Diagnostics最新文献

{"title":"A Study on the Establishment of Diagnostic Reference Levels for Cardiovascular Angiography and Interventional Procedures: Korean General Hospital.","authors":"Daeho Kim, Jungsu Kim","doi":"10.3390/diagnostics16081243","DOIUrl":"10.3390/diagnostics16081243","url":null,"abstract":"<p><p><b>Background/Objectives</b>: Cardiovascular interventions require prolonged fluoroscopy, which increases the risk of radiation. Diagnostic Reference Levels (DRLs), set at the 75th percentile of the dose distribution, are vital benchmarks for dose optimization. Following the release of national DRLs by the Korea Disease Control and Prevention Agency in March 2025, this study established institutional DRLs at a tertiary center to evaluate local optimization against national and international standards. <b>Methods</b>: This study analyzed radiation doses from 2022 to 2024 using DICOM Radiation Dose Structured Reports data from a single center's angiography system. The total kerma-area product values and fluoroscopy times were evaluated across the categorized procedures. Following the International Commission on Radiological Protection guidelines, institutional DRLs were established at the 75th percentile of the dose distribution to benchmark against national and international DRLs. <b>Results</b>: Analysis of 1663 radiation dose structured reports established institutional DRLs, with the total kerma-area product ranging from 23.43 Gy·cm² for coronary angiography to 329.45 Gy·cm² for chronic total occlusion interventions. Complexity significantly increased the radiation burden; multivessel percutaneous coronary intervention and acute myocardial infarction nearly doubled the doses and fluoroscopy times in single-vessel interventions. Although the diagnostic procedures were cine image-driven, for moderate-complexity interventions, the contribution of fluoroscopy was greater. <b>Conclusions</b>: These findings support institutional optimization and development of safety guidelines to enhance patient protection during high-complexity cardiovascular procedures.</p>","PeriodicalId":11225,"journal":{"name":"Diagnostics","volume":"16 8","pages":""},"PeriodicalIF":3.3,"publicationDate":"2026-04-21","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC13114416/pdf/","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"147812141","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Monitoring the Depth of Sedation During Gastrointestinal Endoscopy: A Narrative Review of Current Evidence and Clinical Recommendations.","authors":"Sonia Elena Popovici, Bogdan Miutescu, Stelian Adrian Ritiu, Tudor Voicu Moga, Ioan Sporea, Dorel Sandesc, Ovidiu Bedreag, Marius Păpurică, Mădălina Butaș, Alina Popescu","doi":"10.3390/diagnostics16081245","DOIUrl":"10.3390/diagnostics16081245","url":null,"abstract":"<p><p>Sedation and anesthesia are integral components of modern gastrointestinal endoscopy, enhancing patient comfort and procedural success while adding risks such as respiratory and cardiovascular complications. Accurate monitoring of sedation depth is essential to balance safety and procedural efficacy. This narrative literature review synthesizes current evidence on monitoring depth of anesthesia during endoscopic procedures, including clinical assessment scales, capnography, and processed electroencephalogram (pEEG)-based technologies. The effects of commonly used sedative agents on monitoring parameters and the impact of different monitoring strategies on clinical outcomes are also discussed. Current evidence indicates that clinical assessment remains the cornerstone of monitoring during moderate sedation, while capnography improves early detection of respiratory compromise during deep sedation. pEEG-based monitoring may provide additional value in selected high-risk or prolonged procedures but should complement, not replace, clinical evaluation. A multimodal monitoring approach tailored to sedation depth and patient risk profile is likely to be the most effective strategy for optimizing patient safety. Future research should focus on standardizing monitoring protocols and identifying populations most likely to benefit from advanced monitoring techniques.