophagogastroduodenoscopies with signs of recent haemorrhage is necessary before endoscopists should be considered proficient to perform emergency procedures independently. Endoscopists might be considered as advanced-qualified experts in managing UGIB after a minimum of 50 haemostatic procedure performed. Implementing recommendations on minimum numbers of emergency endoscopies in education programmes of endoscopy trainees could improve their confidence and competency in managing acute UGIB.
Diseases of the digestive system. Gastroenterology
Pumpkin is a vital crop cultivated globally, and its productivity is crucial for food security, especially in developing regions. Accurate and timely detection of pumpkin leaf diseases is essential to mitigate significant losses in yield and quality. Traditional methods of disease identification rely heavily on subjective judgment by farmers or experts, which can lead to inefficiencies and missed opportunities for intervention. Recent advancements in machine learning and deep learning offer promising solutions for automating and improving the accuracy of plant disease detection. This paper presents a comprehensive analysis of state-of-the-art Convolutional Neural Network (CNN) models for classifying diseases in pumpkin plant leaves. Using a publicly available dataset of 2000 highresolution images, we evaluate the performance of several CNN architectures, including ResNet, DenseNet, and EfficientNet, in recognizing five classes: healthy leaves and four common diseases downy mildew, powdery mildew, mosaic disease, and bacterial leaf spot. We fine-tuned these pretrained models and conducted hyperparameter optimization experiments. ResNet-34, DenseNet-121, and EfficientNet-B7 were identified as top-performing models, each excelling in different classes of leaf diseases. Our analysis revealed DenseNet-121 as the optimal model when considering both accuracy and computational complexity achieving an overall accuracy of 86%. This study underscores the potential of CNNs in automating disease diagnosis for pumpkin plants, offering valuable insights that can contribute to enhancing agricultural productivity and minimizing economic losses.
In response to the COVID-19 pandemic, the integration of interpretable machine learning techniques has garnered significant attention, offering transparent and understandable insights crucial for informed clinical decision making. This literature review delves into the applications of interpretable machine learning in predicting the prognosis of respiratory diseases, particularly focusing on COVID-19 and its implications for future research and clinical practice. We reviewed various machine learning models that are not only capable of incorporating existing clinical domain knowledge but also have the learning capability to explore new information from the data. These models and experiences not only aid in managing the current crisis but also hold promise for addressing future disease outbreaks. By harnessing interpretable machine learning, healthcare systems can enhance their preparedness and response capabilities, thereby improving patient outcomes and mitigating the impact of respiratory diseases in the years to come.
Plant diseases pose significant threats to agriculture. It necessitates proper diagnosis and effective treatment to safeguard crop yields. To automate the diagnosis process, image segmentation is usually adopted for precisely identifying diseased regions, thereby advancing precision agriculture. Developing robust image segmentation models for plant diseases demands high-quality annotations across numerous images. However, existing plant disease datasets typically lack segmentation labels and are often confined to controlled laboratory settings, which do not adequately reflect the complexity of natural environments. Motivated by this fact, we established PlantSeg, a large-scale segmentation dataset for plant diseases. PlantSeg distinguishes itself from existing datasets in three key aspects. (1) Annotation type: Unlike the majority of existing datasets that only contain class labels or bounding boxes, each image in PlantSeg includes detailed and high-quality segmentation masks, associated with plant types and disease names. (2) Image source: Unlike typical datasets that contain images from laboratory settings, PlantSeg primarily comprises in-the-wild plant disease images. This choice enhances the practical applicability, as the trained models can be applied for integrated disease management. (3) Scale: PlantSeg is extensive, featuring 11,400 images with disease segmentation masks and an additional 8,000 healthy plant images categorized by plant type. Extensive technical experiments validate the high quality of PlantSeg's annotations. This dataset not only allows researchers to evaluate their image classification methods but also provides a critical foundation for developing and benchmarking advanced plant disease segmentation algorithms.
The emerging research shows that lncRNAs are associated with a series of complex human diseases. However, most of the existing methods have limitations in identifying nonlinear lncRNA-disease associations (LDAs), and it remains a huge challenge to predict new LDAs. Therefore, the accurate identification of LDAs is very important for the warning and treatment of diseases. In this work, multiple sources of biomedical data are fully utilized to construct characteristics of lncRNAs and diseases, and linear and nonlinear characteristics are effectively integrated. Furthermore, a novel deep learning model based on graph attention automatic encoder is proposed, called HGATELDA. To begin with, the linear characteristics of lncRNAs and diseases are created by the miRNA-lncRNA interaction matrix and miRNA-disease interaction matrix. Following this, the nonlinear features of diseases and lncRNAs are extracted using a graph attention auto-encoder, which largely retains the critical information and effectively aggregates the neighborhood information of nodes. In the end, LDAs can be predicted by fusing the linear and nonlinear characteristics of diseases and lncRNA. The HGATELDA model achieves an impressive AUC value of 0.9692 when evaluated using a 5-fold cross-validation indicating its superior performance in comparison to several recent prediction models. Meanwhile, the effectiveness of HGATELDA in identifying novel LDAs is further demonstrated by case studies. the HGATELDA model appears to be a viable computational model for predicting LDAs.