课程文献中心 更多>>

关键词： Artificial intelligence   technology adoption   computer-aided diagnosis   ophthalmology  

摘要： AI-based computer-aided diagnosis (AI-CAD) systems are transforming medical imaging by augmenting clinicians in disease identification and diagnosis. Nonetheless, little is known about how individual differences, particularly clinicians' expertise, affect their perception, trust, and adoption of such systems. Guided by the Elaborated Likelihood Model (ELM), this study systematically compared Task Experts (TEs;retina specialists;n = 38) and Task Non-Experts (TNs;general ophthalmologists;n = 23). TNs reported higher scores than TEs across all adoption metrics, including perceived accuracy, interpretability, credibility, ease of use, usefulness, and intention to use. For further investigation of underlying cognitive processes, PLS-SEM was conducted. It revealed that perceived usefulness was the sole direct predictor of intention to use in both groups, yet its antecedents differed by expertise. Perceived accuracy and interpretability strongly influenced TEs, reflecting central-route processing, whereas AI optimism shaped TNs' attitude, reflecting peripheral-route processing. These findings highlight the need for considering clinicians' expertise levels in AI-CAD design.

关键词： breast cancer diagnosis   object detection   mammogram   YOLOv5   YOLOv8   YOLOv10   YOLOv11   computer-aided detection   artificial intelligence   CAD  

摘要： Background/Objectives: Breast cancer persists as a leading cause of female mortality globally. Mammograms are a key screening tool for early detection, although many resource-limited hospitals lack access to skilled radiologists and advanced diagnostic tools. Deep learning-based computer-aided detection (CAD) systems can assist radiologists by automating lesion detection and classification. This study investigates the performance of various You Only Look Once (YOLO) models and a Hybrid Convolutional-Transformer Architecture (YOLOv5, YOLOv8, YOLOv10, YOLOv11, and Real-Time-DEtection Transformer (RT-DETR)) for detecting mammographic lesions on an affordable embedded system. Methods: We developed a custom web-based annotation tool to enhance mammogram labeling accuracy, using a dataset of 3169 patients from Thailand and expert annotations from three radiologists. Lesions were classified into six categories: Masses Benign (MB), Calcifications Benign (CB), Associated Features Benign (AFB), Masses Malignant (MM), Calcifications Malignant (CM), and Associated Features Malignant (AFM). Results: Our results show that the YOLOv11n model is the optimal choice for the NVIDIA Jetson Nano, achieving an accuracy of 0.86 and an inference speed of 6.16 +/- 0.31 frames per second. A comparative analysis with a graphics processing unit (GPU)-powered system revealed that the Jetson Nano achieves comparable detection performance at a fraction of the cost. Conclusions: The current research landscape has not yet integrated advanced YOLO versions for embedded deployment in mammography. This method could facilitate screening in clinics without high-end workstations, demonstrating the feasibility of deploying CAD systems in low-resource environments and underscoring its potential for real-world clinical applications.

关键词： bevel   butt joint   CAM unit   chamfer   composite resin   lithium disilicate   onlay  

摘要： Objective: To evaluate the marginal adaptation and margin trueness of lithium disilicate glass-ceramic (LDS) and glass-filler reinforced composite resin (CR) onlays fabricated using three different finish lines and two chairside CAM units (4- and 5-axis).
Methods: LDS (*** CAD) and CR (Tetric CAD) onlays were fabricated on butt joint, chamfer, and bevel finish lines using 4-axis (CEREC MC XL) and 5-axis (Programill One) chairside CAM units (N = 120). Onlays were scanned for margin trueness and marginal adaptation analyses. Data were analyzed with three-way ANOVA and Tukey B post hoc test (alpha = 0.05). Results: A significant interaction was found among finish line, material, and CAM unit for marginal adaptation and margin trueness (p < 0.001). Bevel had the highest marginal adaptation;chamfer the lowest. Butt joint showed the highest margin trueness;bevel the lowest. CR onlays had better marginal adaptation than LDS with bevel finish lines, regardless of CAM unit. Overall, 5-axis CAM units yielded better margin trueness. Conclusions: Finish line design, restorative material, and CAM unit affected the marginal adaptation and margin trueness. The bevel finish line yielded the highest marginal adaptation and the chamfer the lowest, regardless of material or CAM unit. Conversely, the butt joint showed the highest margin trueness, and the bevel the lowest.

