Academic Radiology最新文献_第7页

Deep Learning Based Multiomics Model for Risk Stratification of Postoperative Distant Metastasis in Colorectal Cancer 基于深度学习的结直肠癌术后远处转移风险分层多组学模型。

IF 3.9 2区医学 Q1 RADIOLOGY, NUCLEAR MEDICINE & MEDICAL IMAGING

Academic Radiology

Pub Date : 2026-03-01 Epub Date: 2025-09-04 DOI: 10.1016/j.acra.2025.08.040

Xiuzhen Yao , Xiaoyu Han , Danjiang Huang , Yongfei Zheng , Shuitang Deng , Xiaoxiang Ning , Li Yuan , Weiqun Ao

Rationale and Objectives

To develop deep learning-based multiomics models for predicting postoperative distant metastasis (DM) and evaluating survival prognosis in colorectal cancer (CRC) patients.

Materials and Methods

This retrospective study included 521 CRC patients who underwent curative surgery at two centers. Preoperative CT and postoperative hematoxylin-eosin (HE) stained slides were collected. A total of 381 patients from Center 1 were split (7:3) into training and internal validation sets; 140 patients from Center 2 formed the independent external validation set. Patients were grouped based on DM status during follow-up. Radiological and pathological models were constructed using independent imaging and pathological predictors. Deep features were extracted with a ResNet-101 backbone to build deep learning radiomics (DLRS) and deep learning pathomics (DLPS) models. Two integrated models were developed: Nomogram 1 (radiological + DLRS) and Nomogram 2 (pathological + DLPS).

Results

CT- reported T (cT) stage (OR = 2.00, P = 0.006) and CT-reported N (cN) stage (OR = 1.63, P = 0.023) were identified as independent radiologic predictors for building the radiological model; pN stage (OR = 1.91, P = 0.003) and perineural invasion (OR = 2.07, P = 0.030) were identified as pathological predictors for building the pathological model. DLRS and DLPS incorporated 28 and 30 deep features, respectively. In the training set, area under the curve (AUC) for radiological, pathological, DLRS, DLPS, Nomogram 1, and Nomogram 2 models were 0.657, 0.687, 0.931, 0.914, 0.938, and 0.930. DeLong’s test showed DLRS, DLPS, and both nomograms significantly outperformed conventional models (P<.05). Kaplan–Meier analysis confirmed effective 3-year disease-free survival (DFS) stratification by the nomograms.

Conclusion

Deep learning-based multiomics models provided high accuracy for postoperative DM prediction. Nomogram models enabled reliable DFS risk stratification in CRC patients.

目的：建立基于深度学习的多组学模型，用于预测结直肠癌（CRC）患者术后远处转移（DM）和评估生存预后。材料和方法：本回顾性研究包括521例在两个中心接受治疗性手术的结直肠癌患者。收集术前CT和术后苏木精-伊红（HE）染色玻片。中心1共有381名患者（7:3）被分为训练组和内部验证组；来自第二中心的140例患者形成了独立的外部验证集。根据随访期间的糖尿病状态对患者进行分组。使用独立的影像学和病理预测因子构建影像学和病理模型。利用ResNet-101主干提取深度特征，构建深度学习放射组学（DLRS）和深度学习病理学（DLPS）模型。建立了两种综合模型：Nomogram 1（放射学+ DLRS）和Nomogram 2（病理学+ DLPS）。结果：CT报告的T （CT）分期（OR=2.00， P=0.006）和CT报告的N （cN）分期（OR=1.63， P=0.023）可作为建立放射学模型的独立预测因子；pN分期（OR=1.91， P=0.003）和周围神经浸润（OR=2.07， P=0.030）是建立病理模型的病理预测因子。DLRS和DLPS分别包含28个和30个深度特征。在训练集中，放射学、病理、DLRS、DLPS、Nomogram 1、Nomogram 2模型的曲线下面积（area under the curve， AUC）分别为0.657、0.687、0.931、0.914、0.938、0.930。DeLong的测试显示，DLRS、DLPS和两种模态图都明显优于传统模型(结论：基于深度学习的多组学模型对术后DM的预测具有较高的准确性。Nomogram模型能够在CRC患者中实现可靠的DFS风险分层。

