<?xml version="1.0" encoding="utf-8"?>
    <rss version="2.0">
      <channel xmlns:content="http://purl.org/rss/1.0/modules/content/">
        <title>Frontiers in Radiology | New and Recent Articles</title>
        <link>https://www.frontiersin.org/journals/radiology</link>
        <description>RSS Feed for Frontiers in Radiology | New and Recent Articles</description>
        <language>en-us</language>
        <generator>Frontiers Feed Generator,version:1</generator>
        <pubDate>2026-05-09T11:53:55.412+00:00</pubDate>
        <ttl>60</ttl>
        <item>
        <guid isPermaLink="true">https://www.frontiersin.org/articles/10.3389/fradi.2026.1746296</guid>
        <link>https://www.frontiersin.org/articles/10.3389/fradi.2026.1746296</link>
        <title><![CDATA[Intravenous contrast medium impairs CT-based muscle quality but not quantity assessment: a translational study]]></title>
        <pubdate>2026-05-08T00:00:00Z</pubdate>
        <category>Original Research</category>
        <author>Luca Salhöfer</author><author>Gregor Jost</author><author>Mathias Holtkamp</author><author>Jannis Straus</author><author>Marcel Opitz</author><author>Sebastian Zensen</author><author>Rene Hosch</author><author>Johannes Harmes</author><author>Lale Umutlu</author><author>Michael Forsting</author><author>Felix Nensa</author><author>Hubertus Pietsch</author><author>Johannes Haubold</author>
        <description><![CDATA[ObjectivesCT-based body composition analysis (BCA) provides imaging biomarkers, including muscle volume and surrogates of muscle quality. Concerns over the comparability of Non-contrast and contrast-enhanced CT scans have limited their clinical application. This study aims to assess the influence of various contrast phases on a volumetric CT-based BCA.Materials and methods20 Göttingen minipigs were subjected to a Non-contrast (NC) and five contrast-enhanced [Early Arterial, Late Arterial, Vascular Portal Venous, Parenchymal Portal Venous (PPV), Late] CT scans. 114 tri-phasic (Non-Contrast, Arterial, Venous) CT scans were analyzed for human validation. A volumetric BCA network [Body and Organ Analysis (BOA)] extracted muscle radiodensity and the following features as volumes: Muscle, Subcutaneous Adipose Tissue (SAT), Inter- and Intramuscular Adipose Tissue (IMAT), Visceral Adipose Tissue (VAT), and Total Adipose Tissue (TAT). Significance was assessed by a one-way ANOVA with Tukey's multiple comparisons test.ResultsIn the animal model, there was a tendency toward reduced IMAT volumes after CM injection [NC = 245 ml (±105 ml), e.g., PPV = 241 ml (±105 ml)]. Muscle radiodensity was significantly higher following CM administration [Non-contrast: 51.1 HU (±1.9 HU), e.g., Late: 56.6 HU (±2.4 HU), p < 0.001]. The human validation analysis showed similar tendencies for the IMAT volume [Non-contrast: 1750ml (±729 ml), Venous: 1,552 ml (±696 ml), p = 0.10] and significantly higher muscle radiodensity [Venous: 39.2 HU (±9.1 HU), Non-contrast: 35.6 HU (±7.8), p = 0.007].ConclusionMyosteatosis surrogates, such as muscle radiodensity or IMAT, are susceptible to interference by CM, while quantification of muscle tissue and extramuscular fat remains robust.]]></description>
      </item><item>
        <guid isPermaLink="true">https://www.frontiersin.org/articles/10.3389/fradi.2026.1809871</guid>
        <link>https://www.frontiersin.org/articles/10.3389/fradi.2026.1809871</link>
        <title><![CDATA[Multimodal brain MRI and clinical data in olfactory groove meningioma: a prospective data report]]></title>
        <pubdate>2026-05-08T00:00:00Z</pubdate>
        <category>Data Report</category>
        <author>Elena Filimonova</author><author>Anton Pashkov</author><author>Galina Moysak</author><author>Azniv Martirosyan</author><author>Vladimir Kurilov</author><author>Aleksandra Poptsova</author><author>Renata Morozova</author><author>Jamil Rzaev</author>
        <description><![CDATA[Olfactory groove meningiomas are uncommon skull base tumors that often present at advanced stages and may cause persistent cognitive and behavioral disturbances despite successful surgical treatment. Neuroimaging studies of this tumor entity have been limited, and publicly available multimodal MRI datasets remain scarce. Here, we present a prospective, single-center dataset comprising multimodal magnetic resonance imaging and longitudinal clinical data from patients with olfactory groove meningiomas acquired before and after surgical intervention. The dataset includes high-resolution structural MRI, diffusion MRI with tensor-derived metrics, resting-state functional MRI, tumor and peritumoral edema segmentation masks, and detailed clinical and neuropsychological assessments. Imaging data were acquired using a standardized protocol and processed with reproducible pipelines, including quality control and de-identification procedures, and organized in a BIDS format. This dataset is intended to support reproducible research and secondary analyses focused on tumor-related brain alterations, imaging biomarker development, and postoperative recovery in neuro-oncology.]]></description>
      </item><item>
        <guid isPermaLink="true">https://www.frontiersin.org/articles/10.3389/fradi.2026.1691258</guid>
        <link>https://www.frontiersin.org/articles/10.3389/fradi.2026.1691258</link>
        <title><![CDATA[Enhancing low-dose CT denoising via multi-view knowledge transfer without paired data]]></title>
        <pubdate>2026-05-08T00:00:00Z</pubdate>
        <category>Original Research</category>
        <author>Yueyang You</author><author>Li Xu</author><author>Gen Wei</author><author>Fuzhou Hua</author><author>Yingbo Hu</author>