</p>","PeriodicalId":11225,"journal":{"name":"Diagnostics","volume":"16 8","pages":""},"PeriodicalIF":3.3,"publicationDate":"2026-04-21","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC13115536/pdf/","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"147812286","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Clinical Predictors of Dosimetric Precedence for Deep Inspiratory Breath Hold Radiation Therapy for Breast Cancer.","authors":"Alzahra'a Al Matairi, Abdulla Alzibdeh, Issa Mohamad, Ramiz Abu-Hijlih, Wafa Asha, Fadwa Abdel Rahman, Haitham Kanaan, Soha Ahmad, Hikmat Abdel-Razeq, Fawzi Abuhijla","doi":"10.3390/diagnostics16081236","DOIUrl":"10.3390/diagnostics16081236","url":null,"abstract":"<p><p><b>Background/Objectives</b>: Radiation therapy for left-sided breast cancer exposes the heart and lungs to incidental radiation, increasing long-term risk of cardiopulmonary morbidity. This study evaluated the dosimetric impact of DIBH compared with free breathing (FB) and explored patient- and treatment-related predictors of benefit. <b>Methods</b>: We retrospectively analyzed 90 patients with left-sided breast cancer or chest wall irradiation planned under both FB and DIBH. Dosimetric parameters included mean heart dose (MHD), mean lung dose (MLD), heart V5, and lung V20. Univariable analyses assessed associations between dose reductions (Δ%) and clinical factors (age, BMI, fractionation, boost, nodal fields, and smoking). Multivariable regression identified independent predictors. <b>Results</b>: DIBH significantly reduced mean doses to the heart (FB 4.25 ± 1.47 Gy vs. DIBH 2.64 ± 1.28 Gy; Δ -1.60 Gy, <i>p</i> < 0.001, and Cohen's d = 2.46) and lung (FB 14.00 ± 3.45 Gy vs. DIBH 11.64 ± 3.32 Gy; Δ -2.36 Gy, <i>p</i> < 0.001, and Cohen's d = 2.73). Similarly, heart V5 (median Δ -10.5%, <i>p</i> < 0.001) and lung V20 (median Δ -5.5%, <i>p</i> < 0.001) were significantly improved. Higher BMI was independently associated with smaller relative reductions in MHD (-1.4% per unit, <i>p</i> < 0.001) and MLD (-0.31% per unit, <i>p</i> = 0.019). Hypofractionation was linked to greater MHD and MLD reductions (+8.9%, <i>p</i> = 0.007), (3.7%, <i>p</i> = 0.035), respectively. Supraclavicular field irradiation increased lung exposure, while internal mammary chain fields elevated both lung and heart doses. Age showed no significant influence. <b>Conclusions</b>: DIBH provides robust and consistent reductions in cardiac and pulmonary radiation doses during left-sided breast irradiation. The degree of sparing is modulated by BMI, fractionation, and nodal field inclusion.</p>","PeriodicalId":11225,"journal":{"name":"Diagnostics","volume":"16 8","pages":""},"PeriodicalIF":3.3,"publicationDate":"2026-04-21","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC13115145/pdf/","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"147812006","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Accuracy of Modified Budin Views for the Femoral Neck Anteversion at Different Hip Abduction Angles: An Experimental Study on Dry Bones.","authors":"Murat Yuncu, Emre Mucahit Kartal, Sacide Efsun Urger, Levent Sarıkcıoglu, Serkan Gurcan, Ozkan Kose","doi":"10.3390/diagnostics16081238","DOIUrl":"10.3390/diagnostics16081238","url":null,"abstract":"<p><p><b>Background/Objectives:</b> The modified Budin radiographic technique is a practical alternative to CT for measuring femoral neck anteversion (FNA); however, the impact of hip abduction angle on its accuracy remains unclear. This experimental study examined how varying abduction angles affect agreement between modified Budin measurements and CT. <b>Methods</b>: Twenty-seven dry adult femora underwent CT scanning, and FNA was measured using a validated three-slice superimposition method as the reference standard. Modified Budin radiographs were obtained at 20°, 30°, and 40° of femoral abduction. Two orthopedic surgeons independently measured FNA on all images twice, with at least 15 days between measurements. Intra- and interobserver reliability were assessed using the intraclass correlation coefficient (ICC). Mean values per femur were analyzed. Agreement with CT was evaluated using Pearson correlation, Bland-Altman analysis, and absolute error comparisons across abduction angles. <b>Results:</b> Reliability was excellent across all modalities (ICC, 0.982-0.998). Mean CT-derived FNA was 10.0° ± 8.5°, compared with 9.1° ± 8.0° at 20°, 8.3° ± 7.8° at 30°, and 7.8° ± 7.5° at 40° of abduction (<i>p</i> < 