International Journal of Biomedical Imaging最新文献

{"title":"Classification of Chest X-Ray Images Using Modern Machine Learning Techniques by Orang3 Data Mining Tool.","authors":"Amani Al-Ghraibah, Areen Arabiat, Muneera Altayeb","doi":"10.1155/ijbi/8859105","DOIUrl":"10.1155/ijbi/8859105","url":null,"abstract":"<p><p>During the COVID-19 pandemic, there were many concerns raised in diagnosing chest diseases. Regular diagnosis processes of chest diseases sometimes fail in distinguishing between Corona, viral pneumonia diseases, and normal chest x-rays through regular assessments which also need time and not as accurate as the automated diagnosis based on machine learning. Chest diseases can be sometimes diagnosed using x-ray images where these images are used to identify if the patient is normal or has one of the chest diseases like Corona and viral pneumonia. Modern machine learning techniques have achieved high performance in aiding medical diagnostic processes. The novelty of this work lies in a wavelet-based feature extraction framework coupled with a systematic evaluation of ensemble learning algorithms for accurate and computationally efficient classification of COVID-19, viral pneumonia, and normal chest x-ray images. The number of patients used in each case is 1200 patients and this huge dataset is used to create classification models using adaptive boosting (AdaBoost), gradient boosting (GB), and random forest (RF) machine learning algorithms using Orange3 data mining tool. Using wavelet features, the results of evaluating the proposed models were 99.5% on average. The proposed models are very effective in identifying chest diseases and can be combined with x-ray machines in hospitals, so physicians can determine the best treatment for patients. The proposed methods beat other researchers found in literature as they are faster and more accurate where each step takes few seconds on average to complete.</p>","PeriodicalId":47063,"journal":{"name":"International Journal of Biomedical Imaging","volume":"2026 ","pages":"8859105"},"PeriodicalIF":2.3,"publicationDate":"2026-10-06","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC13643059/pdf/","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"149128722","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Advancing Ultrasound Beamforming With Deep Learning: A Comprehensive Review of Methods, Datasets, Benchmarks, and Computational Challenges.","authors":"Hamza Hadri, Abderahhim Fail, Mohamed Sadik","doi":"10.1155/ijbi/4617573","DOIUrl":"https://doi.org/10.1155/ijbi/4617573","url":null,"abstract":"<p><p>Ultrasound imaging is a widely used diagnostic tool, and beamforming techniques are integral to its performance. Traditional methods such as delay-and-sum (DAS) and delay-multiply-and sum (DMAS) have limitations in terms of image quality and computational efficiency. Recently, deep learning approaches have shown significant promise in improving ultrasound beamforming, offering enhanced image resolution and quality, alongside potential for real-time processing. This review is aimed at providing a comprehensive overview of the current state-of-the-art in deep learning-based ultrasound beamforming. It focuses on analyzing various deep learning models, their applications, challenges, and the integration of hardware optimization strategies to enhance performance and efficiency. The review examines a range of deep learning models applied to ultrasound beamforming, including convolutional neural networks (CNNs), generative adversarial networks (GANs), transformer-based models, and hybrid approaches. It also discusses the challenges associated with dataset limitations, model interpretability, and the risk of overfitting. Furthermore, the review explores hardware acceleration using field-programmable gate arrays (FPGAs), along with cloud-edge frameworks for real-time inference. The findings highlight the potential of deep learning models to outperform traditional methods in terms of image quality and resolution. However, challenges such as the need for large and diverse datasets, the black box nature of deep learning models, and the risk of overfitting remain. Hardware optimization through FPGAs has proven effective in enabling real-time processing, but the integration of edge computing with cloud-based solutions offers promising avenues for balancing performance, latency, and energy efficiency. Deep learning-based ultrasound beamforming has great potential to advance medical imaging. Future efforts should enhance model generalizability with diverse and synthetic data, optimize hardware for real-time use, and establish standardized validation protocols to support clinical adoption.