Trifascicular block can progress to complete atrioventricular (AV) block and sudden cardiac death. Its clinical and electrocardiographic presentation is usually dramatic but may occasionally be subtle and intermittent, making recognition difficult. A 45-year-old man presented with exertional dyspnea and no structural heart disease. Surface electrocardiography revealed right bundle branch block with alternating left anterior and left posterior fascicular block and intermittent AV conduction delay, consistent with Rosenbaum syndrome and advanced His-Purkinje disease at high risk of progression to complete AV block. Anticipating this risk, a dual-chamber pacemaker was implanted using left bundle branch area pacing. At follow-up, symptoms resolved and ventricular activation normalized. Rosenbaum syndrome is a rare, potentially life-threatening conduction disorder that may present insidiously. To our knowledge, this is the first report of left bundle branch area pacing used to treat it, illustrating the potential of physiological pacing to restore conduction and prevent adverse outcomes. Trifascicular blocks carry a high risk of adverse outcomes. In rare cases, they present insidiously and intermittently. Physiological pacing via the left bundle branch area appears to be an appropriate therapeutic strategy in Rosenbaum syndrome.
Leishmaniases are infectious diseases whose vectors are female sand flies that depend on the blood of vertebrates for egg maturation. In this study, we investigated the sand fly species and DNA from Leishmania spp. This entomological/epidemiological study was conducted in a rural area in Nova Andradina, Mato Grosso do Sul, Brazil. From March 2019 to February 2020, five modified Falcão light traps were installed at night once a month. The collected insects were separated by sex into pools of 10 specimens. A sample was used for species identification via microscopic analysis, and another sample was used to detect the DNA of Leishmania in females via polymerase chain reaction (PCR). A total of 42,933 insects were collected, 77.6% (n = 33,313) of which were female. A total of 4,993 (11.6%) insects were identified, most were Nyssomyia neivai (n = 4,839; 96.9%). For PCR, 4,130 (9.6%) insects were analyzed, and four pools were positive for Leishmania spp. (with a natural infection rate of 0.1%). When the number of sand flies collected and the climatic variables were correlated, a significant difference was detected when precipitation > 60 mm (p = 0.0314) and relative humidity ≤ 60% (p = 0.0201). The analyzed area is inhabited by one of the main vectors of cutaneous leishmaniasis in Brazil, and we also found that some specimens were infected with Leishmania DNA, which might favor the occurrence of the disease in the region. This information is essential for understanding the dissemination of leishmaniasis and planning actions to prevent in the municipality.
The cross-sectional area (CSA) of the cervical spinal cord (CSC) is a key biomarker for monitoring neurodegenerative diseases such as multiple sclerosis (MS). However, its clinical utility is limited by inter-subject variability (I-SV) due to anthropometric factors (height, sex) and MRI scanner differences, which can obscure true anatomical changes. This motivates the development of a normalization strategy to enhance the sensitivity of this established biomarker to pathological changes. Existing normalization methods rely on complex imaging features unavailable in many clinical datasets, thereby motivating the need for simpler approaches. We propose a lightweight regression-based normalization model for CSC-CSA using routinely available predictors. We evaluated the effect of these predictors across cervical spinal levels (CSL), demonstrating that the model captures a quantifiable proportion of I-SV reflected by a 15-23% reduction in the coefficient of variation (COV). Prior to normalization, CSC-CSA is measured using a novel fully automated landmark-based method. It is independent of vertebral labeling and spinal curvature, and uses spinal nerve midpoints as anatomical references. Model development was conducted using 267 healthy controls (HC) from the Spine Generic dataset. For validation, two external datasets were used: the public OpenNeuro dataset, comprising 10 HC, and a private dataset consisting of 23 scans acquired from 5 patients with MS. Bidirectional stepwise regression per CSL yielded adjusted [Formula: see text] of 0.09-0.17. This approach offers a practical, generalizable framework for normalizing CSC-CSA, enabling accurate cross-sectional and longitudinal analyses in MS cohorts, as well as quantifying atrophy linked to progressive MS.