        <description><![CDATA[IntroductionDespite remarkable advancements in deep learning for low-dose computed tomography (LDCT) denoising, two significant challenges persist: (i) the requirement for paired LDCT and high-dose computed tomography (HDCT) images, which are often impractical to obtain in clinical settings; and (ii) the tendency of existing methods to train models using a single axial view, thereby overlooking complementary information from other views and consequently limiting their performance.MethodsTo address these issues, we propose a Multi-view-to-Single Knowledge Transfer (MvSKT) framework for unpaired LDCT denoising. Our approach involves splitting the 3D unpaired computed tomography (CT) data into 2D images from various views, including axial, sagittal, and coronal. This allows us to train three view-independent 2D GAN models in an unsupervised manner. By stacking successive 2D outputs from each view-independent model into a volumetric format and splitting them into axial-view images, we generate multiple complementary predictions for each axial CT image. Leveraging these predictions as priors, we transfer multi-view knowledge to a single-view model through pseudo-supervision. This process involves fusing multiple view-complementary predictions into reliable pseudo-images using a cycle-consistency-weighted method.ResultsExtensive experiments on the AAPM-Mayo dataset demonstrate that MvSKT outperforms other unpaired denoising methods and even achieves performance comparable to supervised approaches.DiscussionConsequently, the MvSKT framework effectively harnesses multi-view information from unpaired data to enhance LDCT denoising without the strict requirement of paired clinical data.]]></description>
      </item><item>
        <guid isPermaLink="true">https://www.frontiersin.org/articles/10.3389/fradi.2026.1821920</guid>
        <link>https://www.frontiersin.org/articles/10.3389/fradi.2026.1821920</link>
        <title><![CDATA[A preoperative nomogram for predicting 2-year postoperative recurrence after percutaneous transforaminal endoscopic decompression in degenerative lumbar spinal stenosis]]></title>
        <pubdate>2026-05-04T00:00:00Z</pubdate>
        <category>Original Research</category>
        <author>Xinyi Luo</author><author>Yiwen Wang</author><author>Lele Xue</author><author>Qi Liu</author><author>Yue Yang</author><author>Qin Zhang</author><author>Shiwu Yin</author>
        <description><![CDATA[Background and objectivesPostoperative recurrence after percutaneous transforaminal endoscopic decompression (PTED) for degenerative lumbar spinal stenosis (DLSS) remains a clinically relevant challenge, complicating preoperative counseling and long-term management. Reliable tools for predicting individual 2-year recurrence risk using routinely available preoperative data are currently lacking. This study aimed to develop and internally validate a practical preoperative nomogram for individualized recurrence risk prediction after PTED.MethodsWe conducted a retrospective cohort study including 206 patients with DLSS who underwent single-level PTED between August 2021 and August 2023. Preoperative clinical and imaging variables were extracted to construct a multivariable logistic regression model. Candidate predictors were prespecified based on clinical relevance and routine availability. Model performance was evaluated in terms of discrimination, calibration, and clinical utility. Internal validation was performed using 1000 bootstrap resamples and leave-one-out cross-validation (LOOCV).ResultsDuring the 2-year follow-up period, 29 patients (14.08%) experienced postoperative recurrence. The final nomogram incorporated five preoperative predictors: body mass index, diabetes mellitus, lumbosacral transitional vertebrae, number of levels with senior grade facet degeneration, and paraspinal skeletal muscle index. The model showed good discrimination, with an area under the receiver operating characteristic curve (AUC) of 0.845 (95% CI, 0.778–0.912). Bootstrap validation showed a mean AUC of 0.842 (95% CI, 0.772–0.912), and LOOCV yielded an AUC of 0.797 (95% CI, 0.716–0.878). Calibration was satisfactory, and decision curve analysis demonstrated net clinical benefit across a wide range of threshold probabilities.ConclusionsWe developed a clinically interpretable preoperative nomogram that reliably predicts 2-year postoperative recurrence after PTED in patients with DLSS. By integrating routinely assessed clinical and imaging factors, this tool may facilitate individualized risk stratification, support informed preoperative counseling, and guide risk-adapted perioperative management. External validation in independent cohorts is warranted.]]></description>
      </item><item>
        <guid isPermaLink="true">https://www.frontiersin.org/articles/10.3389/fradi.2026.1802335</guid>
        <link>https://www.frontiersin.org/articles/10.3389/fradi.2026.1802335</link>
        <title><![CDATA[Prenatal MRI features of fetal complete agenesis of the corpus callosum associated with unilateral hemispheric cortical malformation: a retrospective study]]></title>
        <pubdate>2026-05-01T00:00:00Z</pubdate>
        <category>Original Research</category>
        <author>Huihui Lin</author><author>Xiaoyu Wang</author><author>Chang Wang</author><author>Gengwu Li</author><author>Xu Li</author>
        <description><![CDATA[IntroductionTo explore the prenatal MRI features of fetal complete agenesis of the corpus callosum (cACC) associated with unilateral hemispheric cortical malformation.MethodsA retrospective analysis was conducted on 101 cases diagnosed with cACC via prenatal MRI. Among them, 16 cases were found to be associated with unilateral hemispheric cortical malformation. The imaging characteristics of this specific malformation were analyzed, and its associations with fetal gender and gestational age were investigated.ResultsIn this group of 16 cases of cACC with unilateral cortical malformation, 13 fetuses were male and 3 were female. The gestational age at diagnosis ranged from 23 weeks to 27 weeks and 2 days. The malformation was located in the left hemisphere in 7 cases and in the right hemisphere in 9 cases. Type I (C1 type: extensive cortical twisting/folding) was observed in 11 cases, all of which exhibited the “rake sign” and “garland sign”. The “rake sign” manifests as abnormally folded dysplastic cortex resembling a rake on axial or coronal views, while the “garland sign” appears as abnormally folded dysplastic cortex resembling a garland on sagittal views. Type II (C4 type: transcortical cleft formation) was observed in 1 case, and Type III (C5 type: focal cortical indentation or serrated changes) was observed in 4 cases. The unilateral cortical malformation involved the frontal lobe in 16 cases, the parietal lobe in 10 cases, and the occipital lobe in 8 cases, with no involvement of the temporal lobe.DiscussionFetal agenesis of the corpus callosum has a relatively high probability of being associated with unilateral cortical malformation. This malformation occurs more frequently in males, is mostly detected during the mid-trimester, and most commonly involves the frontal lobe. The “rake sign” and “garland sign” are its characteristic imaging features. Prenatal definitive diagnosis aids in the perinatal management of this malformation. Prenatal MRI is a crucial supplementary examination for fetal cACC diagnosed by prenatal ultrasound, as it can accurately detect associated unilateral hemispheric cortical malformations that are easily missed by ultrasound, and thus is essential for optimizing perinatal management and parental genetic counseling.]]></description>