0.001). Correlation with CT was strong at all positions, but systematic underestimation increased with abduction angle. Among the tested positions, 20° abduction showed the smallest bias, the narrowest limits of agreement, and the lowest absolute error. <b>Conclusions</b>: Hip abduction angle significantly influences the accuracy of the modified Budin view. Under controlled experimental conditions, 20° abduction provided the closest agreement with CT among the tested positions. These findings suggest that lower abduction angles may improve geometric accuracy, although clinical feasibility and performance must be confirmed in vivo before routine clinical application can be recommended.</p>","PeriodicalId":11225,"journal":{"name":"Diagnostics","volume":"16 8","pages":""},"PeriodicalIF":3.3,"publicationDate":"2026-04-21","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC13115102/pdf/","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"147812206","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Diagnostic Accuracy of 2 cm Versus 4 cm Insertion Depth for Nasal Swabs for SARS-CoV-2 Rapid Antigen Testing-A Randomized Controlled Trial.","authors":"Rasmus Eið Callesen, Tobias Todsen, Rebekka Consuelo Eið, Sabrina Dandanell Stange, Tobias Gredal, Nikolai Kirkby, Michael Papesch, Christian von Buchwald, Kathrine K Jakobsen","doi":"10.3390/diagnostics16081225","DOIUrl":"10.3390/diagnostics16081225","url":null,"abstract":"<p><p><b>Background/Objectives</b>: Optimal specimen collection is essential for accurate diagnostic tests for upper airway infections. Rapid antigen diagnostic tests (RDTs) are commonly used for SARS-CoV-2 testing, yet the optimal sampling depth remains unclear. This study aimed to compare the diagnostic sensitivity and patient discomfort associated with two nasal swab depths: 2 cm (anterior nasal) and 4 cm (proposed mid-turbinate). <b>Methods</b>: In this randomized, paired clinical trial conducted at a public COVID-19 test center in Copenhagen, Denmark, 309 adults presenting for SARS-CoV-2 RT-PCR testing were enrolled. Each participant underwent bilateral nasal sampling using RDTs: one nostril with a 2 cm swab and the other with a 4 cm swab, randomized by side. RT-PCR from oropharyngeal swabs served as the reference standard. Discomfort was rated using a 10-point visual analog scale (VAS). <b>Results</b>: Among the 309 participants, 57 (18.4%) tested positive for SARS-CoV-2 by RT-PCR. RDT sensitivity was 62.1% (95% CI: 48.4-74.5%) for 2 cm swabs and 70.2% (95% CI: 56.6-81.6%) for 4 cm swabs, a non-significant difference (<i>p</i> = 0.34). Among symptomatic individuals, sensitivity increased to 74.4% (2 cm) and 86.0% (4 cm), though the difference also remained non-significant (<i>p</i> = 0.17). Discomfort scores were significantly higher for the 4 cm swab (mean VAS: 5.2) compared to 2 cm (mean VAS: 3.8; <i>p</i> < 0.001). <b>Conclusions</b>: While not statistically significant, deeper mid-turbinate swabbing (4 cm) showed higher diagnostic sensitivity than anterior nasal swabbing (2 cm), especially in symptomatic individuals. However, this came at the cost of increased discomfort. These findings highlight the importance of balancing diagnostic performance and patient tolerability in pandemic testing strategies. The study contributes valuable evidence to inform future guideline development, particularly regarding swab technique, test accuracy, and feasibility in clinical and public health settings.</p>","PeriodicalId":11225,"journal":{"name":"Diagnostics","volume":"16 8","pages":""},"PeriodicalIF":3.3,"publicationDate":"2026-04-20","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC13114395/pdf/","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"147812137","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Quantitative Imaging Biomarkers of PRP-Induced Tendon Remodelling in Chronic Tendinopathy: Review and Single-Centre Experience with Ultrasound Radiomics and MRI T2 Profiling.","authors":"Živa Miriam Geršak, Karlo Pintarić, Jernej Vidmar, Vladka Salapura","doi":"10.3390/diagnostics16081233","DOIUrl":"10.3390/diagnostics16081233","url":null,"abstract":"<p><p>Platelet-rich plasma (PRP) is widely used as a second-line treatment for chronic tendinopathy that persists despite structured conservative care, yet outcomes and imaging correlates remain heterogeneous. This review