To evaluate the temporal behavior and relative variability of a margin-based metric, lesion front displacement, for quantifying geographic atrophy (GA) progression secondary to age-related macular degeneration and to compare its performance with conventional metrics. This retrospective longitudinal analysis included 103 untreated fellow eyes from the MAHALO clinical trial with GA and 18-month follow-up. Fundus autofluorescence images were obtained at baseline and 6, 12, and 18 months. Five progression metrics were calculated: total area change, square root area change, total perimeter change, perimeter-adjusted growth rate (λ'PA), and lesion front displacement. Linearity was assessed using simple linear regression and the Harvey-Collier test. Variability was evaluated using standard deviation, variance ratios, and the Fligner-Killeen test. Linearity was rejected for area and square root area change but not for lesion front displacement, perimeter change, or λ'PA. Lesion front displacement showed a relatively higher coefficient of determination (R2 = 0.504), whereas area (0.116), square root area (0.104), perimeter (0.001), and λ'PA (0.018) exhibited lower goodness of fit to a linear model. Lesion front displacement demonstrated lower dispersion across follow-up intervals. Variance ratio analysis indicated greater variability for all comparator metrics. Variability was significantly lower for lesion front displacement than for perimeter change and λ'PA at 12 and 18 months (P < 0.05). Lesion front displacement demonstrated more consistent linear approximation and lower variability compared with several conventional GA progression metrics. This margin-based approach may serve as a complementary metric for longitudinal assessment of GA progression.
As the Gulf of Maine warms, characterizing seasonal baseline levels of the human pathogen Vibrio parahaemolyticus (Vp) in Eastern oysters (Crassostrea virginica) is crucial, along with synoptic environmental data, to develop predictive capacity and risk management strategies. In Maine, USA, from May to November 2023, oyster samples were collected from Scarborough River, Maquoit Bay, and the Damariscotta River. Vp was quantified using the U.S. Food and Drug Administration-approved three-tube most-probable-number (MPN) real-time PCR method, detecting the presence of total Vp (i.e., tlh gene) and potentially pathogenic Vp (i.e., tdh and/or trh genes). tlh was detected at all sites once water temperatures reached 16°C, with densities ranging from <3 to >1,100 MPN g-1. Generalized additive models revealed significant positive relationships between both potential pathogenic markers (tdh and trh) and sea surface temperature. tdh in the Damariscotta River, the largest oyster-growing area in northern New England, was the most responsive to water temperature, and trh in Maquoit Bay was the most responsive to salinity. Statewide Vp regulations for harvesting and handling oysters were implemented in Maine in 2024, and these baseline data already demonstrate the importance of maintaining cold chain storage, as potentially pathogenic strains were detected at all three sites. Given the growing number of Vp-associated illnesses and the clear response of Vp to warming conditions in the region, these baseline data provide timely insights for harvesters and inform policies to safeguard public health.IMPORTANCEThe Gulf of Maine is both a hotspot for expanding aquaculture production and one of the fastest warming marine regions in the world. Rising numbers of vibriosis illnesses in warmer months associated with consumption of raw oysters, including Maine's first outbreak in 2023, highlight critical gaps in baseline data on the human pathogen Vibrio parahaemolyticus (Vp) (State of Maine, Department of Marine Resources, Vibrio parahaemolyticus control plan, chapter 115.02.a, 2016). The state of Maine requires up-to-date baseline information on Vp in productive growing areas to respond effectively to future illnesses and potential growing area closures. Our results support the need for a statewide Vp control plan that addresses harvest and post-harvest practices, particularly minimizing temperature abuse of live oysters. As climate change continues to warm coastal waters and alter estuarine conditions, systematic data on Vp abundance and seasonality are increasingly essential for managing the public health dimensions of shellfish aquaculture in Maine and similar high-latitude regions.