      </item><item>
        <guid isPermaLink="true">https://www.frontiersin.org/articles/10.3389/fradi.2026.1788985</guid>
        <link>https://www.frontiersin.org/articles/10.3389/fradi.2026.1788985</link>
        <title><![CDATA[Anatomical characteristics of the styloid process associated with internal carotid artery dissection: a systematic review and meta-analysis of controlled trials]]></title>
        <pubdate>2026-05-01T00:00:00Z</pubdate>
        <category>Systematic Review</category>
        <author>Tomas Klail</author><author>Emmanouil Kalioras</author><author>Marc von Gernler</author><author>Roland Giger</author><author>Aristomenis K. Exadaktylos</author><author>Martin Müller</author><author>Franca Wagner</author>
        <description><![CDATA[PurposeInternal carotid artery dissection (ICA-D) is an important cause of stroke in adults. The styloid process (SP) may be associated with ICA-D due to potential (micro)trauma resulting from its close anatomical proximity to the internal carotid artery (ICA). The aim of this systematic review with meta-analysis is to investigate the association between SP characteristics –particularly SP-ICA distance – and ICA-D.MethodsA systematic review was conducted across six databases to identify observational studies comparing ICA-D patients to controls. The primary outcome of interest was the association between SP-ICA distance and ICA-D. Secondary outcomes included associations between ICA-D and the SP length or angulation. A random-effects meta-analysis was performed, including a subgroup analysis of moderate/high-quality studies. Effect sizes were expressed as standardized mean differences (SMD, Hedges' g).ResultsSix studies were included in the systematic review, of which five were eligible in the meta-analysis. The pooled analysis of all five case-control studies (270 ICA-D patients and 377 controls) showed no significant difference in SP-ICA distance (SMD = −0.92, p = 0.143); with a high degree of heterogeneity (I2 = 98%). Subgroup analysis of moderate/high-quality studies evaluating the SP-ICA distance ipsilateral to the ICA-D (4 studies) yielded a negative pooled SMD (−0.29, p = 0.047; moderate heterogeneity: I2 = 64%), consistent with a shorter SP-ICA distance in ICA-D cases. Meta-analysis of the SP length (3 studies) found no significant association (SMD 0.24, p = 0.139) and two studies also found no significant relationship between ICA-D and SP angulation.ConclusionA shorter SP–ICA distance was associated with ICA-D, whereas no significant associations were observed for SP length or angulation. However, the available evidence remains limited and heterogeneous.Systematic Review RegistrationCRD42024582594]]></description>
      </item><item>
        <guid isPermaLink="true">https://www.frontiersin.org/articles/10.3389/fradi.2026.1764357</guid>
        <link>https://www.frontiersin.org/articles/10.3389/fradi.2026.1764357</link>
        <title><![CDATA[Deep learning-based denoising in cardiac CT: effects on image quality, calcium scoring interchangeability, and reporting workflow]]></title>
        <pubdate>2026-04-29T00:00:00Z</pubdate>
        <category>Original Research</category>
        <author>Daniel Wessling</author><author>Jan Magnus</author><author>Jan M. Brendel</author><author>Sebastian Werner</author><author>Patrick Krumm</author><author>Konstantin Nikolaou</author><author>Saif Afat</author><author>Andreas S. Brendlin</author>
        <description><![CDATA[ObjectivesIschemic heart disease is a major global health burden requiring timely diagnosis. Although coronary artery calcium (CAC) scoring and coronary computed tomography angiography (CCTA) are valuable tools, image noise can distort assessment. Deep learning-based denoising (DLD) algorithms may enhance quality, yet their impact on cardiac CT workflows remains unclear. This study evaluated DLD effects on CAC and CCTA image quality, clinical interchangeability, and workflow efficiency compared with iterative reconstruction (IR).Materials and methodsA retrospective analysis of 100 patients with CAC and CCTA scans from the same CT scanner was performed. IR and DLD reconstructions yielded 400 datasets, rated by two radiologists using a semiquantitative scoring system. Objective metrics (CT number stability, noise, contrast-to-noise ratio) were measured. Clinical interchangeability and Workflow efficiency were evaluated.ResultsDLD showed significantly higher overall image quality than IR (p < 0.001), while preserving CT attenuation and improving noise and CNR. Although cardiac age classifications did not differ between reconstruction methods, Agatston scores were significantly higher in IR before manual correction (p < 0.001. After manual correction, Agatston scores were not significantly different between IR and DLD (p ≥ 0.158), supporting clinical comparability of the derived clinical metrics between the two approaches. DLD also reduced manual correction time (p < 0.001).ConclusionsThe investigated DLD algorithm improves image quality and radiological workflows thus potentially enhancing overall patient care. Key limitations include the retrospective single-center design and inherent subjectivity in image-quality evaluation; therefore, findings should be confirmed in prospective studies with more objective measures.]]></description>
      </item><item>
        <guid isPermaLink="true">https://www.frontiersin.org/articles/10.3389/fradi.2026.1763042</guid>
        <link>https://www.frontiersin.org/articles/10.3389/fradi.2026.1763042</link>
