European Journal of Radiology最新文献_第5页

Comparison of perfusion imaging parameters with and without symptom progression in mild stroke patients due to large/medium vessel occlusion 轻度脑卒中大/中血管闭塞伴与无症状进展患者灌注成像参数的比较

IF 3.3 3区医学 Q1 RADIOLOGY, NUCLEAR MEDICINE & MEDICAL IMAGING

European Journal of Radiology

Pub Date : 2025-12-15 DOI: 10.1016/j.ejrad.2025.112617

Koichiro Shindo , Kazutaka Uchida , Fumihiro Sakakibara , Manabu Shirakawa , Yoji Kuramoto , Shinichi Yoshimura

Background and purpose

The association between stroke progression in mild stroke due to large or medium vessel occlusion and perfusion imaging parameters remain uncertain. This study aimed to identify perfusion imaging parameters that predict progressive and/or disabling stroke after hospitalization.

Materials and methods

This retrospective study analyzed 527 consecutive patients with acute large or medium vessel occlusion (April 2019–March 2024) using Rapid processing of Perfusion and Diffusion (RAPID) imaging. Patients with National Institutes of Health Stroke Scale (NIHSS) ≤ 5 and cervical/internal carotid artery or middle cerebral artery occlusion were classified into the progressive/disabling stroke (PD) group (worsening NIHSS ≥ 4 or disabling deficit despite best medical treatment without EVT) or the non-PD group. Patient characteristics and RAPID imaging parameters (relative cerebral blood flow < 30 %; time-to-maximum [Tmax] > 4, >6, >8, and > 10 s; and mismatch volume) were compared. Modified Rankin scale (mRS) 0–2 and 0–1 at 90 days were defined as functional outcomes. Statistical analyses included Student’s t-test, Wilcoxon rank-sum test, and Youden index.

Results

Among 37 patients (mean age, 70.6 ± 15.1 years; 64.9 % male), Tmax > 6 s (73 mL vs. 17 mL, p = 0.007) and mismatch volume (73 mL vs. 16 mL, p = 0.005) were significantly higher in the PD group (n = 9) than in the non-PD group (n = 28). Receiver operating characteristic analysis identified 63 mL as the optimal cutoff value for Tmax > 6 s (sensitivity: 0.78, specificity: 0.82, area under the curve [AUC]: 0.80) and mismatch volume (sensitivity: 0.78, specificity: 0.82, AUC: 0.82). The PD group had significantly worse functional outcomes at 90 days than the non-PD group (mRS 0–2: 3/8 [37.5 %] vs. 23/26 [88.5 %], p = 0.009; mRS 0–1: 2/8 [25.0 %] vs. 19/26 [73.1 %], p = 0.03).

Conclusions

The prediction of progressive stroke and/or disabling stroke based on perfusion imaging parameters upon admission would allow triage of patients requiring endovascular treatment.

背景和目的：由于大血管或中等血管闭塞引起的轻度卒中进展与灌注成像参数之间的关系尚不确定。本研究旨在确定预测住院后进行性和/或致残性卒中的灌注成像参数。材料与方法：本研究采用快速灌注扩散成像（Rapid processing of Perfusion and Diffusion, Rapid）对527例急性大、中血管闭塞患者（2019年4月- 2024年3月）进行回顾性分析。将美国国立卫生研究院卒中量表(NIHSS)≤5分及颈/颈内动脉或大脑中动脉闭塞的患者分为进行性/致残性卒中（PD）组（NIHSS≥4分恶化或虽经最佳治疗但无EVT仍存在致残性缺陷）或非PD组。比较患者特征和快速成像参数（相对脑血流量4、脑血流量6、脑血流量8和脑血流量10 s，以及错配量）。修正Rankin量表（mRS） 0-2和0-1在90天被定义为功能结局。统计分析包括Student’st检验、Wilcoxon秩和检验、Youden指数。结果：37例患者（平均年龄70.6±15.1岁，男性64.9%）中，PD组（n = 9） Tmax bbb6.0 s （73 mL vs. 17 mL, p = 0.007）和错配体积（73 mL vs. 16 mL, p = 0.005）显著高于非PD组（n = 28）。经受试者工作特征分析，确定63 mL为Tmax bbb6 s（灵敏度：0.78，特异性：0.82，曲线下面积[AUC]: 0.80）和错配体积（灵敏度：0.78，特异性：0.82,AUC: 0.82）的最佳截止值。PD组在90天的功能结局明显差于非PD组（mRS 0- 1:3 /8[37.5%]比23/26 [88.5%],p = 0.009; mRS 0- 1:2 /8[25.0%]比19/26 [73.1%],p = 0.03）。结论：根据入院时的灌注成像参数预测进行性卒中和/或致残性卒中，将有助于对需要血管内治疗的患者进行分诊。

{"title":"Comparison of perfusion imaging parameters with and without symptom progression in mild stroke patients due to large/medium vessel occlusion","authors":"Koichiro Shindo , Kazutaka Uchida , Fumihiro Sakakibara , Manabu Shirakawa , Yoji Kuramoto , Shinichi Yoshimura","doi":"10.1016/j.ejrad.2025.112617","DOIUrl":"10.1016/j.ejrad.2025.112617","url":null,"abstract":"<div><h3>Background and purpose</h3><div>The association between stroke progression in mild stroke due to large or medium vessel occlusion and perfusion imaging parameters remain uncertain. This study aimed to identify perfusion imaging parameters that predict progressive and/or disabling stroke after hospitalization.</div></div><div><h3>Materials and methods</h3><div>This retrospective study analyzed 527 consecutive patients with acute large or medium vessel occlusion (April 2019–March 2024) using Rapid processing of Perfusion and Diffusion (RAPID) imaging. Patients with National Institutes of Health Stroke Scale (NIHSS) ≤ 5 and cervical/internal carotid artery or middle cerebral artery occlusion were classified into the progressive/disabling stroke (PD) group (worsening NIHSS ≥ 4 or disabling deficit despite best medical treatment without EVT) or the non-PD group. Patient characteristics and RAPID imaging parameters (relative cerebral blood flow < 30 %; time-to-maximum [Tmax] > 4, >6, >8, and > 10 s; and mismatch volume) were compared. Modified Rankin scale (mRS) 0–2 and 0–1 at 90 days were defined as functional outcomes. Statistical analyses included Student’s <em>t</em>-test, Wilcoxon rank-sum test, and Youden index.</div></div><div><h3>Results</h3><div>Among 37 patients (mean age, 70.6 ± 15.1 years; 64.9 % male), Tmax > 6 s (73 mL vs. 17 mL, p = 0.007) and mismatch volume (73 mL vs. 16 mL, p = 0.005) were significantly higher in the PD group (n = 9) than in the non-PD group (n = 28). Receiver operating characteristic analysis identified 63 mL as the optimal cutoff value for Tmax > 6 s (sensitivity: 0.78, specificity: 0.82, area under the curve [AUC]: 0.80) and mismatch volume (sensitivity: 0.78, specificity: 0.82, AUC: 0.82). The PD group had significantly worse functional outcomes at 90 days than the non-PD group (mRS 0–2: 3/8 [37.5 %] vs. 23/26 [88.5 %], p = 0.009; mRS 0–1: 2/8 [25.0 %] vs. 19/26 [73.1 %], p = 0.03).</div></div><div><h3>Conclusions</h3><div>The prediction of progressive stroke and/or disabling stroke based on perfusion imaging parameters upon admission would allow triage of patients requiring endovascular treatment.</div></div>","PeriodicalId":12063,"journal":{"name":"European Journal of Radiology","volume":"195 ","pages":"Article 112617"},"PeriodicalIF":3.3,"publicationDate":"2025-12-15","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145780484","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

Interpretable fusion deep learning on super-resolution MRI for perineural invasion prediction in pancreatic ductal adenocarcinoma: a multicenter study 基于超分辨率MRI的可解释融合深度学习预测胰腺导管腺癌的神经周围浸润：一项多中心研究。

IF 3.3 3区医学 Q1 RADIOLOGY, NUCLEAR MEDICINE & MEDICAL IMAGING

European Journal of Radiology

Pub Date : 2025-12-15 DOI: 10.1016/j.ejrad.2025.112610

Jianbo Li , Jie Xu , Tingting Shen , Jianjun Zhou , Peng Wang , Tianyu Lu , Dalong Li , Mengsu Zeng , Qiying Tang , Qiuyang Zhu , Haitao Sun

Purpose

To develop an interpretable fusion deep learning model based on super-resolution (SR) MRI for predicting preoperative perineural invasion (PNI) in pancreatic ductal adenocarcinoma (PDAC) and to evaluate its role in guiding postoperative prognostic and therapeutic decision-making.