Postoperative complications after gastrectomy for gastric cancer (GC) increase mortality and delay recovery. Simple, objective tools for preoperative risk stratification using routine data remain limited. This study aimed to develop a clinically practical risk score to predict postoperative complications from standard preoperative laboratory tests. The study analyzed the data of 490 patients who underwent curative gastrectomy for GC between January 2017 and September 2025. Preoperative hematologic and biochemical parameters were systematically screened to predict postoperative complications of Clavien-Dindo grade II or higher. A composite predictive scoring system was developed, and model performance was assessed using multivariate logistic regression analysis and bootstrap internal validation. Postoperative complications of Clavien-Dindo grade II or higher occurred for 89 patients (18%). Based on discriminative performance, the C-reactive protein level, monocyte count, serum albumin level, and prothrombin time-international normalized ratio were identified as independent predictors of postoperative complications. An 8-point scoring system, termed the G-CMAP score, was constructed using odds ratio-based weighting. The G-CMAP score demonstrated modest predictive ability (area under the curve, 0.68) and remained an independent predictor in multivariate analysis (odds ratio, 3.42; 95% confidence interval, 2.08-5.61; P < 0.001). Finally, bootstrap internal validation confirmed good discrimination and calibration, with minimal overfitting. The G-CMAP score is a simple and objective blood-based tool that enables modest preoperative prediction of postoperative complications after gastrectomy for GC.
Hair is a resilient forensic element capable of preserving morphological and biochemical information and may provide relevant evidence in cases of mechanical violence involving forcible traction or avulsion. This report describes the application of trichoscopy as a complementary tool in forensic medicine for the evaluation and documentation of traumatic hair-pulling injuries. A 66-year-old woman underwent forensic medical examination after reporting a single episode of hair traction injury resulting from interpersonal violence. Clinical assessment was performed, followed by trichoscopic examination of the affected scalp area to document hair shaft and follicular alterations. Trichoscopy revealed short terminal hairs of varying lengths, black dots, yellow dots, and the presence of vellus hairs. Petechial haemorrhages were not observed, probably due to the four-week interval between the traumatic event and forensic examination. The findings were compatible with post-traumatic alopecia after forcible traction. This case suggests that trichoscopy may represent a useful, painless, non-invasive, and reproducible complementary method for the objective documentation of traumatic scalp injuries, allowing digital image recording and highlighting its potential role as an auxiliary tool in forensic medicine.
Training and body composition requirements in elite sports may elevate eating disorder (ED) risk. Current estimates suggest a range of 1-28% prevalence of ED in elite athletes, reflecting methodological heterogeneity, underreporting, and limited research focus. EDs adversely affect physical health (e.g., osteoporosis, fatigue, injury), psychological well-being, and athletic performance, often persisting beyond athletic careers. Sport-specific demands may obscure ED symptoms, making underdiagnoses likely. This longitudinal study's objective is to furnish proof for the theoretical model of disordered eating in elite athletes by examining the role of sport-specific indicators, to improve early identification and psychological diagnostics. Three-hundred elite athletes from weight-sensitive (ballet, bodybuilding) and less weight-sensitive (soccer, racket sports, basketball) sports will complete assessments at four time points. A subgroup of 90 athletes will participate in clinical interviews. Logistic and hierarchical regressions will identify ED risk indicators and estimate the prevalence of EDs in athletes. Receiver Operating Characteristic (ROC) analyses will assess diagnostic accuracy of single and combined indicators with reporting of sensitivity, specificity, and the Area Under the Curve (AUC). Group differences and measurement invariance of ED instruments will be tested between the athlete group and 300 non-athlete controls. Identifying reliable ED indicators in elite sports may support early intervention. This theory-based approach aims to enhance diagnostic accuracy and athletes' care. Prospective registration of the study in the German Clinical Trials Register (DRKS00035100) on 03 February 2025.