</p>","PeriodicalId":47063,"journal":{"name":"International Journal of Biomedical Imaging","volume":"2026 ","pages":"4617573"},"PeriodicalIF":2.3,"publicationDate":"2026-10-03","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC13633549/pdf/","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"149116440","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Big Data Analytics Framework for Early Breast Cancer Detection Using Optimized MapReduce and Deep Learning on Large-Scale Mammographic Data.","authors":"Mohammed M Ahmed, Ehab Hassanien","doi":"10.1155/ijbi/8325888","DOIUrl":"https://doi.org/10.1155/ijbi/8325888","url":null,"abstract":"<p><p>The paper presents a framework that is aimed at addressing the large-scale data challenge with a particular focus on early prediction of breast cancer based on mammography image data. The abundance of large-scale medical imaging data in healthcare has opened the way to intelligent diagnostic systems through its ability to analyze and process large-scale mammography data. The proposed framework, the metaheuristic optimization based large-scale breast cancer detection (MO-LSBCD) framework provides a complete solution to the extraction of discriminative features that can be used to predict the disease correctly. Early and accurate diagnosis of breast cancer is one of the most critical factors in improving survival rates and reducing treatment costs for patients. The MO-LSBCD framework is divided into three phases: (1) The CBIS-DDSM mammography dataset consists of 10,239 image files with a size of 6 GB, but for classification, the analytical unit is the annotated region of interest (ROI), with 3568 ROI-level abnormality records from 1566 unique patients (normal, malignant, benign, and benign-without-callback); (2) solving the large-scale data problem by implementing an optimized MapReduce model to efficiently distribute and process the large volumes of breast cancer mammography images in parallel; and (3) effective deep learning techniques to classify breast cancer pathology. The experimental findings indicate that the enhanced MapReduce model reduces total processing time and improves F-measure in comparison with the standard MapReduce model. Among different deep learning models tested on the CBIS-DDSM dataset, the classification of EfficientNet-B3 and ResNet-50 was better with an accuracy of 99.64% and 98.91%, respectively, compared with the other models on all evaluation metrics. The results of this paper confirm that MO-LSBCD framework can be used to provide a robust and scalable solution to large-scale breast cancer detection to help clinicians in the early diagnosis of breast cancer and eventually help in the improved patient outcome.</p>","PeriodicalId":47063,"journal":{"name":"International Journal of Biomedical Imaging","volume":"2026 ","pages":"8325888"},"PeriodicalIF":2.3,"publicationDate":"2026-09-30","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC13624933/pdf/","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"149116580","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Leveraging CNN and Transfer Learning With EfficientNet for Enhanced Optical Coherence Tomography (OCT).","authors":"Siddartha Arekanti, Ch Madhu Bhushan, Irfan Alam, Priya Koppuravuri, Fazal Shaik, Md Muzakkir Hussain, Firoj Gazi","doi":"10.1155/ijbi/1272954","DOIUrl":"10.1155/ijbi/1272954","url":null,"abstract":"<p><p>Optical coherence tomography (OCT) is an important imaging modality for detecting retinal disorders, including diabetic macular edema (DME), choroidal neovascularization (CNV), and drusen, as well as distinguishing healthy retinal patterns. Nevertheless, visual assessment of OCT scans is labor-intensive, observer-dependent, and dependent on expert clinical knowledge, encouraging reliable automated screening tools. This work introduces a hybrid Deep Transferred EfficientNet and Support Vector Machine (DTE-SVM) approach for multiclass OCT image classification. In contrast to standard CNN models that commonly use a softmax layer for final prediction, the proposed framework employs transfer-learned EfficientNet representations with an SVM decision model to strengthen interclass discrimination and robustness within the extracted feature space. Experiments used a working subset of 8767 OCT images selected from the OCT2017 dataset, representing nearly 10.4% of the complete 84,495-image