Materials and methods

In this multicenter retrospective study, we enrolled 714 eligible patients, allocating 608 to a development/internal validation set and 106 from three external centers to an external validation set. The fusion clinical-radiomics-deep transfer learning (FCDR) was developed to predict PNI by integrating imaging signatures derived from deep learning and radiomics on SR-MRI with clinical risk factors and was optimized by selecting the best-performing among seven machine learning algorithms. The final model was subsequently validated for its incremental value in predicting postoperative prognosis and guiding adjuvant therapy.

Results

The FCDR model achieved superior performance of PNI prediction with AUCs of 0.929, 0.886, and 0.832 across development, internal and external validation sets, significantly outperforming single clinical, DL, or radiomics models. The FCDR model-stratified high-risk group was associated with significantly worse postoperative OS and RFS (p < 0.05). Moreover, the high-risk PNI subgroup stratified by this model derived a significant OS and RFS benefit from adjuvant therapy. Model interpretability was affirmed by SHAP analysis.

Conclusion

The proposed interpretable fusion model serves as an effective tool for PNI evaluation, prognostic stratification, and tailoring of adjuvant therapy in PDAC, holding significant promise for personalized precision medicine.

Critical relevance statement

The constructed fusion model offers a robust, non-invasive tool for identifying PNI in pancreatic ductal adenocarcinoma, showing significant potential to guide personalized treatment strategies and improve patient outcomes.

目的：建立一种基于超分辨率（SR） MRI的可解释融合深度学习模型，用于预测胰腺导管腺癌（PDAC）术前神经周围浸润（PNI），并评估其在指导术后预后和治疗决策中的作用。材料和方法：在这项多中心回顾性研究中，我们纳入了714例符合条件的患者，其中608例分配到开发/内部验证组，106例分配到三个外部中心的外部验证组。融合临床-放射组学-深度迁移学习（FCDR）通过整合深度学习和放射组学在SR-MRI上获得的成像特征与临床危险因素来预测PNI，并通过在7种机器学习算法中选择表现最佳的算法进行优化。最终模型在预测术后预后和指导辅助治疗方面的增量价值随后得到验证。结果：FCDR模型在PNI预测方面表现优异，开发、内部和外部验证集的auc分别为0.929、0.886和0.832，显著优于单一临床、DL或放射组学模型。结论：提出的可解释融合模型可作为PNI评估、预后分层和PDAC辅助治疗定制的有效工具，在个性化精准医疗方面具有重要前景。关键相关性声明：构建的融合模型为识别胰腺导管腺癌的PNI提供了一个强大的、非侵入性的工具，显示出指导个性化治疗策略和改善患者预后的巨大潜力。

{"title":"Interpretable fusion deep learning on super-resolution MRI for perineural invasion prediction in pancreatic ductal adenocarcinoma: a multicenter study","authors":"Jianbo Li , Jie Xu , Tingting Shen , Jianjun Zhou , Peng Wang , Tianyu Lu , Dalong Li , Mengsu Zeng , Qiying Tang , Qiuyang Zhu , Haitao Sun","doi":"10.1016/j.ejrad.2025.112610","DOIUrl":"10.1016/j.ejrad.2025.112610","url":null,"abstract":"<div><h3>Purpose</h3><div>To develop an interpretable fusion deep learning model based on super-resolution (SR) MRI for predicting preoperative perineural invasion (PNI) in pancreatic ductal adenocarcinoma (PDAC) and to evaluate its role in guiding postoperative prognostic and therapeutic decision-making.</div></div><div><h3>Materials and methods</h3><div>In this multicenter retrospective study, we enrolled 714 eligible patients, allocating 608 to a development/internal validation set and 106 from three external centers to an external validation set. The fusion clinical-radiomics-deep transfer learning (FCDR) was developed to predict PNI by integrating imaging signatures derived from deep learning and radiomics on SR-MRI with clinical risk factors and was optimized by selecting the best-performing among seven machine learning algorithms. The final model was subsequently validated for its incremental value in predicting postoperative prognosis and guiding adjuvant therapy.</div></div><div><h3>Results</h3><div>The FCDR model achieved superior performance of PNI prediction with AUCs of 0.929, 0.886, and 0.832 across development, internal and external validation sets, significantly outperforming single clinical, DL, or radiomics models. The FCDR model-stratified high-risk group was associated with significantly worse postoperative OS and RFS (p < 0.05). Moreover, the high-risk PNI subgroup stratified by this model derived a significant OS and RFS benefit from adjuvant therapy. Model interpretability was affirmed by SHAP analysis.</div></div><div><h3>Conclusion</h3><div>The proposed interpretable fusion model serves as an effective tool for PNI evaluation, prognostic stratification, and tailoring of adjuvant therapy in PDAC, holding significant promise for personalized precision medicine.</div></div><div><h3>Critical relevance statement</h3><div>The constructed fusion model offers a robust, non-invasive tool for identifying PNI in pancreatic ductal adenocarcinoma, showing significant potential to guide personalized treatment strategies and improve patient outcome<em>s.</em></div></div>","PeriodicalId":12063,"journal":{"name":"European Journal of Radiology","volume":"195 ","pages":"Article 112610"},"PeriodicalIF":3.3,"publicationDate":"2025-12-15","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145793608","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

Pericoronary adipose tissue attenuation on CCTA as a marker of cardiovascular risk: A systematic review and meta-analysis 冠状动脉周围脂肪组织衰减在CCTA上作为心血管风险的标志：一项系统回顾和荟萃分析。

IF 3.3 3区医学 Q1 RADIOLOGY, NUCLEAR MEDICINE & MEDICAL IMAGING

European Journal of Radiology

Pub Date : 2025-12-15 DOI: 10.1016/j.ejrad.2025.112607

Navraj S. Sagoo , Rajveer Sagoo , Jaden LeGate , Mohanakrishnan Sathyamoorthy

Background

Pericoronary adipose tissue (PCAT) attenuation on coronary computed tomography angiography (CCTA) reflects local vascular inflammation, but its prognostic performance across heterogeneous imaging protocols and follow-up durations remains uncertain.

Methods

MEDLINE via Pubmed, Scopus, and Cochrane were searched through June 2025 for longitudinal CCTA studies reporting multivariable-adjusted associations between attenuation-based PCAT metrics and major adverse cardiovascular events (MACE). Hazard ratios (HRs) were standardized to a per–1–Hounsfield unit (HU) scale and pooled using random-effects models, with estimates presented per 5 HU. Follow-up was stratified as short (≤3 years), intermediate (>3–5 years), and long-term (≥5 years). Studies evaluating fat attenuation index (FAI), change in FAI (ΔFAI), thresholds, or artificial intelligence (AI)–derived perivascular metrics were synthesized qualitatively.

Results

Seventeen studies (n = 57,862) met inclusion criteria; seven cohorts (n = 5,922) contributed to the HU-based meta-analysis. Higher PCAT attenuation was associated with increased MACE risk (HR 1.29 per 5 HU; 95 % confidence interval [CI] 1.09–1.53; I²≈84 %). Across follow-up strata, effect estimates remained directionally consistent: short-term HR 1.22 per 5 HU (95 % CI 1.04–1.43), intermediate-term HR 1.38 (95 % CI 0.98–1.95), and long-term HR 1.50 (95 % CI 0.86–2.62), without significant between-group differences. Subgroup analyses showed similar effect sizes in studies using standardized right coronary artery (RCA) PCAT versus lesion-specific or multivessel approaches. Ten additional studies evaluating FAI, ΔFAI, HU thresholds, or composite or AI-based metrics showed higher risk with more inflamed perivascular attenuation phenotypes.

Conclusion

Higher PCAT attenuation is associated with increased MACE risk across diverse imaging platforms and follow-up horizons and appears to be a marker of inflammation-related coronary vulnerability.

