Journal of imaging informatics in medicine最新文献

{"title":"OcuViT: A Vision Transformer-Based Approach for Automated Diabetic Retinopathy and AMD Classification.","authors":"Faisal Ahmed, M D Joshem Uddin","doi":"10.1007/s10278-025-01676-3","DOIUrl":"https://doi.org/10.1007/s10278-025-01676-3","url":null,"abstract":"<p><p>Early detection and accurate classification of retinal diseases, such as diabetic retinopathy (DR) and age-related macular degeneration (AMD), are essential to preventing vision loss and improving patient outcomes. Traditional methods for analyzing retinal fundus images are often manual, prolonged, and rely on the expertise of the clinician, leading to delays in diagnosis and treatment. Recent advances in machine learning, particularly deep learning, have introduced automated systems to assist in retinal disease detection; however, challenges such as computational inefficiency and robustness still remain. This paper proposes a novel approach that utilizes vision transformers (ViT) through transfer learning to address challenges in ophthalmic diagnostics. Using a pre-trained ViT-Base-Patch16-224 model, we fine-tune it for diabetic retinopathy (DR) and age-related macular degeneration (AMD) classification tasks. To adapt the model for retinal fundus images, we implement a streamlined preprocessing pipeline that converts the images into PyTorch tensors and standardizes them, ensuring compatibility with the ViT architecture and improving model performance. We validated our model, OcuViT, on two datasets. We used the APTOS dataset to perform binary and five-level severity classification and the IChallenge-AMD dataset for grading age-related macular degeneration (AMD). In the five-class DR and AMD grading tasks, OcuViT outperforms all existing CNN- and ViT-based methods across multiple metrics, achieving superior accuracy and robustness. For the binary DR task, it delivers highly competitive performance. These results demonstrate that OcuViT effectively leverages ViT-based transfer learning with an efficient preprocessing pipeline, significantly improving the precision and reliability of automated ophthalmic diagnosis.</p>","PeriodicalId":516858,"journal":{"name":"Journal of imaging informatics in medicine","volume":" ","pages":""},"PeriodicalIF":0.0,"publicationDate":"2025-09-19","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145093383","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"MFFC-Net: Multi-feature Fusion Deep Networks for Classifying Pulmonary Edema of a Pilot Study by Using Lung Ultrasound Image with Texture Analysis and Transfer Learning Technique.","authors":"Ngoc Thang Bui, Charlie E Luoma, Xiaoming Zhang","doi":"10.1007/s10278-025-01673-6","DOIUrl":"https://doi.org/10.1007/s10278-025-01673-6","url":null,"abstract":"<p><p>Lung ultrasound (LUS) has been widely used by point-of-care systems in both children and adult populations to provide different clinical diagnostics. This research aims to develop an interpretable system that uses a deep fusion network for classifying LUS video/patients based on extracted features by using texture analysis and transfer learning techniques to assist physicians. The pulmonary edema dataset includes 56 LUS videos and 4234 LUS frames. The COVID-BLUES dataset includes 294 LUS videos and 15,826 frames. The proposed multi-feature fusion classification network (MFFC-Net) includes the following: (1) two features extracted from Inception-ResNet-v2, Inception-v3, and 9 texture features of gray-level co-occurrence matrix (GLCM) and histogram of the region of interest (ROI); (2) a neural network for classifying LUS images with feature fusion input; and (3) four models (i.e., ANN, SVM, XGBoost, and kNN) used for classifying COVID/NON COVID patients. The training process was evaluated based on accuracy (0.9969), F1-score (0.9968), sensitivity (0.9967), specificity (0.9990), and precision (0.9970) metrics after the fivefold cross-validation stage. The results of the ANOVA analysis with 9 features of LUS images show that there was a significant difference between pulmonary edema and normal lungs (p < 0.01). The test results at the frame level of the MFFC-Net model achieved an accuracy of 100% and ROC-AUC (1.000) compared with ground truth at the video level with 4 groups of LUS videos. Test results at the patient level with the COVID-BLUES dataset achieved the highest accuracy of 81.25% with the kNN model. The proposed MFFC-Net model has 125 times higher information density (ID) compared to Inception-ResNet-v2 and 53.2 times compared with Inception-v3.