        <title><![CDATA[Vietnamese expert consensus on the treatment of hepatocellular carcinoma with transarterial chemoembolization]]></title>
        <pubdate>2026-04-21T00:00:00Z</pubdate>
        <category>Policy and Practice Reviews</category>
        <author>Vu Dang Luu</author><author>Do Dang Tan</author><author>Pham Minh Thong</author><author>Le Van Khang</author><author>Le Thanh Dung</author><author>Le Trong Binh</author><author>Nguyen Ngoc Trang</author><author>Nguyen Ngoc Cuong</author><author>Nguyen Dinh Luan</author><author>Ngo Le Lam</author><author>Trinh Ha Chau</author><author>Than Van Sy</author><author>Nguyen Duy Anh</author><author>Le Duc Tho</author><author>Tran Duc Huy</author><author>Dang Ngoc Hieu</author><author>Pham Quang Son</author><author>Le Doan Tri</author><author>Nguyen Dinh Huong</author><author>Pham Cam Phuong</author><author>Nguyen Tien Thinh</author><author>Vo Duy Thong</author><author>Ho Tan Phat</author><author>Phan Thi Hong Duc</author><author>Pham Gia Anh</author><author>Vo Hoi Trung Truc</author><author>Lam Quoc Trung</author><author>Le Ba Thao</author><author>Le Tuan Anh</author><author>Tran Cong Duy Long</author>
        <description><![CDATA[Hepatocellular carcinoma (HCC) remains a major public health challenge in Vietnam and, according to GLOBOCAN 2022, accounted for 24,502 new cases in 2022, making it the second most common cancer and the leading cause of cancer-related mortality. Transarterial chemoembolization (TACE), introduced in 1977, is a pivotal treatment for unresectable HCC, delivering chemotherapeutic agents and embolic materials to induce tumor necrosis through cytotoxicity and ischemia. TACE is versatile and, in selected early-stage HCC (BCLC 0-A) treated with a highly selective approach, may achieve complete response; in intermediate-stage disease (BCLC B), it can prolong survival and enable downstaging, and in advanced-stage disease (BCLC C), it may complement systemic therapy. Despite its global use, TACE practices vary, and Vietnam lacks standardized guidelines tailored to its unique clinical landscape. The Vietnamese Society of Radiology and Nuclear Medicine and The Vietnamese Society of Interventional Radiology conducted three workshops in Da Nang, An Giang, and Can Tho between August and December 2024. These workshops engaged interventional radiologists from major hospitals and regional centers to develop consensus-based TACE recommendations. Through literature reviews, discussions, and physician voting, the recommendations cover pre-TACE diagnostic imaging, patient selection criteria, TACE techniques, choice of embolic materials and chemotherapeutic agents, intra- and post-procedure monitoring, and integration with combination therapies. This initiative aligns international standards with local needs, aiming to enhance the efficacy and consistency of TACE for HCC management in Vietnam.]]></description>
      </item><item>
        <guid isPermaLink="true">https://www.frontiersin.org/articles/10.3389/fradi.2026.1782678</guid>
        <link>https://www.frontiersin.org/articles/10.3389/fradi.2026.1782678</link>
        <title><![CDATA[Development and validation of a nomogram using interpretable machine learning to integrate CT radiomics and PET metabolic parameters for predicting benign-malignant differentiation of pulmonary space-occupying lesions]]></title>
        <pubdate>2026-04-21T00:00:00Z</pubdate>
        <category>Original Research</category>
        <author>Xue Liu</author><author>Xinghua Liu</author><author>Li Bin</author><author>Yu Zhang</author><author>Huiting Liu</author><author>Cailiang Gao</author>
        <description><![CDATA[ObjectiveTo construct a multimodal machine learning model integrating computed tomography (CT) radiomics, Positron Emission Tomography (PET) metabolic parameters, and clinical data for differentiating benign from malignant pulmonary space-occupying lesions (PSOLs), and develop an interpretable nomogram for clinical application.MethodologyThis study enrolled 384 patients with PSOLs who underwent dual-time-point 1⁸F-FDG PET/CT examinations. The cohort was divided into a training set (n = 268, 145 malignant, 123 benign) and an independent temporal validation set (n = 116, 69 malignant, 47 benign) at a 7:3 ratio according to the chronological order of patient enrollment, to avoid data leakage and rigorously assess model generalizability. All malignant lesions were confirmed by pathological examination, while benign lesions were confirmed by pathology (82%) or clinical-imaging follow-up for at least 12 months (18%). CT radiomic features with Intraclass Correlation Coefficient (ICC) values >0.75 were selected, and a Radiomics-score (Rad-score) was generated using the Least Absolute Shrinkage and Selection Operator (LASSO) regression algorithm. Integrated models [Logistic regression, random forest (RF), support vector machine (SVM), eXtreme Gradient Boosting (XGBoost)] were developed by fusing the Rad-score, clinical variables, and PET metabolic parameters. Model performance was evaluated using the area under the receiver operating characteristic curve (AUC), accuracy, sensitivity, specificity, F1-score, and Brier score. Model calibration was assessed via calibration curves, and clinical utility was validated by decision curve analysis (DCA). Model interpretability was achieved using SHapley Additive exPlanations (SHAP) values for the optimal XGBoost model, and a clinically applicable, interpretable nomogram was constructed based on the core predictive features identified by SHAP analysis to facilitate clinical translation.ResultsA Rad-score was constructed from 17 optimally selected features. In the independent temporal validation set, the single-modality models achieved AUCs of 0.808 (Radiomics Model), 0.732 (Clinical Model), and 0.874 (Metabolic Model). Among all tested models, the XGBoost integrated model achieved the highest AUC of 0.967, which was significantly higher than that of all other models (Bonferroni-adjusted P = 0.002–0.032, all adjusted P < 0.05). SHAP analysis identified ΔSUVmax, Rad-score, and delayed phase total lesion glycolysis (TLG_d) as the top three key predictive features.ConclusionsThe predictive logic of the optimal XGBoost model was decoded via SHAP analysis to identify core predictive features, and a clinically applicable, interpretable nomogram was further established based on a multivariate logistic regression model using these core features, to facilitate the clinical translation of our model.]]></description>