collection. This smaller yet representative subset was adopted to lower computational cost while maintaining class distribution among CNV, DME, drusen, and normal groups. Five EfficientNet architectures, from B0 through B4, were assessed using stratified 10-fold cross-validation. Among them, DTE-SVM-B0 produced the strongest overall results, obtaining an accuracy of 0.9076, precision of 0.9074, recall of 0.9076, and F1-score of 0.9073. DTE-SVM-B2 showed the smallest fold-level standard deviation, reflecting comparatively stable behavior. Comparative experiments indicated that DTE-SVM-B0 surpassed ResNet50, DenseNet-121, and EfficientNetB0 softmax-based classifiers, whereas ablation testing verified that substituting the softmax layer with SVM improved results for every EfficientNet backbone. Class-level clinical evaluation also revealed strong sensitivity and specificity, with CNV identified most accurately and drusen remaining the hardest category. Overall, these results suggest that compact EfficientNet feature extractors integrated with SVM offer accurate, stable, and computationally practical OCT-driven retinal disease classification for clinical decision-support settings.</p>","PeriodicalId":47063,"journal":{"name":"International Journal of Biomedical Imaging","volume":"2026 ","pages":"1272954"},"PeriodicalIF":2.3,"publicationDate":"2026-09-07","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC13550866/pdf/","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"148913999","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Evaluation of the Relationship Between Facial Hue and Blood Pressure Fluctuations in Hemodialysis Patients.","authors":"Kosuke Oiwa, Takehiro Okama, Satoshi Suzuki, Yoshitaka Maeda, Akio Nozawa","doi":"10.1155/ijbi/8387784","DOIUrl":"10.1155/ijbi/8387784","url":null,"abstract":"<p><strong>Objective: </strong>In hemodialysis, hypotension may occur due to fluid removal, but traditional intermittent blood pressure measurements risk missing its onset. We are aimed to enabling noninvasive and continuous blood pressure monitoring in patients undergoing hemodialysis based on the spatial features of facial visible images. Facial visible images contain not only skin chromaticity information but also facial expression-related variations. To improve estimation, facial color components are separated from expressions, analyzed for their relationship with blood pressure fluctuations, and used to build an individual estimation model.</p><p><strong>Methods: </strong>Facial landmark coordinates were extracted, and eight facial regions of interest (ROIs) were defined on facial images of hemodialysis patients. Mean and differential color values between ROIs served as explanatory variables, whereas mean blood pressure was the dependent variable. Stepwise selection identified key variables for blood pressure estimation, followed by regression modeling.</p><p><strong>Results: </strong>Applying a linear regression model to the mean color values in each ROI resulted in a higher generalization performance across all subjects. The facial hue components of the a ^∗ and b ^∗ values were used as explanatory variables contributing to blood pressure estimation in all subjects, but there were individual differences in the location of ROIs selected as explanatory variables.</p><p><strong>Conclusion: </strong>Applying a linear regression model to the absolute values of facial hue improved generalization performance, and there are individual differences in the relationship between facial hue components and blood pressure fluctuations. This study is significant because these findings could contribute to the realization of remote blood pressure monitoring during hemodialysis.</p>","PeriodicalId":47063,"journal":{"name":"International Journal of Biomedical Imaging","volume":"2026 ","pages":"8387784"},"PeriodicalIF":2.3,"publicationDate":"2026-08-31","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC13530390/pdf/","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"148876262","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"REFA-DINO: A Relative-Efficient Fusion Attention Framework for Binary and Multiclass Diabetic Retinopathy Classification.","authors":"Hussain Dawood, Muniba Noreen, Ali Javed, Hafsa Ilyas, Usman Muhammad, Hassan Dawood","doi":"10.1155/ijbi/4374759","DOIUrl":"10.1155/ijbi/4374759","url":null,"abstract":"<p><p>Diabetic retinopathy (DR) has become a major cause of preventable blindness, and its early detection is crucial for maintaining vision in