背景：冠状动脉ct血管造影（CCTA）显示冠状动脉周围脂肪组织（PCAT）衰减反映了局部血管炎症，但其在不同成像方案和随访时间中的预后表现仍不确定。方法：MEDLINE通过Pubmed、Scopus和Cochrane检索了截至2025年6月的纵向CCTA研究，报告了基于衰减的PCAT指标与主要不良心血管事件（MACE）之间的多变量调整相关性。将风险比（hr）标准化为每1个hounsfield单位（HU），并使用随机效应模型进行汇总，每5个HU给出估计。随访分为短期（≤3年）、中期（3 ~ 5年）和长期（≥5年）。评估脂肪衰减指数（FAI）、FAI变化（ΔFAI）、阈值或人工智能（AI）衍生的血管周围指标的研究进行了定性综合。结果：17项研究（n = 57,862）符合纳入标准；7个队列（n = 5,922）参与了基于hu的meta分析。较高的PCAT衰减与MACE风险增加相关（HR 1.29 / 5 HU； 95%可信区间[CI] 1.09-1.53; I2≈84%）。在随访阶段，效果估计保持方向一致：短期HR为1.22 / 5 HU (95% CI 1.04-1.43)，中期HR为1.38 (95% CI 0.98-1.95)，长期HR为1.50 (95% CI 0.86-2.62)，组间无显著差异。亚组分析显示，标准化右冠状动脉（RCA） PCAT与病变特异性或多血管入路的效应大小相似。另外10项评估FAI、ΔFAI、HU阈值或综合或基于ai的指标的研究显示，血管周围炎症衰减表型越严重，风险越高。结论：在不同的成像平台和随访期间，较高的PCAT衰减与MACE风险增加相关，似乎是炎症相关冠状动脉易损性的标志。

{"title":"Pericoronary adipose tissue attenuation on CCTA as a marker of cardiovascular risk: A systematic review and meta-analysis","authors":"Navraj S. Sagoo , Rajveer Sagoo , Jaden LeGate , Mohanakrishnan Sathyamoorthy","doi":"10.1016/j.ejrad.2025.112607","DOIUrl":"10.1016/j.ejrad.2025.112607","url":null,"abstract":"<div><h3>Background</h3><div>Pericoronary adipose tissue (PCAT) attenuation on coronary computed tomography angiography (CCTA) reflects local vascular inflammation, but its prognostic performance across heterogeneous imaging protocols and follow-up durations remains uncertain.</div></div><div><h3>Methods</h3><div>MEDLINE via Pubmed, Scopus, and Cochrane were searched through June 2025 for longitudinal CCTA studies reporting multivariable-adjusted associations between attenuation-based PCAT metrics and major adverse cardiovascular events (MACE). Hazard ratios (HRs) were standardized to a per–1–Hounsfield unit (HU) scale and pooled using random-effects models, with estimates presented per 5 HU. Follow-up was stratified as short (≤3 years), intermediate (>3–5 years), and long-term (≥5 years). Studies evaluating fat attenuation index (FAI), change in FAI (ΔFAI), thresholds, or artificial intelligence (AI)–derived perivascular metrics were synthesized qualitatively.</div></div><div><h3>Results</h3><div>Seventeen studies (n = 57,862) met inclusion criteria; seven cohorts (n = 5,922) contributed to the HU-based <em>meta</em>-analysis. Higher PCAT attenuation was associated with increased MACE risk (HR 1.29 per 5 HU; 95 % confidence interval [CI] 1.09–1.53; I<sup>2</sup>≈84 %). Across follow-up strata, effect estimates remained directionally consistent: short-term HR 1.22 per 5 HU (95 % CI 1.04–1.43), intermediate-term HR 1.38 (95 % CI 0.98–1.95), and long-term HR 1.50 (95 % CI 0.86–2.62), without significant between-group differences. Subgroup analyses showed similar effect sizes in studies using standardized right coronary artery (RCA) PCAT versus lesion-specific or multivessel approaches. Ten additional studies evaluating FAI, ΔFAI, HU thresholds, or composite or AI-based metrics showed higher risk with more inflamed perivascular attenuation phenotypes.</div></div><div><h3>Conclusion</h3><div>Higher PCAT attenuation is associated with increased MACE risk across diverse imaging platforms and follow-up horizons and appears to be a marker of inflammation-related coronary vulnerability.</div></div>","PeriodicalId":12063,"journal":{"name":"European Journal of Radiology","volume":"195 ","pages":"Article 112607"},"PeriodicalIF":3.3,"publicationDate":"2025-12-15","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145780593","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

Diagnostic Performance of AI-Assisted Coronary CT Angiography: A Systematic Review and Meta-Analysis 人工智能辅助冠状动脉CT血管造影的诊断性能：系统回顾和荟萃分析

IF 3.3 3区医学 Q1 RADIOLOGY, NUCLEAR MEDICINE & MEDICAL IMAGING

European Journal of Radiology

Pub Date : 2025-12-14 DOI: 10.1016/j.ejrad.2025.112612

Abdelrahman Hafez , Ahmed Sobhy , Mennatullah Ashour , Karim Aiash , Abdulrahman Aldemerdash , Amro Mahmoud Radi , Amir Elissawy , Khaled Zahrawi , Tawfik Besheya , Richard Riccelli , Ayaan Arora , Dina Al-Mahmoud , Jessan Jishu , Eman Toraih , Hani Aiash

Background

Coronary computed tomography angiography (CCTA) is vital for diagnosing ischemic heart disease, yet its accuracy is unreliable due to varying reader expertise. Artificial Intelligence (AI)-driven automated stenosis assessment offers promise for enhancing diagnostic consistency. We aim to evaluate an AI-CCTA assessment against invasive coronary angiography, invasive FFR, and expert readings.

Methods

We performed a comprehensive search in Web of Science, Scopus, PubMed, Cochrane Library, and EMBASE from inception until March 2025. Two independent reviewers screened articles and extracted data on study design, patient demographics, AI methodology, stenosis thresholds, and outcomes. For statistical analysis, we constructed summary receiver operating characteristic (SROC) curves and used a bivariate random-effects model to derive pooled sensitivity, specificity, diagnostic odds ratios (DOR), and area under the curve (AUC). Forest plots were generated to visualize these metrics.

Results

Our meta-analysis included 34 studies with 10,067 patients. AI-based CCTA demonstrated excellent diagnostic performance with an AUC of 0.932 for per-patient analysis. The pooled per-patient sensitivity was 0.89 (95 % CI: 0.87–0.91) and specificity was 0.80 (95 % CI: 0.74–0.86) with diagnostic OR of 37.07 (95 % CI: 24.57–55.92). AI validated against expert readers achieved the highest accuracy (0.94, 95 % CI: 0.87–0.98). The > 70 % stenosis threshold demonstrated superior performance (accuracy: 0.90, specificity: 0.96) compared to the > 50 % threshold (accuracy: 0.85, specificity: 0.87). Per-vessel analysis showed comparable results with an AUC of 0.905.

Conclusion

AI-assisted coronary CT angiography delivers high diagnostic performance for coronary stenosis detection, with strong AUC values, high sensitivity and specificity, and robust diagnostic OR across both per-patient and per-vessel assessments.

冠状动脉ct血管造影（CCTA）对于诊断缺血性心脏病至关重要，但由于读者的专业知识不同，其准确性并不可靠。人工智能（AI）驱动的自动狭窄评估为提高诊断一致性提供了希望。我们的目标是评估AI-CCTA对侵入性冠状动脉造影、侵入性FFR和专家读数的评估。方法综合检索Web of Science、Scopus、PubMed、Cochrane Library、EMBASE等数据库，检索时间为Web of Science成立至2025年3月。两名独立审稿人筛选了文章并提取了研究设计、患者人口统计学、人工智能方法、狭窄阈值和结果方面的数据。为了进行统计分析，我们构建了总受试者工作特征（SROC）曲线，并使用双变量随机效应模型得出合并敏感性、特异性、诊断优势比（DOR）和曲线下面积（AUC）。生成森林图来可视化这些指标。我们的荟萃分析包括34项研究，10,067例患者。基于人工智能的CCTA表现出优异的诊断性能，每例分析的AUC为0.932。每个患者的敏感性为0.89 (95% CI: 0.87-0.91)，特异性为0.80 (95% CI: 0.74-0.86)，诊断OR为37.07 （95% CI: 24.57-55.92）。针对专家读者验证的人工智能达到了最高的准确率（0.94,95% CI: 0.87-0.98）。与>； 50%阈值（准确性：0.85，特异性：0.87）相比，>； 70%狭窄阈值表现出更好的性能（准确性：0.90，特异性：0.96）。每根血管的AUC为0.905。结论人工智能辅助冠状动脉CT血管造影对冠状动脉狭窄检测具有较高的诊断性能，具有较强的AUC值、较高的敏感性和特异性，在每例患者和每条血管评估中均具有较强的诊断OR。

