Frontiers in radiology最新文献

{"title":"First steps in clinical implementation of a lung nodule classification method by artificial intelligence.","authors":"Abderrazzak Ajertil, Zineb Farahat, Abla Bouallou, Elisa Le Tertre, Magnanga-Boukandza Joey Ben Elbher, Abdeljalil El Quessar, Nejat Kabbaj, Mohamed Cherkaoui Malki, Latifa Chat, Lahcen Hammani, Nabil Ngote","doi":"10.3389/fradi.2026.1714951","DOIUrl":"10.3389/fradi.2026.1714951","url":null,"abstract":"<p><p>Lung cancer remains one of the leading causes of cancer-related mortality worldwide, and early detection is essential to improve patient survival. Computed tomography (CT) is currently the reference imaging modality for lung cancer screening. Recent advances in Artificial Intelligence (AI), particularly Deep Learning (DL), have significantly improved automated medical image analysis and diagnostic accuracy. In this study, we propose a Convolutional Neural Network (CNN)-based approach for pulmonary nodule classification. The model was trained using a merged dataset composed of public CT image databases (IQ-OTH/NCCD and SPIE-AAPM) and real-world CT scans collected at Cheikh Zaid Hospital, Morocco. The experimental dataset included 1,103 malignant images, 508 benign images, and 427 normal images. Contrast Limited Adaptive Histogram Equalization (CLAHE) was applied to enhance image contrast prior to training. The proposed CNN achieved an accuracy of 99.84%, precision of 99.97%, sensitivity of 99.84%, and specificity of 99.82% for a three-class classification task. These results demonstrate the robustness and clinical potential of the model in assisting radiologists with the classification of indeterminate pulmonary nodules. This work represents an initial translational step toward real-world clinical implementation of AI-based lung nodule classification at Cheikh Zaid Hospital and highlights the feasibility of integrating AI tools into routine radiology workflows.</p>","PeriodicalId":73101,"journal":{"name":"Frontiers in radiology","volume":"6 ","pages":"1714951"},"PeriodicalIF":4.0,"publicationDate":"2026-07-21","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC13435035/pdf/","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"148674901","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":"From plug-and-play to institution-calibrated radiology AI: a practical framework for operationalizing local validation, monitoring and governance.","authors":"Muhammad Awais, Abdul Rehman","doi":"10.3389/fradi.2026.1902217","DOIUrl":"10.3389/fradi.2026.1902217","url":null,"abstract":"<p><p>Radiology artificial intelligence (AI) is increasingly developed on large external datasets and deployed across institutions, but real-world model performance may vary substantially after implementation. Imaging AI interacts with a local ecosystem shaped by scanner hardware, acquisition protocols, reconstruction methods, technologist practices, disease prevalence, patient demographics, reporting conventions, and clinical workflow. These factors can produce domain shift, degrade calibration, alter false-positive and false-negative patterns, and affect clinical utility. In this article, we argue that radiology AI should move beyond a \"plug-and-play\" deployment paradigm toward institution-calibrated AI stewardship. We propose an institution-specific radiology AI performance profile, conceptually analogous to a local antibiogram, to summarize how AI tools perform within a specific clinical environment. Unlike prior MLOps and radiology AI governance frameworks, the proposed profile translates lifecycle management into a radiology-specific, locally maintainable artifact that captures technical context, clinical context, performance metrics, reference standards, equity checks, workflow effects, and governance triggers. The framework applies not only to diagnostic decision support, but also to research cohort generation and radiology education, where local reporting language, imaging protocols, and case mix may strongly influence model reliability. We emphasize that institution-specific calibration does not require every hospital to develop AI models <i>de novo</i>. Rather, externally developed, vendor-based, and foundation-model approaches should be paired with local validation, cautious threshold adjustment, calibration when applicable, surveillance for drift and protocol changes, and multidisciplinary governance. Responsible radiology AI deployment should therefore ask not only whether a model is accurate, but whether it remains accurate, relevant, equitable, and useful locally over time.