Diabetes and periodontal disease, a bacterial infection involving the oral microbiome, have a bidirectional relationship. However, the relationship between poor glycemic control and the oral microbiome remains unclear. The aim of this study was to assess the direct effect of intensive glycemic control on the oral microbiome in patients with type 2 diabetes mellitus (T2DM). Twenty-nine patients with T2DM and 29 controls were enrolled in this study. Patients underwent a 2-week intensive glycemic control regimen. We collected subgingival dental plaque and conducted full-length 16S ribosomal RNA gene sequencing to identify bacterial species. The focus was on the subgingival microbiome, particularly 29 species of oral bacteria related to periodontal disease, and the characteristics of the diabetic cohort. Periodontal inflammation was quantified using the periodontal inflamed surface area (PISA). Prior to diabetes treatment, the relative abundance of Porphyromonas gingivalis (P = 0.018), Tannerella forsythia (P < 0.001), Fusobacterium nucleatum (P = 0.041), and Prevotella intermedia (P < 0.001), known for their high pathogenicity in periodontal disease, was higher in the diabetes group than in the control group. The bacterial species whose relative abundance changed significantly before and after treatment differed between the PISA improved and non-improved groups in patients with diabetes. Oral bacteria changed pre- and post-diabetes treatment, correlating with improvements in PISA.
Information about on-farm goat productivity patterns is crucial for building resilient production systems in marginalised communities. Goat flock dynamics, management practices and production constraints in selected smallholder farming systems of Zimbabwe were explored. A multistage sampling approach was adopted to recruit a total of 200 participants. The effect of environment on productive indices and constraints was analysed using the Mann-Whitney U test and either Ordinal or Nominal logistic regression, respectively, in IBM SPSS version 26. Results highlight the importance of goats to rural communities. On average, 10.01 ± 5.14 goats were maintained per capita and the largest proportion was found in hot-humid environments (p > 0.05). However, reproductive performance in terms of average gestation length (4. 47 ± 0.59 months), age at first service (11.20 ± 2.24 months) and kidding (15.67 ± 2.30 months) was poorer in semi-arid environments than in the hot-humid areas (p > 0.05). Moreover, optimum productivity is hampered by a multitude of factors, including selection of breeding animals mainly based on personal appraisal (Index = 0.120), dependence of animals primarily on natural forages for feed (Index = 0.213), sub-optimal housing conditions like irregular cleaning (Index = 0.123) and disease preponderance (Index = 0.129). Ailments were mainly controlled using ethnoveterinary medicines (Index = 0.103). Taken together, our results underline the crucial role of goats to rural livelihoods, which could be elevated even more through community-based capacity building on proper animal husbandry practices in terms of breeding, nutrition, health and housing management.
Outbreaks of acute infectious conjunctivitis can spread rapidly in overcrowded communities and healthcare settings. Sudan's armed conflict has increased displacement, crowding, water insecurity, and barriers to healthcare access. A facility- and camp-based cross-sectional study included 417 consecutive patients with acute conjunctivitis or keratoconjunctivitis who presented to participating ophthalmic services and internally displaced persons' settings in Kassala State, Sudan, from 15 August 2024 to 15 December 2024. Demographic, environmental, behavioural, clinical, treatment, and follow-up data were collected using a structured form. Diagnosis was clinical; laboratory viral typing was not performed. Descriptive statistics, Spearman correlations, Mann-Whitney U tests, Kruskal-Wallis tests, and multivariable regression models were used as appropriate. Among 417 patients, 231 (55.4%) were female, and 353 (84.7%) lived in urban areas. Inadequate water/sanitation access was recorded in 125 (30.0%), a dusty environment in 224 (53.7%), a dirty environment in 317 (76.0%), a crowded environment in 223 (53.5%), and contact with a symptomatic person in 353 (84.7%). The most frequent symptoms were itching or burning (94.2%), redness (89.4%), foreign-body sensation (81.1%), eye pain (63.1%), and lid oedema (60.0%). The most frequent signs were conjunctival injection (38.8%), subconjunctival injection (35.7%), follicles/papillae (31.2%), periauricular lymphadenopathy (18.0%), punctate keratitis (11.0%), and membranes (3.8%). Clinicians recorded suspected bacterial superinfection complicating viral keratoconjunctivitis (43.9%), suspected adenoviral keratoconjunctivitis (29.3%), and suspected bacterial keratoconjunctivitis (21.3%) as the most frequent categories. Recovery data were available for 238/417 patients (57.1%); in this sub-cohort, complete recovery was associated with urban residence (OR 3.51; 95% CI 1.47-8.42), inadequate water/sanitation access (OR 9.87; 95% CI 2.46-39.62), a crowded environment (OR 2.41; 95% CI 1.12-5.14), and a lower symptom score (OR 0.78; 95% CI 0.63-0.97). All estimates are observational and are not interpreted as causal. This outbreak occurred in a war-affected setting characterized by displacement, overcrowding, and limited access to water and sanitation. The clinical pattern is consistent with acute infectious keratoconjunctivitis, but the lack of laboratory testing prevents definitive classification as adenoviral epidemic keratoconjunctivitis or as enterovirus- or coxsackievirus-associated acute hemorrhagic conjunctivitis. Future outbreak response in similar settings should include laboratory confirmation and viral typing, access to clean water, hand hygiene, separation of cases in eye-care services, and avoidance of shared personal items.