</p>","PeriodicalId":516858,"journal":{"name":"Journal of imaging informatics in medicine","volume":" ","pages":""},"PeriodicalIF":0.0,"publicationDate":"2025-09-19","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145093417","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Lightweight Transfer Learning Models for Multi-Class Brain Tumor Classification: Glioma, Meningioma, Pituitary Tumors, and No Tumor MRI Screening.","authors":"Alon Gorenshtein, Tom Liba, Avner Goren","doi":"10.1007/s10278-025-01686-1","DOIUrl":"https://doi.org/10.1007/s10278-025-01686-1","url":null,"abstract":"<p><p>Glioma, pituitary tumors, and meningiomas constitute the major types of primary brain tumors. The challenge in achieving a definitive diagnosis stem from the brain's complex structure, limited accessibility for precise imaging, and the resemblance between different types of tumors. An alternative and promising solution is the application of artificial intelligence (AI), specifically through deep learning models. We developed multiple lightweight deep learning models ResNet-18 (both pretrained on ImageNet and trained from scratch), ResNet-34, ResNet-50, and a custom CNN to classify glioma, meningioma, pituitary tumor, and no tumor MRI scans. A dataset of 7023 images was employed, split into 5712 for training and 1311 for validation. Each model was evaluated via accuracy, area under the curve (AUC), sensitivity, specificity, and confusion matrices. We compared our models to SOTA methods such as SAlexNet and TumorGANet, highlighting computational efficiency and classification performance. ResNet pretrained achieved 98.5-99.2% accuracy and near-perfect validation metrics, with an overall AUC of 1.0 and average sensitivity and specificity both exceeding 97% across the four classes. In comparison, ResNet-18 trained from scratch and the custom CNN achieved 91.99% and 87.03% accuracy, respectively, with AUCs ranging from 0.94 to 1.00. Error analysis revealed moderate misclassification of meningiomas as gliomas in non-pretrained models. Learning rate optimization facilitated stable convergence, and loss metrics indicated effective generalization with minimal overfitting. Our findings confirm that a moderately sized, transfer-learned network (ResNet-18) can deliver high diagnostic accuracy and robust performance for four-class brain tumor classification. This approach aligns with the goal of providing efficient, accurate, and easily deployable AI solutions, particularly for smaller clinical centers with limited computational resources. Future studies should incorporate multi-sequence MRI and extended patient cohorts to further validate these promising results.</p>","PeriodicalId":516858,"journal":{"name":"Journal of imaging informatics in medicine","volume":" ","pages":""},"PeriodicalIF":0.0,"publicationDate":"2025-09-19","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145093379","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"A New Method of Modeling the Multi-stage Decision-Making Process of CRT Using Machine Learning with Uncertainty Quantification.","authors":"Kristoffer Larsen, Chen Zhao, Zhuo He, Joyce Keyak, Qiuying Sha, Diana Paez, Xinwei Zhang, Guang-Uei Hung, Jiangang Zou, Amalia Peix, Weihua Zhou","doi":"10.1007/s10278-025-01666-5","DOIUrl":"10.1007/s10278-025-01666-5","url":null,"abstract":"<p><p>Current machine learning-based (ML) models usually attempt to utilize all available patient data to predict patient outcomes while ignoring the associated cost and time for data acquisition. The purpose of this study is to create a multi-stage ML model to predict cardiac resynchronization therapy (CRT) response for heart failure (HF) patients. This model exploits uncertainty quantification to recommend additional collection of single-photon emission computed tomography myocardial perfusion imaging (SPECT MPI) variables if baseline clinical variables and features from electrocardiogram (ECG) are not sufficient. Two hundred eighteen patients who underwent rest-gated SPECT MPI were enrolled in this study. CRT response was defined as an increase in left ventricular ejection fraction (LVEF) > 5% at a 6 ± 1 month follow-up. A multi-stage ML model was created by combining two ensemble models: Ensemble 1 was trained with clinical variables and ECG; Ensemble 2 included Ensemble 1 plus SPECT MPI features. Uncertainty quantification from Ensemble 1 allowed for multi-stage decision-making to determine if the acquisition of SPECT data for a patient is necessary. The performance of the multi-stage model was compared with that of Ensemble models 1 and 2. The response rate for CRT was 55.5% (n = 121) with overall male gender 61.0% (n = 133), an average age of 62.0 ± 11.8, and LVEF of 27.7 ± 11.0. The multi-stage model performed similarly to Ensemble 2 (which utilized the additional SPECT data) with AUC of 0.75 vs. 0.77, accuracy of 0.71 vs. 0.69, sensitivity of 0.70 vs. 0.72, and specificity 0.72 vs. 0.65, respectively. However, the multi-stage model only required SPECT MPI data for 52.7% of the patients across all folds. By using rule-based logic stemming from uncertainty quantification, the multi-stage model was able to reduce the need for additional SPECT MPI data acquisition without significantly sacrificing performance.