Search

Abstract Background The screening process for systematic reviews and meta-analyses in medical research is a labor-intensive and time-consuming task. While machine learning and deep learning have been applied to facilitate this process, these methods often require training data and user annotation. This study aims to assess the efficacy of ChatGPT, a large language model based on the Generative Pretrained Transformers (GPT) architecture, in automating the screening process for systematic reviews in radiology without the need for training data. Methods A prospective simulation study was conducted between May 2nd and 24th, 2023, comparing ChatGPT’s performance in screening abstracts against that of general physicians (GPs). A total of 1198 abstracts across three subfields of radiology were evaluated. Metrics such as sensitivity, specificity, positive and negative predictive values (PPV and NPV), workload saving, and others were employed. Statistical analyses included the Kappa coefficient for inter-rater agreement, ROC curve plotting, AUC calculation, and bootstrapping for p-values and confidence intervals. Results ChatGPT completed the screening process within an hour, while GPs took an average of 7–10 days. The AI model achieved a sensitivity of 95% and an NPV of 99%, slightly outperforming the GPs’ sensitive consensus (i.e., including records if at least one person includes them). It also exhibited remarkably low false negative counts and high workload savings, ranging from 40 to 83%. However, ChatGPT had lower specificity and PPV compared to human raters. The average Kappa agreement between ChatGPT and other raters was 0.27. Conclusions ChatGPT shows promise in automating the article screening phase of systematic reviews, achieving high sensitivity and workload savings. While not entirely replacing human expertise, it could serve as an efficient first-line screening tool, particularly in reducing the burden on human resources. Further studies are needed to fine-tune its capabilities and validate its utility across different medical subfields.

Abstract Background and Aims This study examines the relationships between epicardial adipose tissue (EAT), nonalcoholic fatty liver disease (NAFLD), and coronary artery calcium score (CACS) using non‐ECG‐gated CT scans. It aims to determine the effectiveness of EAT measurements and NAFLD as predictors for coronary artery disease (CAD). Methods This cross‐sectional study was conducted at a specialized center, focusing on individuals who underwent non‐ECG‐gated chest CT scans without contrast. We evaluated EAT thickness and density in three areas: the right atrioventricular groove, the free wall of the right ventricle, and the central area of the right anterior interventricular groove. Additionally, we measured CACS and determined the presence of NAFLD by comparing liver‐to‐spleen density ratios. Statistical analyses, including regression models, were performed using SPSS (version 26). Results In this study, we enrolled 365 participants, including 203 males with an average age of 47 ± 17.93 years. EAT thickness was 6.28 ± 1.62 mm, and EAT density was −96.07 ± 12.47 Hounsfield units (HU). The mean CACS was 22.27 ± 79.01, and the mean liver density was 50.01 ± 10.76 HU. A significant positive correlation was observed between EAT thickness and CACS (r = 0.208, p

Background Stroke, a major global health issue, is broadly categorized into ischemic and hemorrhagic types. The volume of hemorrhage on noncontrast computed tomography guides the treatment options and hints at prognosis. Conventional approaches to calculate intracerebral hemorrhage (ICH) volume, like the ABC/2 method, typically rely on an assumed standard shape and might be inaccurate. Advances in deep learning have significantly improved noncontrast computed tomography's capabilities in ICH volume estimation. This study conducts a comprehensive systematic review and meta‐analysis to evaluate the precision of deep learning algorithms in delineating ICH on noncontrast computed tomography. Methods A systematic review and meta‐analysis, adhering to Preferred Reporting Items for Systematic Reviews and Meta‐Analyses guidelines, was conducted on literature from 2000 to October 2023. Studies were selected on the basis of strict inclusion and exclusion criteria. Performance evaluation was done using the Dice Similarity Coefficient, and the Prediction Model Risk of Bias Assessment Tool was used for quality assessment. Statistical analysis was carried out using Stata 17.0. Results The review included 28 studies, mainly retrospective cohorts, with a focus on convolutional neural network architectures, particularly U‐Net variants. A meta‐analysis of 14 studies revealed a combined Dice Similarity Coefficient of 0.85 (95% CI, 0.82–0.88). Performance was consistent across various methodologies but varied on the basis of ICH pathogenesis, with spontaneous ICH having higher accuracy. Conclusion Deep learning models are highly effective in segmenting ICH on noncontrast computed tomography, demonstrating potential improvements in clinical neuroimaging. Despite their efficacy, challenges in segmenting smaller hemorrhages remain. The findings suggest that deep learning could reduce health care professional workloads and enhance patient care, although further research is needed to address limitations and extend clinical utility.