The original intravascular ultrasound (IVUS) calcium score in percutaneous coronary intervention (PCI) applies only to lesions with a maximum calcium angle > 270°, distinguishing it from the optical coherence tomography (OCT) calcium score. We aimed to investigate a new IVUS-derived calcium score to predict stent underexpansion in moderately to severely calcified lesions. We retrospectively analyzed 238 IVUS-guided PCI cases with moderate to severe calcification visible on angiography, including 136 cases with IVUS-guided PCI as a derivation cohort and 102 cases with both IVUS- and OCT-guided PCI as a validation cohort, excluding cases in which atherectomy was used. In the derivation cohort, IVUS-derived maximum calcium angle showed significant area under the curve (AUC) to predict stent underexpansion (< 70%) at the maximum superficial calcium site (cutoff value: 228°, AUC: 0.72). A multivariable linear regression model showed that maximum calcium angle of 225° (regression coefficient: -11.69; p = 0.023) and the presence of calcified nodule (regression coefficient: -4.913; p = 0.004) were associated with stent expansion. A new IVUS calcium score (0-4 points) was defined as 3 points for maximum calcium angle > 225°, and 1 point for presence of a calcified nodule. A cutoff of ≥ 3 showed a good AUC (0.751) and an excellent negative predictive value (0.963) to predict stent underexpansion. In the validation cohort, the incidence of stent underexpansion was 0%, 0%, 8.6%, and 33.3% in patients with the new IVUS calcium score of 0, 1, 3, and 4, respectively (p = 0.004). The new IVUS-derived calcium score can effectively predict stent expansion in moderately to severely calcified lesions.
Traditional Chinese medicines (TCMs) require analytical methods that can handle chemically complex matrices while addressing both efficacy-oriented constituent analysis and safety-oriented contaminant monitoring. Electrochemical sensors have attracted growing interest in this area because they offer low-cost instrumentation, short analysis time, and compatibility with portable formats, but their practical value depends on more than nominal sensitivity. This critical review evaluates recent progress in electrochemical sensing of bioactive constituents and contaminants in TCM-related matrices, with an emphasis on analytical performance, matrix adaptability, and real-sample validation. Representative constituent studies include simultaneous determination of baicalin and baicalein over 0.01-400 µM, with explicitly paired detection limits of 0.056 nM for baicalein and 0.49 nM for baicalin, and recoveries of 93.65-101.52%, as well as portable chlorogenic acid sensing with a 6.2 nM detection limit in herbal materials. Representative contaminant studies include glyphosate monitoring in TCM samples with an 8.8 × 10-10 M detection limit and 94.47-112.23% recoveries, Hg2+ sensing in Sanqi with an electrochemiluminescence detection limit of 0.27 pM, and ochratoxin A detection in medicinal-edible malt with a detection limit of 0.08 ng mL-1. Across the literature, molecularly imprinted polymers, nanocomposite interfaces, paper-based devices, and dual-mode transduction strategies have all shown analytical promise. However, matrix validation, comparator-method benchmarking, pretreatment transparency, and reproducibility reporting remain far less consistent than detection-limit reporting. The field is therefore most likely to advance not through ever-lower LODs alone, but through more standardized, application-oriented electrochemical methods for authentic TCM matrices.