      </item><item>
        <guid isPermaLink="true">https://www.frontiersin.org/articles/10.3389/fradi.2026.1767875</guid>
        <link>https://www.frontiersin.org/articles/10.3389/fradi.2026.1767875</link>
        <title><![CDATA[YOLO11-based detection of manometry sensors in video-fluoroscopy imaging for computer-aided multimodal assessment of swallowing]]></title>
        <pubdate>2026-04-14T00:00:00Z</pubdate>
        <category>Original Research</category>
        <author>Dionne S. Brandsma</author><author>Manuel Maria Loureiro da Rocha</author><author>Lisette van der Molen</author><author>Maarten J. A. van Alphen</author><author>Michiel W. M. van den Brekel</author><author>Françoise J. Siepel</author>
        <description><![CDATA[PurposeAccurate assessment of swallowing function is essential in the diagnosis and monitoring of dysphagia following head and neck cancer (HNC). The simultaneous analysis of video-fluoroscopy swallow studies (VFSS) and high-resolution impedance manometry (HRIM) offers a more comprehensive evaluation, reducing subjectivity in VFSS and improving anatomical context of HRIM in HNC patients.MethodsThe inherently low pharyngeal pressures in post-treatment HNC patients hinder the analysis of HRIM. As such, this study proposes a deep learning method for the automatic detection of HRIM sensors in VFSS using a YOLO11-based detector, aimed at enabling the automatic delineation of manometric regions. Detection performance was evaluated on 268 frames from 8 HNC patients using a leave-one-patient-out cross-validation approach. EigenCAM-based heatmaps were produced to analyze the model’s attention patterns.ResultsThe model achieved 95.8% Precision, 97.4% Recall, 96.6% F1-score with minimal variation between folds. Under different noise levels and bolus-simulated obstructions, performance remained robust. Our method outperformed previous template-matching methods for manometric sensor detection in VFSS. EigenCAM visualizations confirmed consistent attention to catheter regions.ConclusionThe proposed YOLO11-based detector provides accurate and robust localization of manometric sensors in VFSS sequences to facilitate computer-assisted HRIM-VFSS fusion for objective swallowing assessment.]]></description>
      </item><item>
        <guid isPermaLink="true">https://www.frontiersin.org/articles/10.3389/fradi.2026.1793604</guid>
        <link>https://www.frontiersin.org/articles/10.3389/fradi.2026.1793604</link>
        <title><![CDATA[Combining computed tomography features and histogram analysis facilitates differentiation between solitary pulmonary invasive mucinous and non-mucinous adenocarcinomas]]></title>
        <pubdate>2026-04-14T00:00:00Z</pubdate>
        <category>Original Research</category>
        <author>Duiming Yang</author><author>Hao Hu</author><author>Zhimei Li</author><author>Shuilan Zhang</author><author>Jing Ai</author><author>Handan Zhang</author>
        <description><![CDATA[ObjectiveTo assess the value of combining computed tomography (CT) morphological and histogram features for the differentiation of solitary pulmonary invasive mucinous adenocarcinoma (PIMA) from pulmonary invasive non-mucinous adenocarcinoma (PINMA).MethodsThis retrospective study analyzed the CT images and clinical data of 58 and 105 patients with PIMA and PINMA, respectively. CT histogram features were extracted after delineating regions of interest using 3D Slicer software. CT morphological and histogram features were compared between the PIMA and PINMA groups, and those that differed significantly were included in multivariate logistic regression models. The independent predictive factors identified were used to create CT morphological, CT histogram-based, and combined prediction models. The best-performing model was visualized and evaluated by constructing a nomogram.ResultsThe CT morphological prediction model included nodule type, vacuole sign, and tumor location as factors predictive of PIMA and had an area under the curve of 0.754. The CT histogram-based prediction model included kurtosis and the 90th percentile as factors predictive of PIMA and had an area under the curve of 0.820. The combined prediction model, which included tumor location, vacuole sign, kurtosis, and the 90th percentile, had an area under the curve of 0.845, suggesting greater diagnostic accuracy than the separate models. The combined prediction model also exhibited good calibration and high clinical applicability.ConclusionIntegrating CT morphological features and histogram analysis improves the accuracy of differentiating PIMA from PINMA. The nomogram provides a practical and effective tool for the non-invasive diagnosis of lung cancer subtypes.]]></description>
      </item><item>
        <guid isPermaLink="true">https://www.frontiersin.org/articles/10.3389/fradi.2026.1822303</guid>
        <link>https://www.frontiersin.org/articles/10.3389/fradi.2026.1822303</link>
        <title><![CDATA[Coronary artery calcium scoring in 2026: strengths, limitations, and optimized clinical use]]></title>
        <pubdate>2026-04-13T00:00:00Z</pubdate>
        <category>Mini Review</category>
        <author>Pierre Sabouret</author><author>Domenico Mario Giamundo</author><author>Julien Rosencher</author><author>Stefano Figliozzi</author>
        <description><![CDATA[Coronary artery calcium (CAC) scoring on non-contrast ECG-gated CT remains a robust, reproducible marker of total coronary atherosclerotic burden with clear prognostic value and consistent risk reclassification beyond contemporary clinical calculators. Recent studies (2018–2026) reinforce the ‘power of zero’ for near-term risk de-escalation, identify very high CAC (≥1,000) as a distinct, very-high-risk phenotype, and, importantly, provide randomized evidence that CAC-guided treatment reduces plaque progression. Advances in artificial intelligence (AI), spectral CT, and standardized reporting (SCCT/STR; CAC-DRS) expand opportunities for automated and incidental CAC detection. This mini-review summarizes updated strengths, limitations, and practice guidance; synthesizes new evidence (2024–2026); and includes a practical clinical decision flowchart for the selective use of CAC in prevention pathways.]]></description>