{"title":"Diagnostic Performance of AI-Assisted Coronary CT Angiography: A Systematic Review and Meta-Analysis","authors":"Abdelrahman Hafez , Ahmed Sobhy , Mennatullah Ashour , Karim Aiash , Abdulrahman Aldemerdash , Amro Mahmoud Radi , Amir Elissawy , Khaled Zahrawi , Tawfik Besheya , Richard Riccelli , Ayaan Arora , Dina Al-Mahmoud , Jessan Jishu , Eman Toraih , Hani Aiash","doi":"10.1016/j.ejrad.2025.112612","DOIUrl":"10.1016/j.ejrad.2025.112612","url":null,"abstract":"<div><h3>Background</h3><div>Coronary computed tomography angiography (CCTA) is vital for diagnosing ischemic heart disease, yet its accuracy is unreliable due to varying reader expertise. Artificial Intelligence (AI)-driven automated stenosis assessment offers promise for enhancing diagnostic consistency. We aim to evaluate an AI-CCTA assessment against invasive coronary angiography, invasive FFR, and expert readings.</div></div><div><h3>Methods</h3><div>We performed a comprehensive search in Web of Science, Scopus, PubMed, Cochrane Library, and EMBASE from inception until March 2025. Two independent reviewers screened articles and extracted data on study design, patient demographics, AI methodology, stenosis thresholds, and outcomes. For statistical analysis, we constructed summary receiver operating characteristic (SROC) curves and used a bivariate random-effects model to derive pooled sensitivity, specificity, diagnostic odds ratios (DOR), and area under the curve (AUC). Forest plots were generated to visualize these metrics.</div></div><div><h3>Results</h3><div>Our <em>meta</em>-analysis included 34 studies with 10,067 patients. AI-based CCTA demonstrated excellent diagnostic performance with an AUC of 0.932 for per-patient analysis. The pooled per-patient sensitivity was 0.89 (95 % CI: 0.87–0.91) and specificity was 0.80 (95 % CI: 0.74–0.86) with diagnostic OR of 37.07 (95 % CI: 24.57–55.92). AI validated against expert readers achieved the highest accuracy (0.94, 95 % CI: 0.87–0.98). The > 70 % stenosis threshold demonstrated superior performance (accuracy: 0.90, specificity: 0.96) compared to the > 50 % threshold (accuracy: 0.85, specificity: 0.87). Per-vessel analysis showed comparable results with an AUC of 0.905.</div></div><div><h3>Conclusion</h3><div>AI-assisted coronary CT angiography delivers high diagnostic performance for coronary stenosis detection, with strong AUC values, high sensitivity and specificity, and robust diagnostic OR across both per-patient and per-vessel assessments.</div></div>","PeriodicalId":12063,"journal":{"name":"European Journal of Radiology","volume":"195 ","pages":"Article 112612"},"PeriodicalIF":3.3,"publicationDate":"2025-12-14","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145797246","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

Characterization of middle ear ossicular ligaments using photon counting detector CT 光子计数检测器CT表征中耳听骨韧带。

IF 3.3 3区医学 Q1 RADIOLOGY, NUCLEAR MEDICINE & MEDICAL IMAGING

European Journal of Radiology

Pub Date : 2025-12-13 DOI: 10.1016/j.ejrad.2025.112613

Dinesh Rao , Richard D Beegle , Jeet Patel , John V Murray , Mallory J Raymond , Joseph T Breen , Sukhwinder J.S. Sandhu , Vishal N Patel

Background and Purpose

The middle ear suspensory ossicular ligaments have been previously described on traditional energy integrating detector CT scanners. With recent developments in photon counting detector CT, these ligaments can be better visualized due to greater spatial and contrast resolution.

Materials and Methods

50 patients (100 temporalbones) without hearing loss or middle ear cavity pathology were retrospectively reviewed to characterize the middle ear suspensory ossicular ligaments including the anterior, superior, lateral mallear, and posterior incudal ligaments, and stapes and tensor tympani tendons. Structures were classified as: 0 − not visualized, 1 − thin, 2 − thick, 3 – calcified, 4 – bar. Interrater agreement statistics and associated significance levels are reported.

Results

The anterior and superior mallear ligaments have more complex anatomy and variability than what has been published previously in the radiology literature. The anterior mallear ligament forms a complex with the chorda tympani and the discomallelar and sphenomandibular ligaments. An accessory superior anterior mallear ligament was observed in 67 of 100 subjects. Variant anterior and posterior superior mallear ligaments were observed in 28 and 38 temporal bones, respectively. The posterior incudal ligament was virtually always present and thick, while the lateral mallear ligament was present and thin. The stapes tendon was usually thin, while the tensor tympani tendon was thick. Calcified ligaments and bars were observed in a minority of the ligaments.

Conclusions

Photon counting CT allows for characterization of the ossicular ligaments which is qualitatively consistent with anatomic studies that demonstrate greater complexity than what has been described in the radiology literature.

Clinical Impact: Radiologists and otologists must recognize the increased complexity and variability of middle ear ligaments when using photon counting CT, particularly as it becomes more widely adopted and is suited for imaging of the temporal bone. The findings presented in this manuscript allow for recognition of this anatomy which can facilitate image interpretation.

Abbreviations: AML: anterior mallear ligament; S-AML, superior anterior mallear ligament; M−AML, middle anterior mallear ligament; CHL, conductive hearing loss; ChT, chorda tympani; DML, discomallelear ligament; EID-CT, energy Integrated Detector CT; I, Incus; M, Malleus; LML, lateral mallear ligament; MEC: middle ear cavity; PCCT, photon counting detector CT; PIL, posterior incudal ligament; PTF, petrotympanic fissure; SML, superior mallear ligament; SpML, sphenomandibular ligament; A-SML, anterior superior mallear ligament; P-AML, posterior superior mallear ligament; SNHL, sensorineural hearing loss; ST, stapedius tendon; TTT, tensor tympani tendon.

背景与目的：中耳悬垂听骨韧带已在传统的能量积分检测器CT扫描仪上被描述。随着光子计数检测器CT的最新发展，由于更大的空间和对比度分辨率，这些韧带可以更好地可视化。材料和方法：回顾性分析无听力损失和中耳腔病理的50例（100块颞骨）中耳悬垂听骨韧带的特征，包括前、上、外侧、后内侧韧带以及镫骨和鼓室张肌腱。结构分为：0 -不可见，1 -薄，2 -厚，3 -钙化，4 -棒状。报告了相互间协议统计和相关显著性水平。结果：前韧带和上韧带具有比先前在放射学文献中发表的更复杂的解剖结构和变异性。外阴前韧带与鼓室索、韧带紊乱和下颌韧带形成复合体。100例受试者中有67例出现副上前外阴韧带。颞骨前、后上韧带变异28例，颞骨上韧带变异38例。后内侧韧带几乎总是存在且很厚，而外踝韧带存在且很薄。镫骨腱通常较薄，而鼓室张肌腱较厚。在少数韧带中观察到钙化的韧带和条状。结论：光子计数CT允许表征听骨韧带，其定性与解剖研究一致，证明比放射学文献中描述的更复杂。临床影响：放射科医生和耳科医生在使用光子计数CT时必须认识到中耳韧带日益增加的复杂性和可变性，特别是当光子计数CT被广泛采用并适用于颞骨成像时。在这份手稿中提出的发现允许识别这种解剖结构，这可以促进图像解释。AML：前外阴韧带；S-AML，上前核韧带；M-AML，中前外阴韧带；CHL：传导性听力损失；ChT，定音弦；DML，斜韧带；EID-CT，能量集成检测器CT；我,砧骨;米,锤骨;LML，外核韧带；MEC：中耳腔；PCCT，光子计数检测器CT；PIL，后内侧韧带；PTF，岩石鼓室裂隙；SML，颧上韧带；SpML，蝶下颌韧带；A-SML，前上核韧带；P-AML，外阴后上韧带；SNHL，感音神经性听力损失；ST，镫骨肌腱；TTT，鼓室腱张量。