outlines PRP biology and preparation, summarises quantitative imaging techniques for monitoring tendon response, and presents the experience of a single centre integrating these methods into routine supraspinatus and lateral elbow PRP workflows. PRP is described as an autologous platelet concentrate with variable leukocyte and fibrin content, with leukocyte-rich formulations commonly selected for chronic tendinopathy. Quantitative approaches-including ultrasound shear-wave elastography and radiomics, MRI T2/T2* mapping, CT-based bone metrics, PET/CT, and optical techniques-offer numerical biomarkers of tendon structure, mechanics, and inflammation but are rarely implemented in PRP trials. At the authors' centre, leukocyte-rich PRP is injected under ultrasound guidance after failed physiotherapy, and follow-up combines validated questionnaires with grey-level run-length matrix texture analysis of ultrasound and 3.0 T MRI T2 distribution profiling. A pilot ultrasound study in supraspinatus and common extensor tendinosis showed uniform short-term clinical improvement and significant changes in most texture features, with selected parameters correlating with symptom relief. A prospective supraspinatus cohort demonstrated significant six-month clinical gains in both tendinosis and small partial-thickness tears, whereas only the tendinosis group exhibited T2 profile convergence toward asymptomatic patterns. These data indicate that quantitative ultrasound radiomics and whole-length T2 profiling are feasible imaging biomarkers that capture PRP-induced tendon remodelling beyond qualitative imaging and may help tailor PRP protocols to specific tendon phenotypes.</p>","PeriodicalId":11225,"journal":{"name":"Diagnostics","volume":"16 8","pages":""},"PeriodicalIF":3.3,"publicationDate":"2026-04-20","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC13115292/pdf/","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"147812190","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Dermoscopic Interface Features in Melanoma-Seborrheic Keratosis Collision Tumors: A Scoping Review with an Illustrative Case Report on Wood's Lamp Dermoscopy.","authors":"Alexandre Raphael Meduri, Francesca Ambrogio, Lucia Lospalluti, Domenico Bonamonte, Giulia Ciccarese, Gerardo Cazzato, William Andrew Rosato, Paolo Romita, Mario Della Mura, Hugo Guillermou, Caterina Foti","doi":"10.3390/diagnostics16081226","DOIUrl":"10.3390/diagnostics16081226","url":null,"abstract":"<p><p><b>Background/Objectives:</b> Collision tumors between seborrheic keratosis and melanoma represent a well-known diagnostic pitfall, particularly when the benign keratinocytic component constitutes the predominant portion of the lesion. In such cases, melanoma-specific dermoscopic clues may be obscured by typical seborrheic keratosis patterns, leading to potential underestimation. The aim of this scoping review was to map and summarize the dermoscopic interface features reported in melanoma-seborrheic keratosis collision tumors. Secondary aims were to discuss diagnostic pitfalls, explore potential tumor microenvironment considerations, and assess the adjunctive role of Wood's lamp-assisted dermoscopy. <b>Methods:</b> This review was conducted as a scoping review and reported according to the PRISMA-ScR guidelines using PubMed, Scopus, and Web of Science. Studies reporting histologically confirmed melanoma-seborrheic keratosis collision tumors with available dermoscopic documentation were included. Eligible articles consisted of case reports and case series. Dermoscopic features at the interface between seborrheic keratosis and melanoma were qualitatively synthesized. <b>Results:</b> Five studies describing five melanoma-seborrheic keratosis collision tumors met the inclusion criteria. In all cases, the seborrheic keratosis component was dermoscopically recognizable. Asymmetric interface-related hyperpigmentation was consistently observed in collisions involving pigmented melanomas, whereas it was absent in the single reported case of hypopigmented melanoma. <b>Conclusions:</b> Asymmetric interface-related hyperpigmentation within seborrheic keratosis is a recurrent dermoscopic finding in melanoma-seborrheic keratosis collision tumors and could be considered a monitoring clue rather than a melanoma-specific diagnostic criterion. Given the dynamic nature of melanoma growth, longitudinal assessment of the dermoscopic interface may be particularly informative. Adjunctive techniques, including Wood's lamp-assisted dermoscopy, may support interface-focused evaluation in selected equivocal cases.