</p>","PeriodicalId":73101,"journal":{"name":"Frontiers in radiology","volume":"6 ","pages":"1902217"},"PeriodicalIF":4.0,"publicationDate":"2026-07-16","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC13422437/pdf/","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"148632742","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":"Diagnostic value of quantitative MR spectroscopy metabolite ratios for Canavan disease.","authors":"Majid Reza Tahsini, Marzena Wylezinska-Arridge, Mohammad Zare, Sasan Saket, Mohammad Aidin Farahvash, Mahdis Morovvati, Elham Rahimian, Sotirios Bisdas","doi":"10.3389/fradi.2026.1826170","DOIUrl":"10.3389/fradi.2026.1826170","url":null,"abstract":"<p><strong>Purpose: </strong>Canavan disease (CD) is a leukodystrophy marked by deficiency in the aspartoacylase enzyme and accumulation of N-acetyl aspartate (NAA) in the white matter. MRI and MRS are crucial diagnostic tools for CD. MRS shows a distinctive, significant elevation of the NAA peak, which is typically not quantified and thus not objectively assessed as a diagnostic criterion. This study aimed to establish quantitative cutoff values for the elevated NAA peak on proton MR spectroscopy (¹H-MRS) in patients with CD and hence, incorporate MRS into the diagnostic algorithm.</p><p><strong>Methods: </strong>A cross-sectional retrospective diagnostic study involving 18 confirmed CD cases and 10 healthy individuals was conducted to assess metabolic profiles and determine a diagnostic cutoff for NAA in CD. 1H MRS (TE: 135 ms) was utilized, and data were post-processed using specialized software to quantify the ratios of the three primary metabolites: NAA, Cho, and Cr.</p><p><strong>Results: </strong>The analysis of the CD cases demonstrated substantial elevation in NAA levels compared to the standard control group with the NAA/Cr and NAA/Cho ratios significantly increased in CD patients compared to controls (<i>p</i> < 0.001 for both). Diagnostic cut-off values of 2.75 for NAA/Cr and 7.18 for NAA/Cho yielded an optimism-corrected AUC = 0.96 (95% CI 0.95-1.00) after bootstrap validation.</p><p><strong>Conclusion: </strong>This study represents one of the larger single-center cohorts evaluating quantitative MRS metabolite ratios in Canavan disease and proposes preliminary diagnostic thresholds for improved diagnostic objectivity.</p>","PeriodicalId":73101,"journal":{"name":"Frontiers in radiology","volume":"6 ","pages":"1826170"},"PeriodicalIF":4.0,"publicationDate":"2026-07-15","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC13416435/pdf/","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"148622960","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":"Case Report: Alternative biopsy targets plus CEUS-guided omental puncture: enabling CLDN18.2 detection and precision therapy enrollment in pancreatic cancer.","authors":"Hang Yin, Xiaomeng Zhang, Bing Song, Hui Li, Kun Liu, Chengzheng Zhang","doi":"10.3389/fradi.2026.1800543","DOIUrl":"10.3389/fradi.2026.1800543","url":null,"abstract":"<p><strong>Introduction: </strong>Some primary pancreatic cancer lesions are located deep in the abdomen and adjacent to major blood vessels, which leads to procedural risk and low success rates of percutaneous biopsy. This situation creates substantial obstacles for pathological diagnosis and biomarker detection.</p><p><strong>Methods: </strong>We present a 54-year-old female patient diagnosed with malignant pancreatic body carcinoma. The primary tumor was in close proximity to the splenic artery and accompanied by omental metastatic nodules. After multidisciplinary team discussion, we selected the safely accessible omental metastatic lesions as alternative biopsy targets and performed core needle biopsy under contrast-enhanced ultrasound (CEUS) guidance for tissue sampling.