{"title":"Characterization of middle ear ossicular ligaments using photon counting detector CT","authors":"Dinesh Rao , Richard D Beegle , Jeet Patel , John V Murray , Mallory J Raymond , Joseph T Breen , Sukhwinder J.S. Sandhu , Vishal N Patel","doi":"10.1016/j.ejrad.2025.112613","DOIUrl":"10.1016/j.ejrad.2025.112613","url":null,"abstract":"<div><h3>Background and Purpose</h3><div>The middle ear suspensory ossicular ligaments have been previously described on traditional energy integrating detector CT scanners. With recent developments in photon counting detector CT, these ligaments can be better visualized due to greater spatial and contrast resolution.</div></div><div><h3>Materials and Methods</h3><div>50 patients (100 temporalbones) without hearing loss or middle ear cavity pathology were retrospectively reviewed to characterize the middle ear suspensory ossicular ligaments including the anterior, superior, lateral mallear, and posterior incudal ligaments, and stapes and tensor tympani tendons. Structures were classified as: 0 − not visualized, 1 − thin, 2 − thick, 3 – calcified, 4 – bar. Interrater agreement statistics and associated significance levels are reported.</div></div><div><h3>Results</h3><div>The anterior and superior mallear ligaments have more complex anatomy and variability than what has been published previously in the radiology literature. The anterior mallear ligament forms a complex with the chorda tympani and the discomallelar and sphenomandibular ligaments. An accessory superior anterior mallear ligament was observed in 67 of 100 subjects. Variant anterior and posterior superior mallear ligaments were observed in 28 and 38 temporal bones, respectively. The posterior incudal ligament was virtually always present and thick, while the lateral mallear ligament was present and thin. The stapes tendon was usually thin, while the tensor tympani tendon was thick. Calcified ligaments and bars were observed in a minority of the ligaments.</div></div><div><h3>Conclusions</h3><div>Photon counting CT allows for characterization of the ossicular ligaments which is qualitatively consistent with anatomic studies that demonstrate greater complexity than what has been described in the radiology literature.</div><div>Clinical Impact: Radiologists and otologists must recognize the increased complexity and variability of middle ear ligaments when using photon counting CT, particularly as it becomes more widely adopted and is suited for imaging of the temporal bone. The findings presented in this manuscript allow for recognition of this anatomy which can facilitate image interpretation.</div><div>Abbreviations: AML: anterior mallear ligament; S-AML, superior anterior mallear ligament; M−AML, middle anterior mallear ligament; CHL, conductive hearing loss; ChT, chorda tympani; DML, discomallelear ligament; EID-CT, energy Integrated Detector CT; I, Incus; M, Malleus; LML, lateral mallear ligament; MEC: middle ear cavity; PCCT, photon counting detector CT; PIL, posterior incudal ligament; PTF, petrotympanic fissure; SML, superior mallear ligament; SpML, sphenomandibular ligament; A-SML, anterior superior mallear ligament; P-AML, posterior superior mallear ligament; SNHL, sensorineural hearing loss; ST, stapedius tendon; TTT, tensor tympani tendon.</div></div>","PeriodicalId":12063,"journal":{"name":"European Journal of Radiology","volume":"195 ","pages":"Article 112613"},"PeriodicalIF":3.3,"publicationDate":"2025-12-13","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145780526","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

Predictive value of quantitative parameters from dual-tracer FDG and PSMA PET/CT for progression to castration resistance in patients with metastatic hormone-sensitive prostate cancer 双示踪剂FDG和PSMA PET/CT定量参数对转移性激素敏感前列腺癌患者进展为去势抵抗的预测价值

IF 3.3 3区医学 Q1 RADIOLOGY, NUCLEAR MEDICINE & MEDICAL IMAGING

European Journal of Radiology

Pub Date : 2025-12-13 DOI: 10.1016/j.ejrad.2025.112614

Fangzheng Xiang , Yu Zeng , Zhiqiang Zhang , Yubin Zhang , Wenhui Su , Mingkuan Zhou , Jiajie Yu , Cheng Luo , Yukun Wu , Fufu Zheng

Purposes

To explore the prognostic value of dual-tracer PET/CT-derived parameters and develop a predictive model for castration-resistant prostate cancer-free survival (CRPC-FS) in patients with metastatic hormone-sensitive prostate cancer (mHSPC).

Methods

Forty-five mHSPC patients who underwent prostate-specific membrane antigen (PSMA) and fluorodeoxyglucose (FDG) PET/CT prior to novel hormone therapy (NHT) were included between October 2020 and March 2024. A total of 1,661 PSMA-positive and 303 FDG-positive lesions were evaluated to quantify tumor burden, including maximum standardized uptake value, metabolic tumor volume (MTV) and total lesion uptake under different relative fixed threshold values. Cox regression and Kaplan–Meier analyses assessed their association with CRPC-FS. Internal validation was performed using 1000-bootstrap resampling. Classification and regression tree identified optimal cut-offs. Time-dependent receiver operating characteristic analysis evaluated model discrimination.

Results

In univariate analysis, only MTV of prostate cancer was significantly associated with CRPC-FS (p < 0.05). PSMA-derived (C-index = 0.662) MTV and FDG-derived (C-index = 0.712) MTV under a 40 % threshold exhibited the strongest predictive power and remained independent in multivariable analysis. Internal validation demonstrated favorable model performance and good calibration. The cut-off values for the two parameters were identified as 4.77 and 66.425, respectively. Furthermore, patients were stratified into prognostic groups to develop a risk model with AUCs > 0.85.

Conclusions

MTV exhibited superior prognostic value for CRPC-FS in patients with mHSPC compared to other PET/CT-derived and clinical parameters. Moreover, FDG-derived MTV showed stronger prognostic value than PSMA-derived MTV. The dual-tracer PET/CT-derived model could comprehensively assess tumor burden and aid in early prognostic stratification.

目的探讨PET/ ct双示踪参数的预后价值，建立转移性激素敏感性前列腺癌（mHSPC）患者去雄抵抗性前列腺无癌生存（CRPC-FS）的预测模型。方法纳入2020年10月至2024年3月期间接受新型激素治疗（NHT）前前列腺特异性膜抗原（PSMA）和氟脱氧葡萄糖（FDG） PET/CT检查的45例mHSPC患者。共评估1661例psma阳性和303例fdg阳性病变，量化肿瘤负担，包括不同相对固定阈值下的最大标准化摄取值、代谢肿瘤体积（MTV）和病变总摄取。Cox回归和Kaplan-Meier分析评估了它们与CRPC-FS的相关性。内部验证使用1000次bootstrap重采样进行。分类和回归树确定了最优截止点。时变接收机工作特性分析评价了模型判别。结果单因素分析中，只有前列腺癌MTV与CRPC-FS有显著相关性（p < 0.05）。在40%阈值下，psma衍生的MTV （C-index = 0.662）和fdg衍生的MTV （C-index = 0.712）的预测能力最强，在多变量分析中保持独立。内部验证证明了良好的模型性能和良好的校准。这两个参数的临界值分别为4.77和66.425。此外，将患者分为预后组，建立auc >； 0.85的风险模型。结论与其他PET/ ct衍生和临床参数相比，smtv对mHSPC患者的CRPC-FS具有更好的预后价值。此外，fdg衍生的MTV比psma衍生的MTV具有更强的预测价值。PET/ ct衍生的双示踪模型可以全面评估肿瘤负荷并有助于早期预后分层。

{"title":"Predictive value of quantitative parameters from dual-tracer FDG and PSMA PET/CT for progression to castration resistance in patients with metastatic hormone-sensitive prostate cancer","authors":"Fangzheng Xiang , Yu Zeng , Zhiqiang Zhang , Yubin Zhang , Wenhui Su , Mingkuan Zhou , Jiajie Yu , Cheng Luo , Yukun Wu , Fufu Zheng","doi":"10.1016/j.ejrad.2025.112614","DOIUrl":"10.1016/j.ejrad.2025.112614","url":null,"abstract":"<div><h3>Purposes</h3><div>To explore the prognostic value of dual-tracer PET/CT-derived parameters and develop a predictive model for castration-resistant prostate cancer-free survival (CRPC-FS) in patients with metastatic hormone-sensitive prostate cancer (mHSPC).</div></div><div><h3>Methods</h3><div>Forty-five mHSPC patients who underwent prostate-specific membrane antigen (PSMA) and fluorodeoxyglucose (FDG) PET/CT prior to novel hormone therapy (NHT) were included between October 2020 and March 2024. A total of 1,661 PSMA-positive and 303 FDG-positive lesions were evaluated to quantify tumor burden, including maximum standardized uptake value, metabolic tumor volume (MTV) and total lesion uptake under different relative fixed threshold values. Cox regression and Kaplan–Meier analyses assessed their association with CRPC-FS. Internal validation was performed using 1000-bootstrap resampling. Classification and regression tree identified optimal cut-offs. Time-dependent receiver operating characteristic analysis evaluated model discrimination.</div></div><div><h3>Results</h3><div>In univariate analysis, only MTV of prostate cancer was significantly associated with CRPC-FS (<em>p</em> < 0.05). PSMA-derived (C-index = 0.662) MTV and FDG-derived (C-index = 0.712) MTV under a 40 % threshold exhibited the strongest predictive power and remained independent in multivariable analysis. Internal validation demonstrated favorable model performance and good calibration. The cut-off values for the two parameters were identified as 4.77 and 66.425, respectively. Furthermore, patients were stratified into prognostic groups to develop a risk model with AUCs > 0.85.</div></div><div><h3>Conclusions</h3><div>MTV exhibited superior prognostic value for CRPC-FS in patients with mHSPC compared to other PET/CT-derived and clinical parameters. Moreover, FDG-derived MTV showed stronger prognostic value than PSMA-derived MTV. The dual-tracer PET/CT-derived model could comprehensively assess tumor burden and aid in early prognostic stratification.</div></div>","PeriodicalId":12063,"journal":{"name":"European Journal of Radiology","volume":"195 ","pages":"Article 112614"},"PeriodicalIF":3.3,"publicationDate":"2025-12-13","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145797239","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