      </item><item>
        <guid isPermaLink="true">https://www.frontiersin.org/articles/10.3389/fradi.2026.1798348</guid>
        <link>https://www.frontiersin.org/articles/10.3389/fradi.2026.1798348</link>
        <title><![CDATA[CT-based subchondral bone microstructural analysis in knee osteoarthritis via MR-guided distillation learning]]></title>
        <pubdate>2026-04-10T00:00:00Z</pubdate>
        <category>Original Research</category>
        <author>Yuqi Hu</author><author>Xiangyu Zhao</author><author>Gaowei Qing</author><author>Kai Xie</author><author>Chenglei Liu</author><author>Lichi Zhang</author>
        <description><![CDATA[Reliable analysis of subchondral trabecular microstructure is critical for knee osteoarthritis assessment. However, this analysis largely relies on high-resolution MRI acquired using balanced fast field echo (BFFE) sequences, which are rarely included in routine clinical protocols. Clinical CT is widely acquired, yet its limited spatial resolution and soft-tissue contrast makes direct trabecular parameter estimation unreliable. Therefore, it is specifically demanded to enable accurate trabecular microstructural analysis and osteoarthritis diagnosis using routine clinical CT, while also approaching the reliability of MR-based analysis. In this paper, we propose CT-based Subchondral Microstructural Analysis (CT-SMA) method, which utilizes distillation learning technology to transfer high-resolution structural knowledge from MR to CT while enforcing CT-only inference. The core idea of CT-SMA is to transfer microstructural knowledge learned from high-resolution MR to CT through cross-modal knowledge distillation, using a pre-trained MR-based teacher model to supervise CT-based student model on feature maps. To support effective distillation, CT-SMA further introduces a synthesis-based, multi-stage MR–CT registration strategy that establishes patch-level correspondences across modalities, despite substantial differences in resolution, contrast, and appearance. Experiments on a clinical knee imaging cohort demonstrate that CT-SMA substantially improves CT-based trabecular parameter estimation, achieving strong agreement (ICC = 0.742) with MR-derived references across key trabecular biomarkers. Moreover, when aggregated using a Transformer-based model, the regressed CT-derived parameters enable patient-level osteoarthritis diagnosis with an AUC of 0.883, substantially outperforming CT-based prediction without distillation (AUC = 0.778). These results indicate that routine clinical CT can support reliable subchondral bone analysis via proposed CT-SMA, establishing a practical foundation for large-scale studies.]]></description>
      </item><item>
        <guid isPermaLink="true">https://www.frontiersin.org/articles/10.3389/fradi.2026.1737075</guid>
        <link>https://www.frontiersin.org/articles/10.3389/fradi.2026.1737075</link>
        <title><![CDATA[Systematic review of pituitary gland and pituitary adenoma automatic segmentation techniques in magnetic resonance imaging]]></title>
        <pubdate>2026-04-10T00:00:00Z</pubdate>
        <category>Systematic Review</category>
        <author>Mubaraq Yakubu</author><author>Navodini Wijithilake</author><author>Jonathan Shapey</author><author>Andrew King</author><author>Alexander Hammers</author>
        <description><![CDATA[Accurate segmentation of both the pituitary gland and adenomas from magnetic resonance imaging (MRI) is essential for diagnosis and treatment of pituitary adenomas. This systematic review evaluates automatic segmentation methods for improving the accuracy and efficiency of MRI-based segmentation of pituitary adenomas and the gland itself. We analysed 34 studies that employed automatic and semi-automatic segmentation methods out of 353 reviewed studies. We extracted and synthesized data on segmentation techniques and performance metrics (such as Dice overlap scores). The majority of reviewed studies utilized deep learning approaches, with U-Net-based models being the most prevalent. Automatic methods yielded Dice scores of 0%–89% for pituitary gland and 4%–96% for adenoma segmentation. Semi-automatic methods reported 80%–92% for pituitary gland and 75%–88% for adenoma segmentation. Most studies did not report important metrics such as MR field strength, age and adenoma size (macro/micro/giant) or even adenoma type and human subject numbers. Automated segmentation techniques such as U-Net-based models show promise, especially for adenoma segmentation, but further improvements are needed to achieve consistently good performance in small structures like the normal pituitary gland. Future progress will require methodological innovation and larger, more diverse datasets to enhance clinical applicability.Systematic Review Registration:https://www.crd.york.ac.uk/PROSPERO/view/CRD42023407127, PROSPERO CRD42023407127.]]></description>
      </item><item>
        <guid isPermaLink="true">https://www.frontiersin.org/articles/10.3389/fradi.2026.1795775</guid>
        <link>https://www.frontiersin.org/articles/10.3389/fradi.2026.1795775</link>
        <title><![CDATA[Case Report: Embolization of a cognard type V cavernous sinus dural arteriovenous fistula via occluded Inferior petrosal sinus]]></title>
        <pubdate>2026-04-10T00:00:00Z</pubdate>
        <category>Case Report</category>
        <author>Zhaolong Zhang</author><author>Yanan Jiang</author><author>Qi Zhang</author><author>Liming Shao</author><author>Hongzhang Xu</author><author>Yixing Xie</author><author>Xiaolong Zhao</author><author>Chengjian Sun</author><author>Rui Xu</author>