</p><p><strong>Results: </strong>Pathological examination confirmed adenocarcinoma, and the immunophenotype was consistent with a pancreatobiliary origin. Immunohistochemical staining revealed high CLDN18.2 expression. The patient was subsequently enrolled in a clinical trial for CLDN18.2-directed combination therapy. Serial imaging follow-up demonstrated measurable shrinkage of both the primary tumor and metastatic lesions.</p><p><strong>Discussion: </strong>When imaging and clinical evaluations strongly indicate a solitary pancreatic malignancy and safely accessible metastatic lesions are present, the strategy of CEUS-guided alternative target biopsy serves as a viable approach to acquire tissue specimens and conduct biomarker testing. Nonetheless, its diagnostic concordance and clinical generalizability remain to be validated in large-sample studies.</p>","PeriodicalId":73101,"journal":{"name":"Frontiers in radiology","volume":"6 ","pages":"1800543"},"PeriodicalIF":4.0,"publicationDate":"2026-07-15","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC13416683/pdf/","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"148622876","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":"Multimodal assessment of fetal head circumference and biparietal diameter: systematic differences between MRI and US measurements and development of an US-equivalent prediction regression model.","authors":"Kai Ye, Kexin Wang, Pengsheng Wu, Xiaoying Wang","doi":"10.3389/fradi.2026.1828351","DOIUrl":"10.3389/fradi.2026.1828351","url":null,"abstract":"<p><strong>Objective: </strong>To investigate the differences between fetal magnetic resonance imaging (MRI) and ultrasonography (US) in measuring head circumference (HC) and biparietal diameter (BPD), and to develop and validate a regression model for predicting US-equivalent values based on MRI measurements.</p><p><strong>Methods: </strong>Imaging data of 124 mid-to-late trimester fetuses were retrospectively included. The differences in HC and BPD measurements between the two modalities were quantitatively analyzed. A regression model was constructed using MRI measurements, gestational age (GA), and the time interval between examinations as independent variables to predict the US-equivalent values (USpred).</p><p><strong>Results: </strong>Compared with US, MRI demonstrated significant systematic overestimation for both HC (mean difference: 14.0 mm, <i>p</i> < 0.001) and BPD (median difference: 3.18 mm, <i>p</i> < 0.001). The established USpred model showed high goodness-of-fit (adjusted <i>R</i> ² > 0.95). When classifying based on US standards, the USpred values exhibited significantly improved classification efficacy compared to direct MRI measurements: overall accuracy (HC: 0.782 vs. 0.629; BPD: 0.863 vs. 0.782) and Matthews correlation coefficient (HC: 0.512 vs. 0.322; BPD: 0.639 vs. 0.528).</p><p><strong>Conclusion: </strong>Fetal MRI measurements of HC and BPD show a systematic overestimation compared to US. The USpred model can effectively correct for this systematic bias in MRI measurements and the influence of examination time intervals, significantly improving agreement with commonly used clinical US standards. It provides a reliable tool for the precise assessment of prenatal MRI.</p>","PeriodicalId":73101,"journal":{"name":"Frontiers in radiology","volume":"6 ","pages":"1828351"},"PeriodicalIF":4.0,"publicationDate":"2026-07-14","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC13407774/pdf/","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"148610736","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":"Case Report: Uremic encephalopathy in acute kidney injury caused by abdominal sepsis.","authors":"Sergey Karasev, Rustam Talybov, Tatyana Trofimova, Irina Karaseva","doi":"10.3389/fradi.2026.1875949","DOIUrl":"10.3389/fradi.2026.1875949","url":null,"abstract":"<p><p>Uremic encephalopathy encompasses a wide range of central nervous system disorders resulting from acute or chronic kidney damage. The pathogenesis of this syndrome is multifactorial and includes the accumulation of uremic toxins, hormonal imbalance, electrolyte and acid-base imbalance, blood-brain barrier dysfunction, and systemic inflammation. There are no specific clinical or laboratory markers for uremic encephalopathy, and the diagnosis is often made retrospectively, based on the dynamics of