diabetic patients. Existing computer-aided diagnosis (CAD) systems still face challenges in detecting early-stage lesions, integrating local and global features, and maintaining consistent performance in binary and multiclass settings. To address these limitations, this study proposes a relative-efficient fusion attention framework (REFA-DINO) that integrates a convolutional local-feature learning module with a transformer-based global contextual method. The architecture employs a relative-efficient fusion adapter (REFA) block comprising two branches, EfficientNet and attention-enhanced patch embedding. EfficientNet extracts lesion-sensitive retinal features, and an attention-enhanced patch-embedding module is used to generate token representations. An adaptive relative position-aware cross-attention mechanism is introduced to enable effective interaction among heterogeneous features, which are further processed through the DINO-based transformer module. The extracted global representations are passed to the linear classification head for DR classification. REFA-DINO is evaluated on two diverse and large-scale fundus imaging datasets, EyePACS and APTOS, for binary and multiclass classification. The proposed framework achieved accuracies of 98.60% and 93.77% for binary classification and 85.52% and 87.51% for multiclass classification on the APTOS and EyePACS datasets, respectively. A comparative analysis against state-of-the-art methods, cross-corpora evaluation, and an ablation study are also conducted to demonstrate the effectiveness of the proposed REFA-DINO. Compared with state-of-the-art methods, the REFA-DINO framework enhances binary classification on both the EyePACS and APTOS datasets, with significant improvements over the strongest baselines. The proposed approach attains up to 6.19% and 9.39% improvement in the binary and multiclass classification accuracy on EyePACS. Likewise, it achieves an improvement of 2.64% on the binary classification score on APTOS with competitive multiclass classification performance. Experimental findings highlight the robustness and generalizability of the REFA-DINO framework in classifying DR.</p>","PeriodicalId":47063,"journal":{"name":"International Journal of Biomedical Imaging","volume":"2026 ","pages":"4374759"},"PeriodicalIF":2.3,"publicationDate":"2026-08-25","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC13507030/pdf/","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"148833569","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"A Clinical-Radiomics Nomogram Based on Pretreatment Magnetic Resonance Imaging Predicting Tumor Residual at the End of Radiotherapy in Patients With Nasopharyngeal Carcinoma.","authors":"Pingyan Liao, Min Zeng, Haitao Sun, Rongliang Xi, Guosen Huang","doi":"10.1155/ijbi/7443570","DOIUrl":"10.1155/ijbi/7443570","url":null,"abstract":"<p><strong>Background: </strong>The TNM staging system is a main tool for treatment stratification and prognosis prediction of nasopharyngeal carcinoma (NPC). However, patients with identical clinical stages often show different tumor regression patterns.</p><p><strong>Purpose: </strong>This study is aimed at constructing a nomogram model integrating MRI and clinical variables to predict primary tumor residual status at the completion of radiotherapy for NPC cases.</p><p><strong>Study type: </strong>The present study is a retrospective design.</p><p><strong>Population: </strong>Two hundred NPC patients who accepted radical radiotherapy in the present hospital from December 2017 to March 2023 were systematically reviewed. The selected 200 NPC cases were divided into the training cohort (<i>n</i> = 160) and validation cohort (<i>n</i> = 40).</p><p><strong>Field strength/sequence: </strong>1.5T MRI scan, contrast-enhanced T1-weighted (CE-T1WI) and T2-weighted (T2WI) sequences.</p><p><strong>Assessment: </strong>For the training cohort, the LASSO with logistic regression was conducted to choose important features and construct a radiomics model. The clinical variables of patients were selected on the basis of statistical analysis and previous research. The nomogram was developed by combining the clinical factors with the radiomics features.</p><p><strong>Statistical tests: </strong>Statistical analyses were conducted using R software for correlation analysis, chi-squared test, multivariable regression analysis, Mann-Whitney <i>U</i> test, ROC curve analysis, and DeLong test. Significant level was <i>p</i> < 0.05.