A novel and highly sensitive electrochemical sensor based on an electropolymerized bromocresol green (BCG) film-modified pencil graphite electrode (poly(BCG)@PGE) was developed for the determination of tinidazole (TNDZ). The electropolymerization of BCG on the electrode surface significantly enhanced the electrochemical response toward TNDZ by improving electron-transfer kinetics and increasing the effective surface area. The fabrication and surface characteristics of the modified electrode were systematically investigated using cyclic voltammetry (CV), electrochemical impedance spectroscopy (EIS), scanning electron microscopy (SEM), energy-dispersive X-ray spectroscopy (EDX), and Fourier-transform infrared spectroscopy (FT-IR). The electrochemical behavior of TNDZ was studied using CV and differential pulse voltammetry (DPV). Key experimental parameters affecting sensor performance, including the pH of the electrolyte solution, monomer concentration, and number of electropolymerization cycles, were optimized. Under the optimized conditions, the proposed sensor exhibited a wide linear dynamic range (LDR) from 0.4 to 500 µM for TNDZ determination, with a high sensitivity of 1402 µA·mM-1·cm-2 and a low limit of detection (LOD) of 0.11 µM. The practical applicability of the developed voltammetric sensor was successfully demonstrated by the determination of TNDZ in artificial blood, pharmaceutical formulation, and spring water samples. The obtained results showed satisfactory recoveries with high precision and accuracy, confirming the reliability and potential applicability of the poly(BCG)@PGE sensor for sensitive electrochemical monitoring of TNDZ in complex matrices.
Severe asthma is biologically heterogeneous, and airway-derived biomarkers for disease stratification remain limited. We aimed to identify a mitochondria-related induced sputum transcriptomic signature associated with severe asthma. Two public induced sputum microarray datasets were analyzed: GSE76262 as the discovery cohort and GSE147880 as the external validation cohort. Mitochondrial activity was quantified by single-sample gene set enrichment analysis using the MitoCarta 3.0 gene set. In differential expression analysis, severe asthma was compared against a combined reference group of non-severe asthma and healthy controls (i.e., non-severe asthma and healthy controls combined into a single reference group). Weighted gene co-expression network analysis, and LASSO logistic regression were subsequently used to derive a mitochondria-related gene signature. External validation, multivariable logistic regression, immune infiltration analysis, and nomogram construction were further performed. Severe asthma showed broad transcriptomic remodeling, including 2,000 differentially expressed genes, of which 196 were mitochondria-related. Mito Score was significantly reduced in severe asthma and discriminated severe disease with an area under the curve (AUC) of 0.791. A key co-expression module associated with Mito Score yielded candidate genes enriched in oxidative phosphorylation, thermogenesis, cellular respiration, and reactive oxygen species-related pathways. LASSO identified a 9-gene signature consisting of CLC, IL18R1, SLC7A11, MARCH1, RBMS2, MAN2B2, SMOX, TTC15, and CD4. The signature showed good discrimination in the discovery cohort (AUC 0.892) and retained performance in the external validation cohort (AUC 0.758). Risk Score remained independently associated with severe asthma after adjustment for age and sex. Induced sputum transcriptomics identified a reproducible mitochondria-related 9-gene signature associated with severe asthma, supporting mitochondrial dysfunction as a clinically relevant airway feature and a candidate biomarker framework for disease stratification. Severe asthma showed reduced mitochondrial activity in induced sputum transcriptomes.A mitochondria-related 9-gene signature discriminated severe asthma with good accuracy.The gene panel retained discriminatory value in an independent external cohort.The signature was linked to eosinophilic, neutrophilic, and T-cell immune remodeling.