neurological status. The diagnostic process is further complicated by the presence of comorbid and overlapping diseases. Consequently, renal replacement therapy is often initiated at an suboptimal time. Neuroimaging plays a key role in confirming the diagnosis, providing high sensitivity in assessing the nature and extent of brain tissue damage. The \"lenticular fork sign\" is recognized as the most significant MRI marker and can serve as an early and reliable indicator of uremic encephalopathy. The implementation of an interdisciplinary strategy, including a comprehensive laboratory assessment and timely neuroimaging, allows not only to confirm the pathogenetic mechanisms underlying central nervous system damage, but also to promptly adjust therapeutic treatment. This publication presents a clinical case of uremic encephalopathy resulting from acute kidney injury associated with abdominal sepsis in a 50-year-old patient.</p>","PeriodicalId":73101,"journal":{"name":"Frontiers in radiology","volume":"6 ","pages":"1875949"},"PeriodicalIF":4.0,"publicationDate":"2026-07-10","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC13395866/pdf/","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"148586270","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":"Development of a deep learning model for intussusception using point-of-care ultrasound.","authors":"Anand Thyagachandran, Brian Lefchak, Hema A Murthy, Kelly R Bergmann, Manu Madhok","doi":"10.3389/fradi.2026.1812587","DOIUrl":"10.3389/fradi.2026.1812587","url":null,"abstract":"<p><strong>Objective: </strong>Intussusception is a pediatric emergency with delays in diagnosis. We sought to develop a novel deep learning model for the detection of target sign on point-of-care ultrasound (POCUS) images.</p><p><strong>Materials and methods: </strong>POCUS image/video clip media files obtained during emergency department (ED) visits were included for study. Images underwent preprocessing to enhance and resize the region of interest (ROI), ImageNet high dimensional feature extraction and further training using various machine learning models, including classical machine learning, ensemble learning, transfer learning, and fine-tuning. Outputs were analyzed on a patient case, individual image frame and relative threshold bases.</p><p><strong>Results: </strong>A POCUS image database of originally 785 media files from 49 patients, 8 of whom were positive for intussusception, was converted to 1,582 intussusception and 1,965 normal images for training. The output results show that the fine-tuning models performed better than the classical machine, ensemble and transfer learning models across three different analyses, and that the threshold-based approach to intussusception cases resulted in the greatest predictive performance.</p><p><strong>Discussion: </strong>Intussusception is a potential candidate for development of deep learning tools due to limited capacity for pediatric focused imaging and accurate diagnosis. Prior studies are limited and have used either large private datasets or formal radiology studies. Modeling involved converting dynamic video files into several static images. Fine-tuning models were best adapted to the screening nature of POCUS images.</p><p><strong>Conclusion: </strong>Our work demonstrated the feasibility of developing a deep learning model for the detection of intussusception using a smaller dataset of POCUS images.</p>","PeriodicalId":73101,"journal":{"name":"Frontiers in radiology","volume":"6 ","pages":"1812587"},"PeriodicalIF":4.0,"publicationDate":"2026-07-06","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC13381468/pdf/","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"148610772","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":"Potential of Image2Image translation in reducing AI bias attributed to differences in CT reconstruction methods: proof-of-concept study on a paired dataset.","authors":"Feifei Li, Liliana Lourenco Caldeira, Astha Jaiswal, Nils Große Hokamp, Oya Beyan, Ekaterina Kutafina","doi":"10.3389/fradi.2026.1875867","DOIUrl":"10.3389/fradi.2026.1875867","url":null,"abstract":"&lt;p&gt;&lt;strong&gt;Background: &lt;/strong&gt;The choice of image reconstruction approach, such as Iterative Model Reconstruction (IMR) and iDose &lt;math&gt;&lt;msup&gt;&lt;mi&gt;&lt;/mi&gt; &lt;mn&gt;4&lt;/mn&gt;&lt;/msup&gt; &lt;/math&gt; , is critical in computed tomography (CT) as distribution shifts between them can introduce systematic bias in clinical AI models, posing a barrier to efficient, scalable AI deployment in radiology. While Image-to-Image (I2I) translation offers a path toward harmonization, efficient and translation-feasible AI in radiology demands that such approaches be validated not only on image quality metrics but also on downstream clinical task performance.