</p>","PeriodicalId":11225,"journal":{"name":"Diagnostics","volume":"16 8","pages":""},"PeriodicalIF":3.3,"publicationDate":"2026-04-20","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC13114341/pdf/","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"147812192","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Radiomics in the Evaluation of Cystic and Neoplastic Lytic Lesions of the Jaws.","authors":"Paola Di Giacomo, Pasquale Frisina, Alberto Fratocchi, Pierluigi Barra, Cira Rosaria Tiziana Di Gioia, Flavia Adotti, Giovanni Falisi, Fabrizio Spallaccia, Iole Vozza, Antonella Polimeni, Carlo Di Paolo, Daniela Messineo","doi":"10.3390/diagnostics16081222","DOIUrl":"10.3390/diagnostics16081222","url":null,"abstract":"<p><p><b>Background/Objectives.</b> Radiomics is an emerging imaging-based tool that enhances lesion characterization beyond conventional diagnostic approaches. Its potential in evaluating osteolytic lesions of the jaws lies in improving discrimination between benign and malignant entities. This study aimed at developing a predictive model to identify radiomic features capable of distinguishing benign from malignant lesions. <b>Methods.</b> Subjects with preoperative CT or CBCT and histopathological confirmation were included. A pilot cohort was used for feature selection via LASSO regression, which ranked features by frequency and absolute coefficient. Malignancy was coded as class 1, benign lesions as class 0. Positive coefficients indicated association with malignancy, while negative coefficients with benign characteristics. The most stable features were initially trained on the pilot cohort and then validated on an independent test set through machine learning classifiers as LASSO, support vector machine, artificial neural network, random forest e XGboost. <b>Results.</b> The sample comprised 69 subjects (pilot cohort = 57, test cohort = 12). The predictors selected from LASSO regression were: DifferenceEntropy_GLCM (-0.768), CenterOfMassShift_MORPHOLOGICAL (-1.390), INTENSITY-HISTOGRAM_MaximumHistogramGradientGrayLevel (1.139), GLRLM_ShortRunLowGrayLevelEmphasis (-0.742), and Maximum3DDiameter_MORPHOLOGICAL (0.932). As for model performance on test, LASSO achieved the best performance (AUC 0.83), with perfect specificity and sensitivity of 0.71. SVM showed good AUC but poor sensitivity, while random forest and XGBoost performed poorly (AUC 0.57 and 0.37, respectively). <b>Conclusions.</b> The LASSO model proved to be a transparent and robust classifier, suitable for both feature selection and external validation. The selected features demonstrated strong discriminative ability, supporting the potential of radiomics in improving lesion assessment and guiding clinical decision-making.</p>","PeriodicalId":11225,"journal":{"name":"Diagnostics","volume":"16 8","pages":""},"PeriodicalIF":3.3,"publicationDate":"2026-04-20","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC13114418/pdf/","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"147812204","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Global and Regional Diagnostic Results of Progress Toward Cervical Cancer Elimination, According to the WHO Strategy: A Systematic Review.","authors":"Dan Cristian Luca, Ciprian Cirimbei, Sinziana Octavia Ionescu, Vlad Rotaru, Dan Nicolae Straja, Mihnea Alecu, Elena Chitoran, Daniela Cristina Stefan, Laurentiu Simion","doi":"10.3390/diagnostics16081224","DOIUrl":"10.3390/diagnostics16081224","url":null,"abstract":"<p><p><b>Background/Objectives</b>: In 2020, the World Health Organization (WHO) launched a strategy to eliminate cervical cancer by ensuring high vaccination coverage, effective screening, and access to treatment. This systematic review assesses how this strategy has been implemented in different world regions, what results have been achieved, and what disparities and challenges have been faced. <b>Methods</b>: A systematic search was performed in PubMed, Web of Science, GLOBOCAN, IARC, ASCO, and ESMO for publications issued between November 2020 and December 2024. A total of 721 records were reviewed and assessed, resulting