</p>","PeriodicalId":516858,"journal":{"name":"Journal of imaging informatics in medicine","volume":" ","pages":""},"PeriodicalIF":0.0,"publicationDate":"2025-09-19","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145093410","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Optimising Generalisable Deep Learning Models for CT Coronary Segmentation: A Multifactorial Evaluation.","authors":"Shisheng Zhang, Ramtin Gharleghi, Sonit Singh, Chi Shen, Dona Adikari, Mingzi Zhang, Daniel Moses, Dominic Vickers, Arcot Sowmya, Susann Beier","doi":"10.1007/s10278-025-01677-2","DOIUrl":"https://doi.org/10.1007/s10278-025-01677-2","url":null,"abstract":"&lt;p&gt;&lt;p&gt;Coronary artery disease (CAD) remains a leading cause of morbidity and mortality worldwide, with incidence rates continuing to rise. Automated coronary artery medical image segmentation can ultimately improve CAD management by enabling more advanced and efficient diagnostic assessments. Deep learning-based segmentation methods have shown significant promise and offered higher accuracy while reducing reliance on manual inputs. However, achieving consistent performance across diverse datasets remains a persistent challenge due to substantial variability in imaging protocols, equipment and patient-specific factors, such as signal intensities, anatomical differences and disease severity. This study investigates the influence of image quality and resolution, governed by vessel size and common disease characteristics that introduce artefacts, such as calcification, on coronary artery segmentation accuracy in computed tomography coronary angiography (CTCA). Two datasets were utilised for model training and validation, including the publicly available ASOCA dataset (40 cases) and a GeoCAD dataset (70 cases) with more cases of coronary disease. Coronary artery segmentations were generated using three deep learning frameworks/architectures: default U-Net, Swin-UNETR, and EfficientNet-LinkNet. The impact of various factors on model generalisation was evaluated, focusing on imaging characteristics (contrast-to-noise ratio, artery contrast enhancement, and edge sharpness) and the extent of calcification at both the coronary tree and individual vessel branch levels. The calcification ranges considered were 0 (no calcification), 1-99 (low), 100-399 (moderate), and &gt; 400 (high). The findings demonstrated that image features, including artery contrast enhancement (r = 0.408, p &lt; 0.001) and edge sharpness (r = 0.239, p = 0.046), were significantly correlated with improved segmentation performance in test cases. Regardless of severity, calcification had a negative impact on segmentation accuracy, with low calcification affecting the segmentation most poorly (p &lt; 0.05). This may be because smaller calcified lesions produce less distinct contrast against the bright lumen, making it harder for the model to accurately identify and segment these lesions. Additionally, in males, a larger diameter of the first obtuse marginal branch (OM1) (p = 0.036) was associated with improved segmentation performance for OM1. Similarly, in females, larger diameters of left main (LM) coronary artery (p = 0.008) and right coronary artery (RCA) (p &lt; 0.001) were associated with better segmentation performance for LM and RCA, respectively. These findings emphasise the importance of accounting for imaging characteristics and anatomical variability when developing generalisable deep learning models for coronary artery segmentation. Unlike previous studies, which broadly acknowledge the role of image quality in segmentation, our work quantitatively demonstrates the extent to which contra","PeriodicalId":516858,"journal":{"name":"Journal of imaging informatics in medicine","volume":" ","pages":""},"PeriodicalIF":0.0,"publicationDate":"2025-09-18","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145088722","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Decision Strategies in AI-Based Ensemble Models in Opportunistic Alzheimer's Detection from Structural MRI.","authors":"Solveig Kristina Hammonds, Trygve Eftestøl, Kathinka Daehli Kurz, Alvaro Fernandez-Quilez","doi":"10.1007/s10278-025-01604-5","DOIUrl":"https://doi.org/10.1007/s10278-025-01604-5","url":null,"abstract":"<p><p>Alzheimer's disease (AD) is a neurodegenerative condition and the most common form of dementia. Recent developments in AD treatment call for robust diagnostic tools to facilitate medical decision-making. Despite progress for early diagnostic tests, there remains uncertainty about clinical use. Structural magnetic resonance imaging (MRI), as a readily available imaging tool in the current AD diagnostic pathway, in combination with artificial intelligence, offers opportunities of added value beyond symptomatic evaluation. However, MRI studies in AD tend to suffer from small datasets and consequently limited generalizability. Although ensemble models take advantage of the strengths of several models to improve performance and generalizability, there is little knowledge of how the different ensemble models compare performance-wise and the relationship between detection performance and model calibration. The latter is especially relevant for clinical translatability. In our study, we applied three ensemble decision strategies with three different deep learning architectures for multi-class AD detection with structural MRI. For two of the three architectures, the weighted average was the best decision strategy in terms of balanced accuracy and calibration error. In contrast to the base models, the results of the ensemble models showed that the best detection performance corresponded to the lowest calibration error, independent of the architecture. For each architecture, the best ensemble model reduced the estimated calibration error compared to the base model average from (1) 0.174±0.01 to 0.164±0.04, (2) 0.182±0.02 to 0.141±0.04, and (3) 0.269±0.08 to 0.240±0.04 and increased the balanced accuracy from (1) 0.527±0.05 to 0.608±0.06, (2) 0.417±0.03 to 0.456±0.04, and (3) 0.348±0.02 to 0.371±0.03.