&lt;/p&gt;&lt;p&gt;&lt;strong&gt;Methods: &lt;/strong&gt;We compared three harmonization strategies: CycleGAN, Denoising Diffusion Models, and a conventional Gaussian filter, using a paired CT dataset. Performance was evaluated using task-agnostic metrics, including Structural Similarity Index Measure (SSIM), Peak Signal-to-Noise Ratio (PSNR), Frechet Inception Distance (FID), and Sliced Wasserstein Distance (SWD). Crucially, we assessed clinical utility by measuring the downstream impact on multi-organ segmentation using Mean Relative Volume Error (RVE), focusing on consistency across reconstructed modes.&lt;/p&gt;&lt;p&gt;&lt;strong&gt;Results: &lt;/strong&gt;Our findings reveal a significant discrepancy between visual fidelity and clinical utility. While Diffusion and CycleGAN models achieve competitive feature-level similarity, they introduce catastrophic systematic biases in downstream segmentation, particularly in adipose tissues (Body Composition Analysis, BCA), where volume errors can reach &lt;math&gt;&lt;mo&gt;-&lt;/mo&gt;&lt;/math&gt; 100%, indicating a failure to preserve critical anatomical fat structures. In contrast, the computationally efficient Gaussian filter served as the most robust baseline, maintaining minimal RVE and the highest consistency between IMR and iDose &lt;math&gt;&lt;msup&gt;&lt;mi&gt;&lt;/mi&gt; &lt;mn&gt;4&lt;/mn&gt;&lt;/msup&gt; &lt;/math&gt; modes.&lt;/p&gt;&lt;p&gt;&lt;strong&gt;Conclusions: &lt;/strong&gt;The results suggest that in the context of efficient AI for radiology imaging, where efficiency encompasses not only computational cost but also reliability of clinical outputs, task-agnostic image quality metrics are insufficient proxies for downstream clinical performance. High-parameter generative models may introduce structural losses that remain invisible to such metrics yet critically affect clinical utility. The underperformance of generative approaches observed here is likely attributable to 3D slice inconsistencies during volumetric reconstruction and suboptimal model adaptation, rather than to an inherent limitation of I2I methods, which retain significant potential for mitigating reconstruction-related bias in radiology AI. Future work developing harmonization pipelines must therefore prioritize task-specific, clinically grounded validation as the primary benchmark-and carefully consider whether computationally simpler approaches, such as conventional filtering, may already provide comparably stable harmonization for s","PeriodicalId":73101,"journal":{"name":"Frontiers in radiology","volume":"6 ","pages":"1875867"},"PeriodicalIF":4.0,"publicationDate":"2026-07-03","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC13375877/pdf/","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"148581502","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":"Early identification of PRISm from chest CT: a multimodal for three-class stratification of normal, PRISm, and COPD.","authors":"Yan Mou, Xueqing Bian, Na Tian, Zheng Xia, Jinbo Shi, Ming Li, Zhimin Shao, Lin Qi, Haiyan Ge, Miao Huang","doi":"10.3389/fradi.2026.1829853","DOIUrl":"10.3389/fradi.2026.1829853","url":null,"abstract":"<p><strong>Objective: </strong>Preserved ratio impaired spirometry (PRISm), a precursor to chronic obstructive pulmonary disease (COPD), is difficult to diagnose, primarily due to its subtle and inconspicuous changes in lung structure. This study aims to develop an attention-driven multi-instance learning framework with ResNet34 (MIL-R34) that integrates CT imaging with clinical data to automatically classify normal, PRISm, and COPD, thereby addressing current limitations in early PRISm detection.