</p><p><strong>Results: </strong>Up to 109 radiomics characteristics were extracted, from which seven stable characteristics were selected to generate the Rad-score. According to the regression result, T stage can be regarded as an independent predictor of primary tumor residual status. A nomogram combining the Rad-score and clinical variables was established and proved, yielding an AUC of 0.836 (95% CI: 0.766-0.889, <i>p</i> < 0.001) in the training set and 0.843 (<i>p</i> < 0.001) in the validation set, respectively. The DeLong test indicates that the nomogram model is superior to both the individual clinical variables and the Rad-score alone. The calibration curve and decision curve analyses proved the validity and clinical utility of the proposed model.</p><p><strong>Data conclusion: </strong>A clinical-radiomics nomogram has good efficacy in predicting primary tumor residual after radiotherapy, providing individual treatment strategies for NPC cases.</p>","PeriodicalId":47063,"journal":{"name":"International Journal of Biomedical Imaging","volume":"2026 ","pages":"7443570"},"PeriodicalIF":2.3,"publicationDate":"2026-08-07","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC13449176/pdf/","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"148689945","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Automated Semisupervised Measurement of Optic Nerve Sheath Diameter From CT Following Traumatic Brain Injury.","authors":"Emily Wittrup, Alan Kay, Elizabeth Lombard, Jett Rosen, Ethan Schnathorst, Christine Geng, Haoyuan Ma, Erica B Stein, Kevin R Ward, Craig Williamson, Kayvan Najarian","doi":"10.1155/ijbi/9929121","DOIUrl":"10.1155/ijbi/9929121","url":null,"abstract":"<p><p>Automated measurement of optic nerve sheath diameter (ONSD) from computed tomography (CT) scans is clinically important, but deep learning methods are often limited by the scarcity of high-quality, expert-labeled data given the labor-intensive nature of manual segmentations. To address this, we present a fully automated modular pipeline combining deep learning and semisupervised learning techniques for ONSD estimation from axial head CT scans. The pipeline features three stages: (1) deep learning-based slice selection, (2) semisupervised machine learning for optic nerve segmentation utilizing both labeled and unlabeled publicly available data, and (3) geometric and morphological ONSD measurement. When validated on public and internal datasets, our approach generalized better than traditional supervised segmentation methods, achieving an intersection over union (IoU) of 0.561 ± 0.141 and Dice coefficient of 0.707 ± 0.131 on previously unseen data. Slice-wise measurement accuracy varied based on measurement distance from the ocular globe, yielding mean absolute errors as low as 1.779 and 1.899 mm for the right and left ONSD, respectively. Our findings highlight the potential for semisupervised deep learning to deliver fully automated ONSD measurements and the framework's adaptability to difficult medical imaging tasks even with limited, low-quality ground truth for training.</p>","PeriodicalId":47063,"journal":{"name":"International Journal of Biomedical Imaging","volume":"2026 ","pages":"9929121"},"PeriodicalIF":2.3,"publicationDate":"2026-08-05","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC13442929/pdf/","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"148686426","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Deep Learning for Brain Tumour Analysis: A Systematic Review of CNN-Transformer Hybrids in Multimodal Imaging.","authors":"Solomon Buabeng Antwi, Peter Appiahene, Ben Beklisi Kwame Ayawli, Peter Nimbe","doi":"10.1155/ijbi/4763936","DOIUrl":"10.1155/ijbi/4763936","url":null,"abstract":"&lt;p&gt;&lt;strong&gt;Background: &lt;/strong&gt;Brain tumour detection and analysis using medical imaging requires the extraction of both local spatial features and global contextual representations. Although convolutional neural networks (CNNs) excel at capturing local spatial patterns and Transformer-based architectures model long-range dependencies effectively, the optimal architectural paradigm for clinical deployment remains unresolved. This systematic review and meta-analysis evaluates hybrid CNN-Transformer architectures for brain tumour detection, focusing on the integration of local and global feature learning, diagnostic accuracy and computational efficiency. The roles of generative adversarial networks (GANs) for addressing data scarcity and multimodal imaging fusion for diagnostic completeness are also critically examined.