To assess the correlation between photoreceptor mosaic integrity in chronic central serous chorioretinopathy (cCSCR) and multiple morpho-functional parameters, including photoreceptor density (PD), normalized reflectivity of photoreceptor layers (NR-PRL), normalized reflectivity of the retinal pigment epithelium (NR-RPE), choriocapillaris vascular perfusion density (CC-VPD), and retinal sensitivity (RS). This observational cross-sectional study was conducted between June 2023 and September 2024. Patients with cCSCR and healthy controls underwent imaging with Spectralis High Magnification Module (HMM), structural OCT, and OCT-Angiography in the macular area. Regions of interest were selected and topographically superimposed across modalities. Retinal sensitivity was assessed using fundus microperimetry with a 4-2-1 automated staircase strategy. Mean photoreceptor density was significantly lower in cCSCR eyes than in healthy controls (6617.8 vs. 11162.3 cells/mm², p = 0.0004). Mean NR-PRL was 0.6 in cCSCR and 0.9 in controls (p = 0.0074), while NR-RPE was also significantly reduced in cCSCR (p = 0.0465). Mean CC-VPD was 41.8% in cCSCR compared with 63.7% in controls (p = 0.0011). Mean retinal sensitivity was significantly decreased in cCSCR (10.74 ± 3.90 dB) versus controls (19.39 ± 3.90 dB; p = 0.0001) and strongly correlated with photoreceptor density. A strong correlation between NR-PRL and NR-RPE was observed in both controls (r = 0.883, p = 0.014) and cCSCR eyes (r = 0.884, p = 0.015). Photoreceptor mosaic integrity significantly correlates with retinal sensitivity in cCSCR. HMM represents a reliable and accessible in vivo tool for quantifying photoreceptor loss and its functional impact.
We developed and evaluated a deep learning (DL) model for image-based detection of acute pancreatitis (AP) on abdominal contrast-enhanced CT (CECT). A total of 552 patients from two university centers (January 2010-January 2026) were included. The internal dataset comprised 207 patients with clinically and radiologically confirmed AP (499 scans) and 250 control patients with suspected AP (368 scans). An independent external validation cohort included 95 patients. Convolutional neural network-based models were trained using monophasic and biphasic CECT data. The final model was evaluated on a 20% patient-level hold-out test set from the internal cohort and on the external cohort. Performance was assessed using the F1 score and area under the receiver operating characteristic curve (AUROC). A single-input multiphase model incorporating arterial and portal venous phase scans achieved the best performance, with ensembling applied for biphasic studies. On the internal hold-out test set (n = 116), the model achieved an F1 score of 0.83 (95% confidence interval 0.75-0.89) and an AUROC of 0.89 (0.82-0.95). Performance remained robust on external validation (n = 95), with an AUROC of 0.99 (0.96-1.00) and an F1 score of 0.92 (0.86-0.97). DL enabled accurate CECT-based identification of AP in this retrospective multicenter cohort, with performance maintained in an independent external dataset. Prospective validation using broader and independently adjudicated clinical populations remains necessary. The model showed promising performance for CECT-based acute pancreatitis detection but was not designed or tested as a triage system. Diagnostic uncertainty in acute pancreatitis often arises from nonspecific abdominal symptoms and inter-reader variability in CECT interpretation. The DL model achieved high internal accuracy (AUROC 0.89) and maintained robust performance in an independent external validation cohort (AUROC 0.99).
Developing professional expertise warrants deliberate practice, where skills are developed and refined through repeated engagement with targeted tasks. Deliberate practice is difficult to implement and rarely available in the field of surgery. Digital twins, which are virtual replications of anatomy and behavior, show promise in facilitating deliberate practice and addressing the limitations presented by augmented and virtual reality systems. In this work, we investigated CholeCoach, a scalable surgical education platform for laparoscopic cholecystectomy in teaching retraction through comparison between trainee and expert surgeons and gaining their end-user experience. Digital twin simulations were generated using a machine learning pipeline from six laparoscopic cholecystectomy videos. Surgical trainees and experts were asked to virtually retract the gallbladder to expose target areas of dissection (15 tasks, 5 cases). Retraction vector data were collected and compared between groups. A post-study 5-point Likert scale survey was used to evaluate the educational value, usability, and design of the platform. Eighteen participants (11 trainees, 7 experts) engaged with the CholeCoach platform. On average, trainees pulled with 11% [95% CI 4-19%, p = 0.035] less force, had 1.8 [95% CI 1.5-2.2, p < 0.001] times higher within-task variability in force, and covered 31% [95% CI 16-48%, p = 0.004] larger gallbladder surface area than experts. Qualitative assessment of the platform through surveys was positive. Most participants found the platform to be a useful adjunct to current surgical training, with potential to improve trainee performance. Participants suggested additional functionalities and automated comparison with expert retraction. This study revealed significant discrepancy between expert and trainee approaches to retraction, objectively demonstrating the potential of CholeCoach in enabling deliberate practice and automated assessment. Qualitative assessment was positive and will be used to inform further development of CholeCoach. Digital twin-based simulation has potential to facilitate surgical training and democratize high-quality surgical education across healthcare institutions.