</p><p><strong>Materials and methods: </strong>A total of 1,063 participants from two centers who underwent thin-slice chest CT (<1 mm) and pulmonary function tests within 2 weeks were retrospectively analyzed. After screening, 966 participants were included (Center A: 776; Center B: 190 for external testing). Participants were divided into normal, PRISm, and COPD groups. For each subject, 20 CT slices were selected and processed using a denoising strategy. Four deep learning models (ResNet34, CNN, Swin Transformer, EfficientNet) were compared. Clinical data were encoded and combined with imaging features to construct a multimodal model. Performance was evaluated using AUC, accuracy, sensitivity, and specificity.</p><p><strong>Results: </strong>Among CT-only models, ResNet34 achieved the best performance (AUC 0.873, accuracy 0.699). The clinical-only model showed limited performance (AUC 0.656, accuracy 0.423), whereas the multimodal model further improved performance (AUC 0.879, accuracy 0.724), outperforming both unimodal approaches. External validation yielded an AUC of 0.808 and an accuracy of 0.642, demonstrating reasonable generalizability despite performance decline across centers.</p><p><strong>Conclusion: </strong>This study demonstrates that chest CT imaging, when combined with clinical demographic data, can reliably identify PRISm as a transitional state between normal lung function and COPD. Furthermore, the proposed multimodal provides interpretable imaging biomarkers that enhance COPD risk stratification and address current diagnostic limitations in the early detection of PRISm.</p>","PeriodicalId":73101,"journal":{"name":"Frontiers in radiology","volume":"6 ","pages":"1829853"},"PeriodicalIF":4.0,"publicationDate":"2026-07-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC13368798/pdf/","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"148457643","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":"Harmonization of slice thickness through resampling improves comparability of MRI-derived neonatal brain volumes.","authors":"Anouk S Verschuur, Martijn F Boomsma, Alexander Leemans, Ingrid M Nijholt, Leonora Hendson, Selma Low, Amy Metcalfe, Donna M Slater, Chantal M W Tax, Gerda Meijler, Lara M Leijser","doi":"10.3389/fradi.2026.1868602","DOIUrl":"10.3389/fradi.2026.1868602","url":null,"abstract":"<p><strong>Purpose: </strong>To assess reliability and agreement of brain volume measurements from neonatal T2-weighted magnetic resonance imaging (MRI) scans acquired with different slice thickness settings (i.e., 2 mm, 3 mm and 3 mm resampled).</p><p><strong>Methods: </strong>We used term-equivalent (38-45 weeks postmenstrual age) T2-weighted MRI scans with slice thicknesses of 2 mm and 3 mm from a subset of 40 preterm and term infants (27-40 weeks' gestation) from two existing neonatal imaging cohorts. Resampling of the 2 mm scans to a 3 mm grid provided a third set of scans, referred to as \"3 mm resampled\". Automated brain segmentation provided intracranial volume, total brain volume, and volumes for eight brain structures. Reliability and agreement of volume measurements were assessed using intraclass correlation coefficients (ICC) and Bland-Altman plots, respectively.</p><p><strong>Results: </strong>For all volumes, inter-scan reliability was higher for 3 mm vs. 3 mm resampled scans (mean observer ICC >0.92) compared to 2 mm vs. 3 mm scans (mean observer ICC >0.87). Bland-Altman plots revealed smaller mean volume differences between 3 mm and 3 mm resampled scans than between 2 mm and 3 mm scans. However, a proportional bias was found for deep gray matter volume for 3 mm vs. 3 mm resampled scans.</p><p><strong>Conclusion: </strong>Slice thickness influences brain volume measurements from neonatal MRI. Harmonization of imaging data acquired with different settings through image resampling improves reliability and agreement of brain volume measurements.</p>","PeriodicalId":73101,"journal":{"name":"Frontiers in radiology","volume":"6 ","pages":"1868602"},"PeriodicalIF":4.0,"publicationDate":"2026-06-29","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC13357915/pdf/","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"148439246","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}