        <description><![CDATA[BackgroundCavernous sinus dural arteriovenous fistula (CS-DAVF) with cortical venous drainage to the spinal perimedullary veins (Cognard type V) is a rare clinical entity. The treatment becomes particularly challenging when such a case is complicated by occlusion of the inferior petrosal sinus. We report a rare case of Cognard type V cavernous sinus dural arteriovenous fistula (CS-DAVF) embolized with Onyx via the occluded inferior petrosal sinus (IPS).Case descriptionA 62-year-old woman presented to our hospital with nausea, vomiting and progressive weakness in both lower limbs. Magnetic resonance imaging (MRI) revealed extensive edema of pons and medulla oblongata with flow voids. Cerebral angiography demenstrated a dural arteriovenous fistula (DAVF) in the left cavernous sinus, supplied by the meningohypophyseal trunk and branches of the middle meningeal artery. The fistula drained into the intercavernous sinus and right cavernous sinus, ultimately draining into the spinal perimedullary veins. Transarterial approach was tried first. Onyx-18 was injected into the branches of middle meningeal artery, partially casting the fistula. However, the DAVF was not completely occluded due to residual supply from the meningohypophyseal trunk. The second operation was performed using transvenous approach. The microcatheter was successfully navigated to the fistulous point via the occluded right IPS and the intercavernous sinus. Left cavernous sinus (CS) and intercavernous sinus were casted using Onyx-18 and the DAVF was completely eliminated.ConclusionWe report a Cognard type V CS-DAVF that was embolized with Onyx through an occluded IPS, demonstrating that this technique is a feasible and effective treatment option.]]></description>
      </item><item>
        <guid isPermaLink="true">https://www.frontiersin.org/articles/10.3389/fradi.2026.1719448</guid>
        <link>https://www.frontiersin.org/articles/10.3389/fradi.2026.1719448</link>
        <title><![CDATA[Prophylactic transarterial embolization after endoscopic hemostasis in patients with non-variceal upper gastrointestinal bleeding - is it time to act?]]></title>
        <pubdate>2026-04-02T00:00:00Z</pubdate>
        <category>Mini Review</category>
        <author>Francesco Tiralongo</author><author>Roberto Minici</author><author>Makoto Taninokuchi Tomassoni</author><author>Corrado Ini'</author><author>Davide Giuseppe Castiglione</author><author>Francesco Vacirca</author><author>Cristina Mosconi</author><author>Stefania Tamburrini</author><author>Giuseppe Messina</author><author>Emanuele David</author><author>Pietro Valerio Foti</author><author>Stefano Palmucci</author><author>Antonio Basile</author>
        <description><![CDATA[BackgroundNon-variceal upper gastrointestinal bleeding (NVUGIB) continues to present a significant clinical burden due to rebleeding after apparently successful endoscopic hemostasis, particularly in ulcers overlying large-caliber arterial territories. Prophylactic transarterial embolization (pTAE) has been proposed as a strategy to prevent rebleeding in high-risk patients. This mini-review evaluates the evidence for pTAE after successful endoscopic control in NVUGIB, focusing on patient selection, technical approaches, outcomes, and complications.MethodsA literature search of PubMed and Scopus (January 2010–September 2025) was conducted, yielding 10 studies (two randomized trials, three prospective, and five retrospective) evaluating pTAE. Only studies addressing prophylactic, not empiric, embolization were included.ResultsEvidence suggests that pTAE is technically feasible and generally safe when guided by ulcer location, size (≥15–20 mm), Rockall score (≥5), and arterial territory (GDA or LGA). While randomized trials did not show overall superiority in intention-to-treat analyses, per-protocol data and observational studies suggest reduced rebleeding and a lower need for surgical rescue in well-selected patients. Complications are infrequent when standardized techniques and early timing (≤24 h) are applied.ConclusionRoutine pTAE is not supported by current guidelines or RCT-level evidence. However, in anatomically and clinically high-risk ulcers, pTAE may offer meaningful benefits. Further multicenter randomized trials with uniform protocols are warranted to clarify its role and optimize patient selection.]]></description>
      </item><item>
        <guid isPermaLink="true">https://www.frontiersin.org/articles/10.3389/fradi.2026.1782068</guid>
        <link>https://www.frontiersin.org/articles/10.3389/fradi.2026.1782068</link>
        <title><![CDATA[Persistent vessel wall enhancement and progression of cerebral microbleeds in CADASIL: a case report]]></title>
        <pubdate>2026-04-02T00:00:00Z</pubdate>
        <category>Case Report</category>
        <author>Wenjuan Xu</author><author>Chao Zhang</author><author>Xiaomin Liu</author><author>Xiaoyu Zhang</author>
        <description><![CDATA[Cerebral autosomal dominant arteriopathy with subcortical infarcts and leukoencephalopathy (CADASIL) is the most prevalent hereditary cerebral small vessel disease (cSVD), primarily caused by pathogenic variants in the NOTCH3 gene. Neuroimaging features, including white matter hyperintensities (WMH), lacunar infarcts, and cerebral microbleeds (CMBs), typically progress with disease advancement. However, the pathophysiological mechanisms underlying this neuroimaging progression remain poorly understood. Here we present a patient with a NOTCH3 mutation who exhibited persistent vessel wall enhancement on serial high-resolution vessel wall MRI (VWMRI), alongside progression of CMBs. This finding supports a critical role of blood-brain barrier (BBB) disruption in CADASIL pathophysiology. In conclusion, persistent intracranial vessel wall enhancement may be observed in patients with CADASIL. Further studies are needed to investigate the relationship between this imaging biomarker and clinical as well as neuroimaging outcomes.]]></description>
      </item><item>
        <guid isPermaLink="true">https://www.frontiersin.org/articles/10.3389/fradi.2026.1744940</guid>
        <link>https://www.frontiersin.org/articles/10.3389/fradi.2026.1744940</link>
        <title><![CDATA[A retrospective study of 96 cases comparing x-ray radiography and MRI for diagnosing paediatric subacute osteomyelitis]]></title>
        <pubdate>2026-04-01T00:00:00Z</pubdate>
        <category>Original Research</category>
        <author>Elio Paris</author><author>Ahmer A. Khan</author><author>Giacomo De Marco</author><author>Anne Tabard-Fougère</author><author>Oscar Vazquez</author><author>Christina Steiger</author><author>Romain Dayer</author><author>Dimitri Ceroni</author>