关键词： automated mechanical design   design optimization   finite element analysis   automated mechanical design   computer-aided design   design optimization   finite element analysis   large language model  

摘要： Mechanical design workflow typically involves iterative cycles of computer-aided design (CAD) modeling, finite element analysis, and optimization, each requiring significant expertise and manual effort. Recent advances in large language models (LLMs) have enabled the automation of individual stages, including CAD modeling and finite element analysis. Moreover, several studies have utilized LLMs for mechanical design. However, fully integrated end-to-end automation remains limited. Therefore, this study proposes a novel end-to-end framework based on LLM agents that achieves full automation of the entire design-analysis-optimization workflow. Driven entirely by natural language inputs, the framework integrates automatic parameter extraction, CAD modeling, meshing, structural analysis, and optimization. The proposed framework performs structural analyses and achieves optimization goals while preserving design constraints. This is demonstrated through four case studies: a cantilever beam, two-section bar, desk, and flanged pipe. Notably, the proposed approach preserves critical design parameters that are often implicit, thereby mimicking the decision-making of experienced engineers. These results demonstrate the feasibility of democratizing mechanical design by enabling non-experts to perform sophisticated tasks. Although the current implementation is confined to linear analysis and exhibits reduced robustness in highly complex scenarios, this work provides a promising foundation for AI-driven automation in engineering design.

关键词： Breast Neoplasms   Digital Mammography   Diagnosis (Computer-Assisted)   Artificial Intelligence  

摘要： ObjectiveA lthough artificial intelligence-based computer-aided diagnosis (AI-CAD) is increasingly applied in screening mammography, its use in diagnostic settings is less established. This study evaluated the diagnostic performance of AI-CAD abnormality scores at optimized thresholds across various diagnostic indications vs radiologists. Materials and methods This retrospective study included 1534 women (mean age, 51.4 +/- 8.8 years) who underwent diagnostic mammography between March 2015 and February 2016. Cases were categorized into three diagnostic indications: (1) symptomatic, (2) BI-RADS 3 follow-up, and (3) referral for abnormal imaging. A commercially available AI-CAD system provided abnormality scores (0-100%). Final diagnosis was confirmed by pathology or >= 2-year imaging stability. AI-CAD performance (sensitivity, specificity, accuracy, PPV, and AUC was evaluated at two thresholds: vendor-recommended 10% for screening and an optimized 50% from ROC analysis (Youden's index), and compared with original radiologist interpretations. Results Among the 1534 patients, 397 (25.9%) were diagnosed with breast cancer. At the 50% threshold, AI-CAD showed significantly higher specificity (95.0% vs 86.2%), accuracy (91.7% vs 87.2%), and PPV (85.1% vs 69.5%) than radiologists (all p < 0.001). AUCs were comparable (AI-CAD: 0.886;radiologists: 0.882;p = 0.75). In symptomatic patients, AUC was significantly higher than radiologists (0.873 vs 0.815, p = 0.002);in BI-RADS 3 follow-ups and asymptomatic imaging-detected abnormalities, specificity was improved with a tradeoff in lower sensitivity. Conclusion AI-CAD demonstrated diagnostic performance comparable to radiologists in mammography and, at an optimized threshold, offered superior specificity, PPV, and accuracy. Especially in symptomatic patients, a higher threshold increased diagnostic performance without compromising *** PointsQuestionAI-CAD has the potential to be applied for diagnostic mammography by a

关键词： cardiometabolic disease   gut-brain-heart axis , heart-brain-metabolism axis   vascular cognitive impairment and dementia   Aβ   amyloid-β   ACS   acute coronary syndrome   AD   Alzheimer’s disease   AF   atrial fibrillation   ANS   autonomic nervous system   BBB   blood-brain barrier   BP   blood pressure   CAD   coronary artery disease   CMR   cardiac magnetic resonance imaging   CNS   central nervous system   CVD   cardiovascular disease   ECG   electrocardiographic   GLP-1   glucagon-like peptide-1   HBA   heart-brain axis   HBMA   heart-brain-metabolism axis   LA   left atrial   LV   left ventricular   MACE   major cardiovascular events   MRI   magnetic resonance imaging   PET   positron emission tomography   RAGE   receptor for advanced glycation end-products   SBI   silent brain infarction   SCD   sudden cardiac death   SGLT2   sodium-glucose cotransporter 2   TCM   takotsubo cardiomyopathy   VCID   vascular cognitive impairment and dementia  

摘要： The heart-brain axis has been widely acknowledged in medicine, but arguably it has not been particularly well understood until relatively recent times. For example, we have only recently come to appreciate the profound impact of the HBA on the development of vascular cognitive impairment and dementia (VCID). The complexity of these relationships is further reinforced by appreciating additional autonomic interconnectedness via the gut and metabolic pathways, as well as the spleen and immune pathways, in the form of the heart-brain-metabolism axis. Furthermore, the directionality of these pathways is complex, and all of these elements exert unique effects on the other. This multifaceted system is susceptible to pathologic processes involving multiple organ systems, namely, but not exclusively, the heart, vasculature, brain, autonomic nervous system, gut, and liver, with common clinical outcomes including VCID and cognitive frailty. In this state-of-the-art review, we explore the anatomic and physiologic interconnectedness between the heart, brain, gut, and metabolic systems. A focus has been placed on how relevant pathologic processes affect the cardiovascular and neurologic organ systems, as well as specific aspects of metabolism and the influence of the immune and gastrointestinal systems, all of which can contribute to both nervous system and cardiac dysfunction. Where relevant, we describe how therapeutic efforts should be focused on a preventative approach, with early identification of relevant cardiovascular and neurologic factors, to allow for the timely introduction of therapeutic measures to mitigate the risk of developing disease. Emerging approaches to cognitive risk prediction in cardiovascular disease include clinically accessible tools for older adults, brain imaging biomarkers linking stress-related neural activity to future cardiovascular events, and novel strategies such as gut microbiome modulation. Evidence supports the cognitive safety and potentia