</p>","PeriodicalId":516858,"journal":{"name":"Journal of imaging informatics in medicine","volume":" ","pages":""},"PeriodicalIF":0.0,"publicationDate":"2025-09-17","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145082845","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Video Transformer for Segmentation of Echocardiography Images in Myocardial Strain Measurement.","authors":"Kuan-Chih Huang, Chang-En Lin, Donna Shu-Han Lin, Ting-Tse Lin, Cho-Kai Wu, Geng-Shi Jeng, Lian-Yu Lin, Lung-Chun Lin","doi":"10.1007/s10278-025-01682-5","DOIUrl":"https://doi.org/10.1007/s10278-025-01682-5","url":null,"abstract":"<p><p>The adoption of left ventricular global longitudinal strain (LVGLS) is still restricted by variability among various vendors and observers, despite advancements from tissue Doppler to speckle tracking imaging, machine learning, and, more recently, convolutional neural network (CNN)-based segmentation strain analysis. While CNNs have enabled fully automated strain measurement, they are inherently constrained by restricted receptive fields and a lack of temporal consistency. Transformer-based networks have emerged as a powerful alternative in medical imaging, offering enhanced global attention. Among these, the Video Swin Transformer (V-SwinT) architecture, with its 3D-shifted windows and locality inductive bias, is particularly well suited for ultrasound imaging, providing temporal consistency while optimizing computational efficiency. In this study, we propose the DTHR-SegStrain model based on a V-SwinT backbone. This model incorporates contour regression and utilizes an FCN-style multiscale feature fusion. As a result, it can generate accurate and temporally consistent left ventricle (LV) contours, allowing for direct calculation of myocardial strain without the need for conversion from segmentation to contours or any additional postprocessing. Compared to EchoNet-dynamic and Unity-GLS, DTHR-SegStrain showed greater efficiency, reliability, and validity in LVGLS measurements. Furthermore, the hybridization experiments assessed the interaction between segmentation models and strain algorithms, reinforcing that consistent segmentation contours over time can simplify strain calculations and decrease measurement variability. These findings emphasize the potential of V-SwinT-based frameworks to enhance the standardization and clinical applicability of LVGLS assessments.</p>","PeriodicalId":516858,"journal":{"name":"Journal of imaging informatics in medicine","volume":" ","pages":""},"PeriodicalIF":0.0,"publicationDate":"2025-09-17","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145082770","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"From Detection to Radiology Report Generation: Fine-Grained Multi-Modal Alignment with Semi-Supervised Learning.","authors":"Qian Tang, Lijun Liu, Xiaobing Yang, Li Liu, Wei Peng","doi":"10.1007/s10278-025-01650-z","DOIUrl":"https://doi.org/10.1007/s10278-025-01650-z","url":null,"abstract":"<p><p>Radiology report generation plays a critical role in supporting diagnosis, alleviating clinicians' workload, and improving diagnostic accuracy by integrating radiological image content with clinical knowledge. However, most existing models primarily establish coarse-grained mappings between global images and textual reports, often overlooking fine-grained associations between lesion regions and corresponding report content. This limitation affects the accuracy and clinical relevance of the generated reports. To address this, we propose D2R-Net, a lesion-aware radiology report generation model. D2R-Net leverages bounding box annotations for 22 chest diseases to guide the model to focus on clinically significant lesion regions. It employs a global-local dual-branch architecture that fuses global image context with localized lesion features and incorporates a Lesion Region Enhancement Module (LERA) to strengthen the recognition of key lesion regions. Additionally, an implicit alignment mechanism, including Local Alignment Blocks (LAB) and Global Alignment Blocks (GAB), is designed to bridge the semantic gap between visual and textual modalities. Experimental results on the benchmark MIMIC-CXR dataset demonstrate the superior performance of D2R-Net in generating accurate and clinically relevant radiology reports.