A combination of High-Pitch turbo FLASH mode and iterative Beam-Hardening correction in cranial computed tomography Facilitates acute small subdural hematoma visualization 颅计算机断层扫描中的高频涡轮闪光模式和迭代波束硬化校正的结合有助于急性小硬膜下血肿的可视化。

IF 3.3 3区医学 Q1 RADIOLOGY, NUCLEAR MEDICINE & MEDICAL IMAGING

European Journal of Radiology

Pub Date : 2025-12-13 DOI: 10.1016/j.ejrad.2025.112611

Xinying Zhang , Yongxin Ou , Qiya He , Xiaotong Xie , Chunyan Wang , Zhen Wu , Lingjiang Mo , Hui Zeng , Jiuping Liang , Li’an Luo , Lveqin Huang , Liheng Ma

Objectives

To assess the impact of combined high-pitch turbo FLASH mode and iterative beam-hardening correction (iBHC) on image quality and diagnostic performance for detecting acute small subdural hematoma in brain CT.

Materials and methods

Fifty-three patients received unenhanced CT scans in FLASH mode, while forty-six patients received scans in non-FLASH mode. The scanning time, radiation dose, and motion artifacts were compared between the two techniques, and iBHC and non-iBHC reconstruction methods were applied to FLASH images. The severity of artifacts in the subcalvarial and posterior fossa was categorized into three grades for qualitative analysis. For quantitative analysis, CT values and standard deviations of gray matter, white matter, cerebrospinal fluid, the skull, and subdural hematoma were measured, and the contrast-to-noise ratio (CNR) and CT value difference rate were calculated. The differences in these parameters between the iBHC and non-iBHC groups were analyzed.

Results

The FLASH group demonstrated significantly reduced scan time, radiation dose, and motion artifacts compared with the non-FLASH group. In the FLASH scanning mode, the iBHC group had significantly lower severity grades of subcalvarial and posterior fossa artifacts than the non-iBHC group (P < 0.001). Moreover, the CNR between the gray matter and white matter and the CNR and CT value difference rates between the skull and subdural hematoma were significantly higher in the iBHC group (P < 0.05).

Conclusion

The combination of FLASH scanning and iBHC reconstruction effectively reduced artifacts, improved image contrast, and enhanced diagnostic confidence for detecting acute subdural hematomas.

目的：评价高频turbo FLASH模式联合迭代波束硬化校正（iBHC）对急性脑硬膜下小血肿CT图像质量和诊断性能的影响。材料与方法：53例患者行FLASH模式CT非增强扫描，46例患者行非FLASH模式扫描。比较两种技术的扫描时间、辐射剂量和运动伪影，并对FLASH图像应用iBHC和非iBHC重建方法。颅下和后窝假影的严重程度分为三个等级进行定性分析。定量分析时，测量脑灰质、脑白质、脑脊液、颅骨、硬膜下血肿的CT值及标准差，计算对比噪声比（CNR）和CT值差率。分析iBHC组和非iBHC组在这些参数上的差异。结果：与非FLASH组相比，FLASH组的扫描时间、辐射剂量和运动伪影明显减少。在FLASH扫描模式下，iBHC组的颅下和后窝伪影严重程度明显低于非iBHC组(P结论：FLASH扫描与iBHC重建相结合有效减少伪影，提高图像对比度，提高诊断急性硬膜下血肿的可信度。

{"title":"A combination of High-Pitch turbo FLASH mode and iterative Beam-Hardening correction in cranial computed tomography Facilitates acute small subdural hematoma visualization","authors":"Xinying Zhang , Yongxin Ou , Qiya He , Xiaotong Xie , Chunyan Wang , Zhen Wu , Lingjiang Mo , Hui Zeng , Jiuping Liang , Li’an Luo , Lveqin Huang , Liheng Ma","doi":"10.1016/j.ejrad.2025.112611","DOIUrl":"10.1016/j.ejrad.2025.112611","url":null,"abstract":"<div><h3>Objectives</h3><div>To assess the impact of combined high-pitch turbo FLASH mode and iterative beam-hardening correction (iBHC) on image quality and diagnostic performance for detecting acute small subdural hematoma in brain CT.</div></div><div><h3>Materials and methods</h3><div>Fifty-three patients received unenhanced CT scans in FLASH mode, while forty-six patients received scans in non-FLASH mode. The scanning time, radiation dose, and motion artifacts were compared between the two techniques, and iBHC and non-iBHC reconstruction methods were applied to FLASH images. The severity of artifacts in the subcalvarial and posterior fossa was categorized into three grades for qualitative analysis. For quantitative analysis, CT values and standard deviations of gray matter, white matter, cerebrospinal fluid, the skull, and subdural hematoma were measured, and the contrast-to-noise ratio (CNR) and CT value difference rate were calculated. The differences in these parameters between the iBHC and non-iBHC groups were analyzed.</div></div><div><h3>Results</h3><div>The FLASH group demonstrated significantly reduced scan time, radiation dose, and motion artifacts compared with the non-FLASH group. In the FLASH scanning mode, the iBHC group had significantly lower severity grades of subcalvarial and posterior fossa artifacts than the non-iBHC group (<em>P</em> < 0.001). Moreover, the CNR between the gray matter and white matter and the CNR and CT value difference rates between the skull and subdural hematoma were significantly higher in the iBHC group (<em>P</em> < 0.05).</div></div><div><h3>Conclusion</h3><div>The combination of FLASH scanning and iBHC reconstruction effectively reduced artifacts, improved image contrast, and enhanced diagnostic confidence for detecting acute subdural hematomas.</div></div>","PeriodicalId":12063,"journal":{"name":"European Journal of Radiology","volume":"195 ","pages":"Article 112611"},"PeriodicalIF":3.3,"publicationDate":"2025-12-13","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145910851","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

Evaluation of DeepSeek-R1 and contemporary large language models on the radiology board examination: A milestone achieved as open-source model matches performance with closed-source model 对DeepSeek-R1和当代大型语言模型在放射学板检查中的评估：开源模型与闭源模型的性能匹配，实现了一个里程碑。

IF 3.3 3区医学 Q1 RADIOLOGY, NUCLEAR MEDICINE & MEDICAL IMAGING

European Journal of Radiology

Pub Date : 2025-12-13 DOI: 10.1016/j.ejrad.2025.112608

Takeshi Nakaura , Naoki Kobayashi , Takanori Masuda , Yasunori Nagayama , Hiroyuki Uetani , Masafumi Kidoh , Seitaro Oda , Yoshinori Funama , Toshinori Hirai

Objectives

Recent advances in large language models (LLMs), especially reasoning LLMs have demonstrated impressive reasoning capabilities in specialized domains. The purpose of this study is to evaluate the performance of new open-source reasoning LLM, DeepSeek-R1 and other contemporary LLMs on radiology board examination questions, comparing their accuracy to human radiologists.

Materials and methods

We assessed 10 LLMs, including both closed-source models (GPT-4o, o1, o3-mini, Claude 3.5 Sonnet, Gemini Flash 2.0) and open-source models (DeepSeek-R1, its distilled versions, and Llama 3.3 70B), on 105 non-image multiple-choice questions from the 2024 official board examination of the Japan Radiological Society. We evaluated accuracies of LLMs and compared those to accuracies of examinees (3rd-year radiology residents).