关键词： Deep Learning   Graph Neural Network   Athlete Movements   Forecast   Dynamic CAD Model  

摘要： High-level athletes need to complete body posture detection during training to promptly identify errors in their movements, reduce mistakes, and improve training performance. Therefore, this paper proposes an automatic method for recognizing and generating athletes' movement postures based on deep learning algorithms and dynamic computer-aided design (CAD) models. The method extracts training movement images through three-dimensional image detection and analyzes movement changes by integrating constraint points and parameters. A deep learning neural network is used to construct a human action recognition model, incorporating attention mechanisms and athlete trajectories to complete prediction tasks. Finally, dynamic CAD models are used to generate athlete training movements to help athletes detect their posture. The research shows that deep learning algorithms and dynamic CAD models are effective in athlete movement recognition and prediction, helping to optimize prediction results and providing valuable action reconstruction and guidance for athletes.

关键词： cMyC   myocardial infarction   troponin   AMI   acute myocardial infarction   AUC   area under the curve   CAD   coronary artery disease   cMyC   cardiac myosin-binding protein C   ED   emergency department   ESC   European Society of Cardiology   hs-cTn   high-sensitivity cardiac troponin   hs-cTnI   high-sensitivity cardiac troponin I   hs-cTnT   high-sensitivity cardiac troponin T   NPV   negative predictive value   NSTEMI   non–ST-segment elevation myocardial infarction   PPV   positive predictive value  

摘要： Cardiac myosin-binding protein C (cMyC) is a cardiac-specific sarcomeric protein with faster release kinetics compared with those of high-sensitivity cardiac troponin (hs-cTn).

关键词： deep learning   computed tomography   craniofacial fractures   pediatric patients   Grad-CAM   2.5D models   ResNet-18  

摘要： Pediatric craniofacial fractures represent a significant diagnostic challenge due to age-dependent anatomical variability, subtle fracture lines, and the limited availability of specialists during emergency care. Computed tomography (CT) remains the gold standard for fracture assessment;however, small or non-displaced fractures may be easily overlooked. Advances in artificial intelligence offer opportunities for automated support in injury detection. In this study, we developed a 2.5D deep learning model that uses multiplanar CT reconstructions (axial, coronal, sagittal) as a three-channel input to a ResNet-18 network with transfer learning. A total of 63 high-quality CT examinations were included and split into a training set (n = 44) and a validation set (n = 19). The preprocessing pipeline included isotropic resampling, bone-window normalization, and the generation of 384 x 384 images. Model performance was evaluated using AUCROC, AUCPR, sensitivity, specificity, precision, F1-score, and threshold optimization using Youden's index. The proposed model achieved AUCROC = 0.822 and AUCPR = 0.855. At the optimal decision threshold (t = 0.54), sensitivity reached 0.556, specificity 1.000, and F1-score 0.714. Grad-CAM visualizations correctly highlighted fracture-related structures, which was confirmed in an independent clinical assessment. The results demonstrate the potential of the proposed model as a decision-support tool for pediatric craniofacial trauma, providing a foundation for future extensions toward radiomics and hybrid deep-radiomic pipelines.

关键词： coronary artery disease   heart failure   high-risk PCI   ischemic heart disease   mechanical circulatory support   percutaneous coronary intervention   ACS   acute coronary syndrome(s)   BCIS   British Cardiovascular Intervention Society   CABG   coronary artery bypass grafting   CAD   coronary artery disease   CV   cardiovascular   GDMT   guideline directed medical therapy   HF   heart failure   IABP   intra-aortic balloon pump   iLVD   ischemic left ventricular dysfunction   LV   left ventricular   LVEF   left ventricular ejection fraction   MCS   mechanical circulatory support   MI   myocardial infarction   PCI   percutaneous coronary intervention   pLVAD   percutaneous left ventricular unloading device   VA-ECMO   veno-arterial extracorporeal membrane oxygenation  

摘要： Ischemic left ventricular dysfunction (iLVD) is the most prevalent cause of heart failure with reduced ejection fraction and is associated with higher mortality than nonischemic cardiomyopathy. Coronary revascularization was once considered routine for patients with iLVD, yet recent randomized trials have questioned this approach. This state-of-the-art review examines the pathophysiology of iLVD, explores the randomized evidence around coronary revascularization, and outlines which patients with iLVD should still be offered coronary artery bypass grafting or percutaneous coronary intervention (PCI). Finally, a step-by-step guide for planning and performing PCI in this population is provided. Attention is given to shared decision-making, nontechnical factors, and considerations for achieving complete revascularization, with a specific focus on upcoming trials of mechanical circulatory support.