        <description><![CDATA[BackgroundSubacute hematogenous osteomyelitis (SAHOM) presents a diagnostic challenge, requiring robust validation of imaging accuracy.PurposeTo determine the superior diagnostic performance of MRI vs. radiography (x-ray) in detecting and classifying SAHOM.MethodsThis retrospective study included 96 proven SAHOM cases (2000–2025). Demographic data, involved bones, and microbiological results were collected. Two independent readers assessed x-Ray and MRI for detection of SAHOM, and classified lesions using the modified Roberts classification. Inter-reader disagreements were resolved by consensus. Sensitivity of x-Ray was evaluated against MRI as the reference standard.Resultsx-ray radiographs and MRI from 96 proven cases of SAHOM involving 49 males and 47 females (mean age 47.1 ± 47.6 months) were evaluated. MRI was markedly more sensitive, with significantly more correct imaging findings than radiography for detecting the features of SAHOM (100% vs. 47.9%). Moreover, 21.3% of the SAHOM lesions on x-ray radiography were misclassified. Radiography's limitations were most pronounced for lesions of the spine, tarsal/carpal bones, pelvis, and epiphysis, as well as for infections caused by Kingella kingae (K. kingae).ConclusionsMRI is a more effective method than x-ray radiography for diagnosing SAHOM; it reveals lesions with higher definition and enables their more precise classification. This is especially true of lesions involving the spine, pelvis, tarsal or carpal bones, and the epiphysis, or when SAHOM is caused by K. kingae. MRI also provides much better imaging of the involvement of growth cartilage and damage to articular cartilage.]]></description>
      </item><item>
        <guid isPermaLink="true">https://www.frontiersin.org/articles/10.3389/fradi.2026.1800804</guid>
        <link>https://www.frontiersin.org/articles/10.3389/fradi.2026.1800804</link>
        <title><![CDATA[Subclinical pulmonary congestion in ST-segment elevation myocardial infarction assessed by computed tomography]]></title>
        <pubdate>2026-04-01T00:00:00Z</pubdate>
        <category>Original Research</category>
        <author>Yufan Gao</author><author>Keyi Cui</author><author>Shuo Liang</author><author>Minghui Hua</author><author>Hong Zhang</author><author>Zhigang Guo</author>
        <description><![CDATA[ObjectivesTo evaluate the prognostic value of subclinical pulmonary congestion, assessed as extravascular lung water (EVLW) by computed tomography (CT), in ST-segment elevation myocardial infarction (STEMI) patients with Killip class 1.MethodsThis retrospective study included Killip class 1 STEMI patients who underwent CT prior to primary percutaneous coronary intervention (PCI). EVLW was derived from mean lung density. The Global Registry of Acute Coronary Events (GRACE) score was calculated. The endpoint was in-hospital major adverse cardiovascular events (MACE), defined as all-cause mortality, acute heart failure, cardiogenic shock, resuscitated cardiac arrest, or stroke. The predictive improvement of adding EVLW to the GRACE score was assessed using receiver operating characteristic (ROC) analysis, net reclassification improvement (NRI), and integrated discrimination improvement (IDI).ResultsAmong 249 patients (mean age 59 ± 11 years; 195 men), 28 experienced MACE. Patients with MACE had a higher GRACE score (144.96 ± 22.95 vs. 133.63 ± 19.84, P = 0.006) and EVLW (24.24% vs. 21.36%, P = 0.001). Adding EVLW to the GRACE score significantly increased the area under the ROC curve (AUC) (0.754 vs. 0.656, P = 0.045), NRI (0.491, P = 0.013), and IDI (0.060, P = 0.019).ConclusionsIn Killip class 1 STEMI patients, CT-identified subclinical pulmonary congestion is associated with an increased risk of in-hospital MACE.]]></description>
      </item><item>
        <guid isPermaLink="true">https://www.frontiersin.org/articles/10.3389/fradi.2026.1740915</guid>
        <link>https://www.frontiersin.org/articles/10.3389/fradi.2026.1740915</link>
        <title><![CDATA[Artificial intelligence in the diagnosis of thyroid diseases: applications and challenges]]></title>
        <pubdate>2026-04-01T00:00:00Z</pubdate>
        <category>Review</category>
        <author>Huayi Zhao</author><author>Tingyu Xue</author><author>Yan Zhang</author><author>Yanchao Lu</author><author>Danning Wang</author><author>Pei Yin</author><author>Lei Jiang</author><author>Dan Liu</author><author>Yong Wang</author>
        <description><![CDATA[Thyroid diseases, a prevalent class of endocrine system disorders, require diagnostic accuracy, which is essential for effective patient treatment and management. In recent years, artificial intelligence (AI) technology has made significant advancements in the medical field, providing new opportunities for the early diagnosis and precise treatment of thyroid diseases. This review discusses the latest applications of AI in the diagnosis of thyroid diseases, with a particular focus on the use of machine learning and deep learning (DL) algorithms in image classification, segmentation, and object detection within thyroid ultrasound, computed tomography, magnetic resonance imaging, and single photon emission computed tomography. Through the integration of cross-modal studies, this article reveals the application of AI across various imaging modalities, highlighting its potential value in feature extraction and risk stratification. Furthermore, we conduct an in-depth analysis of key challenges faced by AI applications, such as data heterogeneity (the decline in model performance due to data differences across institutions and equipment) and insufficient interpretability (DL models often function as “black boxes,” making it difficult to provide transparent decision-making rationale, which limits clinical trust and adoption). In summary, AI technology demonstrates notable advantages and developmental potential in the automated diagnosis of thyroid diseases; however, its clinical translation still requires addressing the aforementioned challenges. The resultant analysis demonstrates that AI holds promise in improving the diagnosis and treatment of thyroid diseases, offering new pathways for personalized medicine and better patient outcomes. Specifically, AI-driven tools can reduce diagnostic variability and errors in thyroid nodule assessment, enhance treatment precision with risk-stratified recommendations, and support more consistent, individualized clinical decisions.]]></description>
      </item>
      </channel>
    </rss>