Results

DeepSeek-R1 and OpenAI o1 each answered 92/105 items correctly (87.6 %, 95 % CI: 80.1–92.5); McNemar’s test showed no significant head-to-head difference (p = 1.00), outperforming human radiologists’ mean accuracy of 67.6 ± 9.3 % (z-score = 2.15, equivalent to 98.4th percentile). DeepSeek-R1 demonstrated superior cost-efficiency with API costs approximately 1/27th of o1. DeepSeek-R1′s distilled 32B model (61.9 % accuracy) outperformed Llama 3.3 70B (57.1 %), despite having fewer parameters.

Conclusion

Open-source DeepSeek-R1 matches the performance of top closed-source models while offering superior cost-efficiency in radiology knowledge assessment.

目的：大型语言模型（llm）的最新进展，特别是推理llm在特定领域展示了令人印象深刻的推理能力。本研究的目的是评估新型开源推理LLM、DeepSeek-R1和其他当代LLM在放射学委员会考试问题上的表现，并将其准确性与人类放射科医生进行比较。材料和方法：我们评估了10个llm，包括闭源模型（gpt - 40, 01, 03 -mini, Claude 3.5 Sonnet, Gemini Flash 2.0）和开源模型（DeepSeek-R1，其提炼版本和Llama 3.3 70B），对来自日本放射学会2024年官方委员会考试的105个非图像选择题进行了评估。我们评估了llm的准确性，并将其与考生（放射学三年级住院医师）的准确性进行了比较。结果：DeepSeek-R1和OpenAI o1分别正确率为92/105 (87.6%,95% CI: 80.1 ~ 92.5)；McNemar的检验结果显示没有显著的头对头差异（p = 1.00），优于人类放射科医生的平均准确率67.6±9.3% （z-score = 2.15，相当于98.4百分位）。DeepSeek-R1表现出卓越的成本效益，API成本约为1/27。DeepSeek-R1的蒸馏32B模型（准确率61.9%）优于Llama 3.3 70B(57.1%)，尽管参数更少。结论：开源DeepSeek-R1的性能与顶级闭源模型相当，同时在放射学知识评估中提供了优越的成本效益。

{"title":"Evaluation of DeepSeek-R1 and contemporary large language models on the radiology board examination: A milestone achieved as open-source model matches performance with closed-source model","authors":"Takeshi Nakaura , Naoki Kobayashi , Takanori Masuda , Yasunori Nagayama , Hiroyuki Uetani , Masafumi Kidoh , Seitaro Oda , Yoshinori Funama , Toshinori Hirai","doi":"10.1016/j.ejrad.2025.112608","DOIUrl":"10.1016/j.ejrad.2025.112608","url":null,"abstract":"<div><h3>Objectives</h3><div>Recent advances in large language models (LLMs), especially reasoning LLMs have demonstrated impressive reasoning capabilities in specialized domains. The purpose of this study is to evaluate the performance of new open-source reasoning LLM, DeepSeek-R1 and other contemporary LLMs on radiology board examination questions, comparing their accuracy to human radiologists.</div></div><div><h3>Materials and methods</h3><div>We assessed 10 LLMs, including both closed-source models (GPT-4o, o1, o3-mini, Claude 3.5 Sonnet, Gemini Flash 2.0) and open-source models (DeepSeek-R1, its distilled versions, and Llama 3.3 70B), on 105 non-image multiple-choice questions from the 2024 official board examination of the Japan Radiological Society. We evaluated accuracies of LLMs and compared those to accuracies of examinees (3rd-year radiology residents).</div></div><div><h3>Results</h3><div>DeepSeek-R1 and OpenAI o1 each answered 92/105 items correctly (87.6 %, 95 % CI: 80.1–92.5); McNemar’s test showed no significant head-to-head difference (p = 1.00), outperforming human radiologists’ mean accuracy of 67.6 ± 9.3 % (z-score = 2.15, equivalent to 98.4th percentile). DeepSeek-R1 demonstrated superior cost-efficiency with API costs approximately 1/27th of o1. DeepSeek-R1′s distilled 32B model (61.9 % accuracy) outperformed Llama 3.3 70B (57.1 %), despite having fewer parameters.</div></div><div><h3>Conclusion</h3><div>Open-source DeepSeek-R1 matches the performance of top closed-source models while offering superior cost-efficiency in radiology knowledge assessment.</div></div>","PeriodicalId":12063,"journal":{"name":"European Journal of Radiology","volume":"195 ","pages":"Article 112608"},"PeriodicalIF":3.3,"publicationDate":"2025-12-13","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145833613","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

Automated lung texture analysis for assessing interstitial lung disease in systemic sclerosis: Diagnostic accuracy in photon-counting-detector and conventional energy-integrating-detector CT 用于评估系统性硬化症间质性肺病的自动肺结构分析：光子计数检测器和传统能量积分检测器CT的诊断准确性。

IF 3.3 3区医学 Q1 RADIOLOGY, NUCLEAR MEDICINE & MEDICAL IMAGING

European Journal of Radiology

Pub Date : 2025-12-10 DOI: 10.1016/j.ejrad.2025.112605

Jasmin Happe , Cosimo Bruni , Lisa Jungblut , Nicholas Landini , Cecilia Strappa , Christian Bluethgen , Muriel Elhai , Rucsandra Dobrota , Carina Mihai , Sinziana Muraru , Anna-Maria Hoffmann-Vold , Anna Rita Larici , Thomas Frauenfelder , Oliver Distler , Jonas Kroschke

Objectives

To evaluate the performance of automated Lung Texture Analysis (LTA) in assessing interstitial lung disease (ILD) in systemic sclerosis (SSc) using low-dose photon-counting detector CT (PCD-CT) compared to conventional low-dose energy-integrating detector CT (EID-CT).

Materials and Methods

In this study of a prospectively enrolled SSc cohort, a post-hoc analysis on 186 patients (93 PCD-CT, 93 EID-CT), matched by propensity scoring, was performed. Visual ILD assessment by three expert radiologists served as the reference standard. Image quality assessment was performed using Likert-scales by expert radiologists and signal-to-noise ratios (SNR). Quantitative ILD features and extent were extracted using LTA (Imbio, CALIPER-based). Diagnostic accuracy was assessed using ROC-AUC analysis.

Results

LTA-based assessment of ILD on PCD-CT demonstrated a higher AUC for detecting ILD presence (AUC = 0.846) compared to EID-CT (AUC = 0.772). PCD-CT also exhibited superior AUCs in identifying specific ILD features, including ground-glass opacities, reticulation, and honeycombing. However, EID-CT showed higher AUCs than PCD-CT in detecting extensive ILD (>20 % lung involvement; AUC = 0.978 vs. 0.842). Despite significantly lower radiation dose, PCD-CT achieved comparable SNR and superior image quality ratings on Likert-scale.

Conclusion

Both EID-CT and PCD-CT demonstrated acceptable to excellent AUC values, indicating their strong applicability in ILD assessment. Further, LTA using PCD-CT consistently provided excellent AUCs for detecting individual ILD features in SSc, supporting its clinical utility despite being trained on PCD-CT data. PCD-CT’s enhanced image quality and lower radiation dose make it a promising tool for longitudinal ILD assessment. Further multicenter validation is warranted.

目的：比较低剂量光子计数检测器CT （PCD-CT）与传统低剂量能量积分检测器CT （EID-CT）在评估系统性硬化症（SSc）间质性肺病（ILD）中的自动肺结构分析（LTA）的性能。材料和方法：在这项前瞻性纳入SSc队列的研究中，对186例患者（93例PCD-CT， 93例EID-CT）进行了事后分析，并进行了倾向评分。三名放射科专家的视觉ILD评估作为参考标准。图像质量评估由放射科专家使用李克特量表和信噪比（SNR）进行。使用LTA （Imbio, CALIPER-based）提取定量ILD特征和范围。采用ROC-AUC分析评估诊断准确性。结果：与EID-CT （AUC = 0.772）相比，PCD-CT上基于lta的ILD检测AUC （AUC = 0.846）更高。PCD-CT在识别特定ILD特征（包括毛玻璃样混浊、网状和蜂窝状）方面也显示出优越的auc。然而，EID-CT在检测广泛ILD时的AUC高于PCD-CT （bbb20 %肺部受累；AUC = 0.978 vs. 0.842）。尽管辐射剂量明显较低，但PCD-CT在李克特量表上获得了相当的信噪比和优越的图像质量评级。结论：EID-CT和PCD-CT均能获得较好的AUC值，在ILD评估中具有较强的适用性。此外，使用PCD-CT的LTA在检测SSc的单个ILD特征方面始终提供出色的auc，尽管是在PCD-CT数据上进行训练，但仍支持其临床应用。PCD-CT图像质量的提高和较低的辐射剂量使其成为一种有前途的纵向ILD评估工具。进一步的多中心验证是必要的。