&lt;/p&gt;&lt;p&gt;&lt;strong&gt;Methods: &lt;/strong&gt;A systematic search was conducted across IEEE Xplore, PubMed, Scopus and Google Scholar for studies published between January 2021 and May 2025. From 1876 initially identified articles, 94 met the prespecified inclusion criteria following quality assessment using the QUADAS-2 and ROBINS-I frameworks. A random-effects meta-analysis of diagnostic accuracy was performed using the DerSimonian-Laird estimator, with statistical heterogeneity quantified using I&lt;sup&gt;2&lt;/sup&gt; and publication bias assessed using funnel plot asymmetry and Egger's test. Computational efficiency was standardised to GigaFLOPs using a reference input of 240 × 240 × 155 voxels (BraTS benchmark), with FLOP estimates derived from primary publications where available and bounded by theoretical complexity formulas otherwise, with estimated values explicitly distinguished throughout.&lt;/p&gt;&lt;p&gt;&lt;strong&gt;Results: &lt;/strong&gt;Across all 94 included studies, the pooled diagnostic accuracy was 93.5% (95% CI: 92.7%-94.4%); however, confirmed publication bias (Egger's &lt;i&gt;p&lt;/i&gt; = 0.043) indicates this represents an upper-bound approximation rather than an unbiased population estimate. Because subgroup study counts were insufficient for formal random-effects pooling (CNN-only: &lt;i&gt;n&lt;/i&gt; = 3; Transformer-only: &lt;i&gt;n&lt;/i&gt; = 2; CNN-Transformer hybrid: &lt;i&gt;n&lt;/i&gt; = 4; minimum recommended &lt;i&gt;n&lt;/i&gt; = 10 per subgroup), no subgroup meta-analysis was performed. Instead, descriptive mean accuracies are reported as hypothesis-generating observations only: CNN-only models 91.7%, Transformer-only models 93.6% and CNN-Transformer hybrid models 94.6%. These figures must not be interpreted as pooled meta-analytic estimates; they reflect mean observed accuracy across a small number of included studies and are reported solely to illustrate directional trends consistent with the mechanistic rationale for hybridisation. Substantial heterogeneity was observed (I&lt;sup&gt;2&lt;/sup&gt; = 78.3&lt;i&gt;%&lt;/i&gt;; &lt;i&gt;p&lt;/i&gt; &lt; 0.001). Three integration paradigms were identified: sequential (45% of models; 93.8% accuracy; 1.8 GFLOPs), parallel (32%; 94.3%; 2.8 GFLOPs) and hierarchical (23%; 94.9%; 3.5 GFLOPs","PeriodicalId":47063,"journal":{"name":"International Journal of Biomedical Imaging","volume":"2026 ","pages":"4763936"},"PeriodicalIF":2.3,"publicationDate":"2026-06-16","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC13270495/pdf/","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"148273390","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Brain Tumor Segmentation Using U-Net With ResNet50 Encoder for Enhanced MRI Analysis.","authors":"Abeer Masfer Alkahtani, Samia Dardouri","doi":"10.1155/ijbi/5318118","DOIUrl":"10.1155/ijbi/5318118","url":null,"abstract":"<p><p>Accurate brain tumor segmentation in magnetic resonance imaging (MRI) remains a challenging task due to the high variability in tumor appearance, shape, and location. Manual segmentation is time-consuming, subjective, and impractical for large-scale clinical use, highlighting the need for robust automated solutions. This study introduces an enhanced U-Net architecture with a ResNet50 encoder, designed to improve feature extraction through deeper convolutional layers and residual connections. By reformulating tumor delineation as a pixel-level segmentation problem rather than image-level classification, the model achieves more precise boundary detection. Trained on the publicly available TCGA-LGG dataset, the proposed model significantly outperformed the baseline U-Net, achieving a Dice score of 0.9659, an Intersection over Union (IoU) of 0.9567, and a Matthews correlation coefficient (MCC) of 0.9253. These results demonstrate superior segmentation capability compared to standard U-Net and are competitive with recent state-of-the-art methods. The findings highlight the potential of the proposed framework as a proof of concept for integration into clinical decision support, while also underscoring the need for future validation on larger, multi-institutional datasets.</p>","PeriodicalId":47063,"journal":{"name":"International Journal of Biomedical Imaging","volume":"2026 ","pages":"5318118"},"PeriodicalIF":2.3,"publicationDate":"2026-06-10","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC13250769/pdf/","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"148228650","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}