in the inclusion of 47 studies in the final analysis. <b>Results</b>: Considerable regional variations were identified between the regions analyzed. Countries with organized vaccination programs, structured screening systems, and stronger treatment infrastructure generally reported more favorable outcomes. In contrast, several settings in Eastern Europe, parts of Asia, and sub-Saharan Africa remained below the WHO targets, with lower vaccination uptake, limited screening participation, and a persistently high cervical cancer burden. Progress was further constrained by inequalities in access to healthcare, differences in program organization, and limited comparability of available national data. <b>Conclusions</b>: Although many countries demonstrate that eliminating cervical cancer is achievable, global alignment with WHO targets remains uneven. Strengthening health infrastructure, improving vaccination uptake, and ensuring equitable access to screening and treatment are essential to turning this strategy into a realistic global outcome.</p>","PeriodicalId":11225,"journal":{"name":"Diagnostics","volume":"16 8","pages":""},"PeriodicalIF":3.3,"publicationDate":"2026-04-20","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC13115489/pdf/","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"147812025","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"An Adaptive Deep Learning Framework for Multi-Label Chest X-Ray Diagnosis Using a Hybrid CNN-Transformer Architecture and Class-Wise Ensemble Fusion.","authors":"Chi-Feng Hsieh, Hsu-Hsia Peng, Yu-Hsiang Tsai, Chia-Ching Chang, Cheng-Hsuan Juan, Hsian-He Hsu, Chun-Jung Juan","doi":"10.3390/diagnostics16081227","DOIUrl":"10.3390/diagnostics16081227","url":null,"abstract":"<p><p><b>Background/Objectives</b>: To develop and externally evaluate a deep learning framework for multi-label thoracic disease classification on chest radiographs using hybrid convolutional neural network (CNN)-transformer architectures, hierarchical scalar-weighted fusion, and ensemble strategies. <b>Methods</b>: This retrospective, multi-center study utilized publicly available datasets: NIH ChestX-ray14 (112,120 images; 30,805 patients) for model development and internal testing, and CheXpert (223,415 images) plus ChestX-Det10 (3578 images) for external validation. Nine CNN-transformer hybrids were systematically benchmarked, and the proposed model incorporated multi-scale DenseNet121 features, scalar-weighted fusion, positional encodings, and cross-attention. Four post hoc ensemble methods were explored, including a class-wise Top-3 Grid Search. Performance was evaluated using AUROC as the primary metric, along with precision, recall, F1-score, accuracy, specificity, positive predictive value, and negative predictive value. Statistical comparisons were performed using bootstrapped resampling and appropriate parametric or non-parametic tests. <b>Results</b>: On the NIH internal test set, the proposed hybrid model achieved a mean AUROC of 0.8495, which was significantly higher than that of the DenseNet121 baseline (0.8441, <i>p</i> = 0.032). The Top-3 Grid Search ensemble further improved internal performance, achieving a mean AUROC of 0.8577 (<i>p</i> < 0.00001). On external validation, the ensemble consistently outperformed DenseNet121, achieving mean AUROCs of 0.6500 on CheXpert (<i>p</i> < 0.001) and 0.6592 on ChestX-Det10 (<i>p</i> < 0.001). Per-class analysis revealed significant improvements for clinically important conditions such as cardiomegaly, mass, and pneumothorax. Grad-CAM visualizations demonstrated the strong alignment of predicted abnormalities with radiologically relevant regions. <b>Conclusions</b>: This CNN-transformer framework, particularly when combined with class-wise ensemble strategies, provided modest but statistically significant improvements in multi-label chest X-ray classification. External validation suggested partial generalizability across datasets, although performance remained moderate under domain shift.</p>","PeriodicalId":11225,"journal":{"name":"Diagnostics","volume":"16 8","pages":""},"PeriodicalIF":3.3,"publicationDate":"2026-04-20","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC13115109/pdf/","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"147812117","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}