{"title":"Automated lung texture analysis for assessing interstitial lung disease in systemic sclerosis: Diagnostic accuracy in photon-counting-detector and conventional energy-integrating-detector CT","authors":"Jasmin Happe , Cosimo Bruni , Lisa Jungblut , Nicholas Landini , Cecilia Strappa , Christian Bluethgen , Muriel Elhai , Rucsandra Dobrota , Carina Mihai , Sinziana Muraru , Anna-Maria Hoffmann-Vold , Anna Rita Larici , Thomas Frauenfelder , Oliver Distler , Jonas Kroschke","doi":"10.1016/j.ejrad.2025.112605","DOIUrl":"10.1016/j.ejrad.2025.112605","url":null,"abstract":"<div><h3>Objectives</h3><div>To evaluate the performance of automated Lung Texture Analysis (LTA) in assessing interstitial lung disease (ILD) in systemic sclerosis (SSc) using low-dose photon-counting detector CT (PCD-CT) compared to conventional low-dose energy-integrating detector CT (EID-CT).</div></div><div><h3>Materials and Methods</h3><div>In this study of a prospectively enrolled SSc cohort, a post-hoc analysis on 186 patients (93 PCD-CT, 93 EID-CT), matched by propensity scoring, was performed. Visual ILD assessment by three expert radiologists served as the reference standard. Image quality assessment was performed using Likert-scales by expert radiologists and signal-to-noise ratios (SNR). Quantitative ILD features and extent were extracted using LTA (Imbio, CALIPER-based). Diagnostic accuracy was assessed using ROC-AUC analysis.</div></div><div><h3>Results</h3><div>LTA-based assessment of ILD on PCD-CT demonstrated a higher AUC for detecting ILD presence (AUC = 0.846) compared to EID-CT (AUC = 0.772). PCD-CT also exhibited superior AUCs in identifying specific ILD features, including ground-glass opacities, reticulation, and honeycombing. However, EID-CT showed higher AUCs than PCD-CT in detecting extensive ILD (>20 % lung involvement; AUC = 0.978 vs. 0.842). Despite significantly lower radiation dose, PCD-CT achieved comparable SNR and superior image quality ratings on Likert-scale.</div></div><div><h3>Conclusion</h3><div>Both EID-CT and PCD-CT demonstrated acceptable to excellent AUC values, indicating their strong applicability in ILD assessment. Further, LTA using PCD-CT consistently provided excellent AUCs for detecting individual ILD features in SSc, supporting its clinical utility despite being trained on PCD-CT data. PCD-CT’s enhanced image quality and lower radiation dose make it a promising tool for longitudinal ILD assessment. Further multicenter validation is warranted.</div></div>","PeriodicalId":12063,"journal":{"name":"European Journal of Radiology","volume":"195 ","pages":"Article 112605"},"PeriodicalIF":3.3,"publicationDate":"2025-12-10","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145767612","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

Safety and performance of a Lipiodol-resistant mixing and injection system in conventional trans-arterial chemoembolization: A post-market clinical follow-up 传统经动脉化疗栓塞中抗脂醇混合和注射系统的安全性和性能：上市后临床随访。

IF 3.3 3区医学 Q1 RADIOLOGY, NUCLEAR MEDICINE & MEDICAL IMAGING

European Journal of Radiology

Pub Date : 2025-12-09 DOI: 10.1016/j.ejrad.2025.112603

Romaric Loffroy , Rafael Duran , Léa Bertin , Rupert Horst Portugaller

Objective

To confirm the safety and performance of Vectorio®, a Lipiodol-resistant mixing and injection system for conventional trans-arterial chemoembolization (cTACE) in clinical practice.

Materials and Methods

This was a prospective, observational, single-arm, multicenter post-market clinical follow-up study conducted between August 2022 and May 2023 (NCT05316077). Adult patients with confirmed hepatocellular carcinoma (HCC) and eligible for cTACE were considered for enrollment. The leakage/breakage rate and the technical performance of Vectorio®, assessed as the ease of use, were evaluated. Adverse events (AE) and device deficiencies were collected up to 24 h after the procedure.

Results

The study included 50 patients (76.0 % male; mean ± SD age: 69.1 ± 9.6 years) with intermediate-stage HCC. One case of leakage, resulting from a user error, was reported. The resulting leakage/breakage rate of Vectorio® was 2.0 % (1/50; 95 %CI: [0.4–––10.5]). The usability of the device was evaluated as ‘easy’ or ‘very easy’ for all the assessed characteristics except for three procedures (6.0 %) during which the disconnection of the syringes from the stopcock, was deemed to be difficult. A total of 7 AEs of mild or moderate intensity were observed in 6 patients. None were related to the device, and none were serious. Study limitations include: small sample size, absence of comparator, and short-follow up period.

Conclusions

This study supports the safety and performance of Vectorio® for the mixing and injection of the Lipiodol / chemotherapeutic agent emulsion for cTACE.

目的：验证抗脂醇混合注射系统Vectorio®用于常规经动脉化疗栓塞（cTACE）临床应用的安全性和性能。材料和方法：这是一项前瞻性、观察性、单臂、多中心上市后临床随访研究，于2022年8月至2023年5月进行（NCT05316077）。已确诊肝细胞癌（HCC）且符合cTACE治疗条件的成年患者被纳入研究对象。评估了泄漏/破损率和Vectorio®的技术性能，评估为易用性。在手术后24小时内收集不良事件（AE）和器械缺陷。结果：该研究纳入了50例中期HCC患者（76.0%为男性，平均±SD年龄：69.1±9.6岁）。据报道，有一个由用户错误导致的泄漏案例。Vectorio®的渗漏/破损率为2.0% （1/50;95% CI:[0.4—10.5]）。该设备的可用性被评估为“容易”或“非常容易”，除了三个程序（6.0%），其中注射器与旋塞的断开被认为是困难的。6例患者共发生轻、中度ae 7例。没有一起与设备有关，也没有一起是严重的。研究的局限性包括：样本量小，缺乏比较者，随访时间短。结论：本研究支持Vectorio®用于混合和注射用于cTACE的脂醇/化疗药物乳剂的安全性和性能。

{"title":"Safety and performance of a Lipiodol-resistant mixing and injection system in conventional trans-arterial chemoembolization: A post-market clinical follow-up","authors":"Romaric Loffroy , Rafael Duran , Léa Bertin , Rupert Horst Portugaller","doi":"10.1016/j.ejrad.2025.112603","DOIUrl":"10.1016/j.ejrad.2025.112603","url":null,"abstract":"<div><h3>Objective</h3><div>To confirm the safety and performance of Vectorio®, a Lipiodol-resistant mixing and injection system for conventional trans-arterial chemoembolization (cTACE) in clinical practice.</div></div><div><h3>Materials and Methods</h3><div>This was a prospective, observational, single-arm, multicenter post-market clinical follow-up study conducted between August 2022 and May 2023 (NCT05316077). Adult patients with confirmed hepatocellular carcinoma (HCC) and eligible for cTACE were considered for enrollment. The leakage/breakage rate and the technical performance of Vectorio®, assessed as the ease of use, were evaluated. Adverse events (AE) and device deficiencies were collected up to 24 h after the procedure.</div></div><div><h3>Results</h3><div>The study included 50 patients (76.0 % male; mean ± SD age: 69.1 ± 9.6 years) with intermediate-stage HCC. One case of leakage, resulting from a user error, was reported. The resulting leakage/breakage rate of Vectorio® was 2.0 % (1/50; 95 %CI: [0.4–––10.5]). The usability of the device was evaluated as ‘easy’ or ‘very easy’ for all the assessed characteristics except for three procedures (6.0 %) during which the disconnection of the syringes from the stopcock, was deemed to be difficult. A total of 7 AEs of mild or moderate intensity were observed in 6 patients. None were related to the device, and none were serious. Study limitations include: small sample size, absence of comparator, and short-follow up period.</div></div><div><h3>Conclusions</h3><div>This study supports the safety and performance of Vectorio® for the mixing and injection of the Lipiodol/chemotherapeutic agent emulsion for cTACE.</div></div>","PeriodicalId":12063,"journal":{"name":"European Journal of Radiology","volume":"195 ","pages":"Article 112603"},"PeriodicalIF":3.3,"publicationDate":"2025-12-09","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145762427","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0