Introduction: Detection of paroxysmal atrial fibrillation (PAF) is crucial for secondary prevention in patients with recent strokes of unknown etiology. This systematic review and meta-analysis assess the predictive power of available risk scores for detecting new PAF after acute ischemic stroke (AIS). Methods: PubMed, Embase, Scopus, and Web of Science databases were searched until September 2023 to identify relevant studies. A bivariate random effects meta-analysis model pooled data on sensitivity, specificity, and area under the curve (AUC) for each score. The QUADAS-2 tool was used for the quality assessment. Results: Eventually, 21 studies with 18 original risk scores were identified. Age, left atrial enlargement, and NIHSS score were the most common predictive factors, respectively. Seven risk scores were meta-analyzed, with iPAB showing the highest pooled sensitivity and AUC (sensitivity: 89.4%, specificity: 74.2%, AUC: 0.83), and HAVOC having the highest pooled specificity (sensitivity: 46.3%, specificity: 82.0%, AUC: 0.82). Altogether, seven risk scores displayed good discriminatory power (AUC ≥0.80) with four of them (HAVOC, iPAB, Fujii, and MVP scores) being externally validated. Conclusion: Available risk scores demonstrate moderate to good predictive accuracy and can help identify patients who would benefit from extended cardiac monitoring after AIS. External validation is essential before widespread clinical adoption.

Abstract Background To assess the efficacy of artificial intelligence (AI) models in diagnosing and prognosticating acute appendicitis (AA) in adult patients compared to traditional methods. AA is a common cause of emergency department visits and abdominal surgeries. It is typically diagnosed through clinical assessments, laboratory tests, and imaging studies. However, traditional diagnostic methods can be time-consuming and inaccurate. Machine learning models have shown promise in improving diagnostic accuracy and predicting outcomes. Main body A systematic review following the PRISMA guidelines was conducted, searching PubMed, Embase, Scopus, and Web of Science databases. Studies were evaluated for risk of bias using the Prediction Model Risk of Bias Assessment Tool. Data points extracted included model type, input features, validation strategies, and key performance metrics. Results In total, 29 studies were analyzed, out of which 21 focused on diagnosis, seven on prognosis, and one on both. Artificial neural networks (ANNs) were the most commonly employed algorithm for diagnosis. Both ANN and logistic regression were also widely used for categorizing types of AA. ANNs showed high performance in most cases, with accuracy rates often exceeding 80% and AUC values peaking at 0.985. The models also demonstrated promising results in predicting postoperative outcomes such as sepsis risk and ICU admission. Risk of bias was identified in a majority of studies, with selection bias and lack of internal validation being the most common issues. Conclusion AI algorithms demonstrate significant promise in diagnosing and prognosticating AA, often surpassing traditional methods and clinical scores such as the Alvarado scoring system in terms of speed and accuracy.

Abstract Background Irritable bowel syndrome (IBS) is a common chronic functional gastrointestinal disorder associated with a wide range of clinical symptoms. Some researchers have used cluster analysis (CA), a group of non-supervised learning methods that identifies homogenous clusters within different entities based on their similarity. Objective and methods This literature review aims to identify published articles that apply CA to IBS patients. We searched relevant keywords in PubMed, Embase, Web of Science, and Scopus. We reviewed studies in terms of the selected variables, participants’ characteristics, data collection, methodology, number of clusters, clusters’ profiles, and results. Results Among the 14 articles focused on the heterogeneity of IBS, eight of them utilized K-means Cluster Analysis (K-means CA), four employed Hierarchical Cluster Analysis, and only two studies utilized Latent Class Analysis. Seven studies focused on clinical symptoms, while four articles examined anocolorectal functions. Two studies were centered around immunological findings, and only one study explored microbial composition. The number of clusters obtained ranged from two to seven, showing variation across the studies. Males exhibited lower symptom severity and fewer psychological findings. The association between symptom severity and rectal perception suggests that altered rectal perception serves as a biological indicator of IBS. Ultra-slow waves observed in IBS patients are linked to increased activity of the anal sphincter, higher anal pressure, dystonia, and dyschezia. Conclusion IBS has different subgroups based on different factors. Most IBS patients have low clinical severity, good QoL, high rectal sensitivity, delayed left colon transit time, increased systemic cytokines, and changes in microbial composition, including increased Firmicutes-associated taxa and depleted Bacteroidetes-related taxa. However, the number of clusters is inconsistent across studies due to the methodological heterogeneity. CA, a valuable non-supervised learning method, is sensitive to hyperparameters like the number of clusters and random initialization of cluster centers. The random nature of these parameters leads to diverse outcomes even with the same algorithm. This has implications for future research and practical applications, necessitating further studies to improve our understanding of IBS and develop personalized treatments.

Introduction. This study investigates the long-term effectiveness and safety of endovenous laser treatment (EVLT) for chronic venous insufficiency (CVI), a condition commonly caused by dysfunctional valves in the venous circulation system. Materials and Methods. In this retrospective cohort study, patients underwent EVLT and were followed up for successive short intervals and one last time after a median duration of 9-year postprocedural. Pre- and postprocedure duplex ultrasound was used to assess changes in the great saphenous vein (GSV) diameter, reflux, and saphenofemoral junction incompetence. Quality of life was evaluated using the SF-36 and Aberdeen Varicose Vein Questionnaire (AVVQ). Results. Sixty-eight patients with a mean age of 52.4±12.4 years were enrolled in the study. The mean follow-up time was 8.9±2.1 years, ranging from 5 to 12 years. The mean GSV diameter significantly decreased in all patients (whole group) across proximal (from 5.8±2.3 mm to 4.2±2.1 mm), middle (from 4.7±1.6 mm to 2.8±2.2 mm), and distal (from 4.5±2.3 mm to 2.2±2.2 mm) segments, with P