Stroke remains a leading cause of mortality and long-term disability worldwide, while currently used clinical risk prediction tools demonstrate only moderate accuracy. Retinal imaging combined with artificial intelligence has emerged as a promising non-invasive approach for improving stroke risk stratification by capturing microvascular changes that reflect cerebrovascular health. This scoping review was conducted in accordance with Joanna Briggs Institute methodology and reported following the Preferred Reporting Items for Systematic Reviews and Meta-Analyses Extension for Scoping Reviews guidelines. Electronic searches were performed in PubMed/MEDLINE, Scopus, and Web of Science to identify studies applying artificial intelligence techniques to retinal imaging for stroke risk prediction. Quantitative human studies reporting predictive performance metrics were eligible for inclusion. Eleven studies were included. Most investigations applied deep learning models to fundus photography or optical coherence tomography angiography. Reported predictive performance varied substantially, with area under the receiver operating characteristic curve values ranging from 0.719 to 0.987. Retinal vascular geometry, morphology, and network complexity were the most consistently identified predictors of stroke risk. Models integrating retinal imaging with clinical variables generally demonstrated improved discrimination compared with image-only approaches. However, the majority of studies relied on internal validation, and more than two-thirds were judged to have a high risk of bias. Several studies were based on large shared datasets, particularly the UK Biobank, which may limit interpretation of the overall sample size. Overall, artificial intelligence-based retinal imaging shows considerable potential as a complementary tool for stroke risk prediction, but the current evidence is limited by methodological heterogeneity, insufficient external validation, and concerns regarding generalizability. Robust prospective and externally validated studies are required before routine clinical implementation can be recommended.
Therapeutic drug monitoring (TDM) may optimize dosing of mycophenolate mofetil (MMF) in patients with nephrotic syndrome, but it is not routinely implemented. When performed, TDM typically relies on total mycophenolic acid (MPA) concentrations, although only unbound MPA is pharmacologically active. This study aimed to develop a population pharmacokinetic (PK) model characterizing the PK and pharmacodynamics (PD) of total and unbound MPA in plasma and saliva of paediatric nephrotic syndrome patients following oral MMF. PK/PD data from 62 children (47 full and 21 limited sampling curves) were analysed using nonlinear mixed-effects modelling. A two-compartment model with first-order absorption best describes unbound MPA PK. Protein binding was captured using a nonlinear saturable binding equation. The final population PK parameter estimates (for unbound MPA) were absorption rate constant (ka), 1.81 h-1 (relative standard error (RSE): 8%); apparent clearance (Cl/F), 2550 L h-1 (RSE: 8%); apparent central volume (Vc/F), 1590 L (RSE: 27%); apparent peripheral volume (Vp/F), 239,000 L (RSE: 52%); apparent intercompartmental clearance (Q/F), 1940 L h-1 (RSE: 13%); maximal binding capacity (Bmax), 196 µmol (RSE: 44%); dissociation constant (Kd), 0.75 µmol L-1 (RSE: 49%). Saliva and unbound plasma MPA showed a proportional relationship (saliva/plasma ratio: 1.34, RSE: 24%). The PK-PD relationship between unbound MPA and inosine monophosphate dehydrogenase (IMPDH) activity was described by a sigmoid inhibitory effect model with a baseline parameter. The population estimates were baseline effect, 66.4 (RSE: 21%); unbound plasma concentration for 50% inhibition, 0.0251 (RSE: 27%); shape parameter, 0.559 (RSE:22%). A limited sampling strategy allowed acceptable prediction of unbound MPA area under the concentration-time curve from 0 to 12 hours (AUC0-12 h) but not IMPDH activity. The developed model provides a foundation for model-informed precision dosing of MMF in children with nephrotic syndrome.