</p>","PeriodicalId":516858,"journal":{"name":"Journal of imaging informatics in medicine","volume":" ","pages":""},"PeriodicalIF":0.0,"publicationDate":"2025-09-16","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145071370","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"AtTheViewBox: Scrolling Past PowerPoints to a Novel Web-Based Solution for Interactive Case-Based Presentations.","authors":"Michael Fei, Dane Van Tassel, Ianto Xi, Vineeth Gangaram","doi":"10.1007/s10278-025-01620-5","DOIUrl":"https://doi.org/10.1007/s10278-025-01620-5","url":null,"abstract":"<p><p>Traditional radiology education relies heavily on PowerPoint presentations with static 2D images, which fail to replicate the interactive nature of reading radiological studies at a workstation. There is a growing need for an interactive, case-based platform that enables real-time collaboration in presentations. This study introduces AtTheViewBox, a web-based application designed to integrate DICOM images into presentations, offering a more dynamic and interactive learning experience. AtTheViewBox was developed using open-source libraries, including React, CornerstoneJS, and Supabase. The application allows users to embed DICOM images in slide presentations via iframes, enabling standard functionalities at a radiology workstation like scrolling, zooming, and windowing. A survey was conducted among radiology residents and educators from four academic institutions to assess the utility and ease of use of AtTheViewBox compared to traditional teaching methods. Among 27 radiology residents surveyed, 100% agreed that AtTheViewBox would enhance their case-based learning experience, with 93% preferring it over static images or videos. Among 30 educators, the application received an average usefulness rating of 9.5/10. Additionally, 63% of educators found AtTheViewBox as easy or easier to use than their current methods. AtTheViewBox effectively modernizes radiology education by enabling interactive DICOM integration in presentations. This tool enhances learning by mimicking workstation experiences and fostering real-time collaboration. The overwhelmingly positive reception suggests that AtTheViewBox addresses key limitations in current teaching methodologies and has the potential to become a standard in radiology education.</p>","PeriodicalId":516858,"journal":{"name":"Journal of imaging informatics in medicine","volume":" ","pages":""},"PeriodicalIF":0.0,"publicationDate":"2025-09-16","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145077001","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Accuracy of AI-Based Algorithms in Pulmonary Embolism Detection on Computed Tomographic Pulmonary Angiography: An Updated Systematic Review and Meta-analysis.","authors":"Seyed Ali Nabipoorashrafi, Arsalan Seyedi, Razman Arabzadeh Bahri, Amirhossein Yadegar, Mostafa Shomal-Zadeh, Fatemeh Mohammadi, Samira Amin Afshari, Negar Firoozeh, Navida Noroozzadeh, Farbod Khosravi, Sanaz Asadian, Hamid Chalian","doi":"10.1007/s10278-025-01645-w","DOIUrl":"https://doi.org/10.1007/s10278-025-01645-w","url":null,"abstract":"<p><p>Several artificial intelligence (AI) algorithms have been designed for detection of pulmonary embolism (PE) using computed tomographic pulmonary angiography (CTPA). Due to the rapid development of this field and the lack of an updated meta-analysis, we aimed to systematically review the available literature about the accuracy of AI-based algorithms to diagnose PE via CTPA. We searched EMBASE, PubMed, Web of Science, and Cochrane for studies assessing the accuracy of AI-based algorithms. Studies that reported sensitivity and specificity were included. The R software was used for univariate meta-analysis and drawing summary receiver operating characteristic (sROC) curves based on bivariate analysis. To explore the source of heterogeneity, sub-group analysis was performed (PROSPERO: CRD42024543107). A total of 1722 articles were found, and after removing duplicated records, 1185 were screened. Twenty studies with 26 AI models/population met inclusion criteria, encompassing 11,950 participants. Univariate meta-analysis showed a pooled sensitivity of 91.5% (95% CI 85.5-95.2) and specificity of 84.3 (95% CI 74.9-90.6) for PE detection. Additionally, in the bivariate sROC, the pooled area under the curved (AUC) was 0.923 out of 1, indicating a very high accuracy of AI algorithms in the detection of PE. Also, subgroup meta-analysis showed geographical area as a potential source of heterogeneity where the I² for sensitivity and specificity in the Asian article subgroup were 60% and 6.9%, respectively. Findings highlight the promising role of AI in accurately diagnosing PE while also emphasizing the need for further research to address regional variations and improve generalizability.</p>","PeriodicalId":516858,"journal":{"name":"Journal of imaging informatics in medicine","volume":" ","pages":""},"PeriodicalIF":0.0,"publicationDate":"2025-09-15","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145070741","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}