{"title":"Deep Learning Based Multiomics Model for Risk Stratification of Postoperative Distant Metastasis in Colorectal Cancer","authors":"Xiuzhen Yao , Xiaoyu Han , Danjiang Huang , Yongfei Zheng , Shuitang Deng , Xiaoxiang Ning , Li Yuan , Weiqun Ao","doi":"10.1016/j.acra.2025.08.040","DOIUrl":"10.1016/j.acra.2025.08.040","url":null,"abstract":"<div><h3>Rationale and Objectives</h3><div>To develop deep learning-based multiomics models for predicting postoperative distant metastasis (DM) and evaluating survival prognosis in colorectal cancer (CRC) patients.</div></div><div><h3>Materials and Methods</h3><div>This retrospective study included 521 CRC patients who underwent curative surgery at two centers. Preoperative CT and postoperative hematoxylin-eosin (HE) stained slides were collected. A total of 381 patients from Center 1 were split (7:3) into training and internal validation sets; 140 patients from Center 2 formed the independent external validation set. Patients were grouped based on DM status during follow-up. Radiological and pathological models were constructed using independent imaging and pathological predictors. Deep features were extracted with a ResNet-101 backbone to build deep learning radiomics (DLRS) and deep learning pathomics (DLPS) models. Two integrated models were developed: Nomogram 1 (radiological + DLRS) and Nomogram 2 (pathological + DLPS).</div></div><div><h3>Results</h3><div>CT- reported T (cT) stage (OR=2.00, P=0.006) and CT-reported N (cN) stage (OR=1.63, P=0.023) were identified as independent radiologic predictors for building the radiological model; pN stage (OR=1.91, P=0.003) and perineural invasion (OR=2.07, P=0.030) were identified as pathological predictors for building the pathological model. DLRS and DLPS incorporated 28 and 30 deep features, respectively. In the training set, area under the curve (AUC) for radiological, pathological, DLRS, DLPS, Nomogram 1, and Nomogram 2 models were 0.657, 0.687, 0.931, 0.914, 0.938, and 0.930. DeLong’s test showed DLRS, DLPS, and both nomograms significantly outperformed conventional models (P<.05). Kaplan–Meier analysis confirmed effective 3-year disease-free survival (DFS) stratification by the nomograms.</div></div><div><h3>Conclusion</h3><div>Deep learning-based multiomics models provided high accuracy for postoperative DM prediction. Nomogram models enabled reliable DFS risk stratification in CRC patients.</div></div>","PeriodicalId":50928,"journal":{"name":"Academic Radiology","volume":"33 3","pages":"Pages 858-871"},"PeriodicalIF":3.9,"publicationDate":"2026-03-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145006768","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"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

Leading with C.A.R.E: A Framework to Foster Belonging and Well-Being in Radiology 以C.A.R.E为先导：培养放射学归属感和幸福感的框架。

IF 3.9 2区医学 Q1 RADIOLOGY, NUCLEAR MEDICINE & MEDICAL IMAGING

Academic Radiology

Pub Date : 2026-03-01 Epub Date: 2025-11-18 DOI: 10.1016/j.acra.2025.10.046

Lily M. Belfi M.D. FACR , Sarah Averill M.D. , Constantine Burgen M.D. , Reni Butler M.D. , Michele Retrouvey M.D. , Lori A. Deitte M.D. FACR FAUR

The field of radiology is experiencing profound shifts driven by technological advancements, evolving workplace structures, and changing generational expectations, contributing to increased rates of burnout, professional isolation, and diminished purpose among radiologists. To address these challenges and promote workforce sustainability, we propose the C.A.R.E. framework—encompassing Community, Advocacy, Recognition, and Empowerment—as a comprehensive, human-centered model to foster belonging and well-being in radiology. This framework emphasizes intentional strategies to build meaningful professional relationships in hybrid and remote environments; offers pathways to engage in advocacy at personal, interpersonal, and systemic levels; underscores the importance of recognition and appreciation in cultivating engagement and combating burnout; and highlights mentorship, coaching, and sponsorship as pivotal tools for professional and personal growth. By implementing the C.A.R.E. framework, radiology departments and organizations can create inclusive, supportive environments that enhance individual fulfillment, professional resilience, and organizational success. Future initiatives should focus on operationalizing and evaluating these practices to ensure sustained improvements in workforce well-being and patient care outcomes.

在技术进步、工作场所结构演变和世代期望变化的推动下，放射学领域正在经历深刻的变革，导致放射科医生职业倦怠率上升、职业孤立和目标降低。为了应对这些挑战并促进劳动力的可持续性，我们提出了C.A.R.E.框架——包括社区、倡导、认可和赋权——作为一个全面的、以人为本的模型，以促进放射学的归属感和福祉。该框架强调在混合和远程环境中建立有意义的专业关系的有意策略；提供在个人、人际和系统层面进行宣传的途径；强调认可和赞赏在培养敬业精神和对抗倦怠方面的重要性；并强调指导、指导和赞助是职业和个人成长的关键工具。通过实施C.A.R.E.框架，放射科和组织可以创造包容的、支持性的环境，从而提高个人的成就感、专业的弹性和组织的成功。未来的举措应侧重于实施和评估这些做法，以确保持续改善劳动力福利和患者护理结果。

{"title":"Leading with C.A.R.E: A Framework to Foster Belonging and Well-Being in Radiology","authors":"Lily M. Belfi M.D. FACR , Sarah Averill M.D. , Constantine Burgen M.D. , Reni Butler M.D. , Michele Retrouvey M.D. , Lori A. Deitte M.D. FACR FAUR","doi":"10.1016/j.acra.2025.10.046","DOIUrl":"10.1016/j.acra.2025.10.046","url":null,"abstract":"<div><div>The field of radiology is experiencing profound shifts driven by technological advancements, evolving workplace structures, and changing generational expectations, contributing to increased rates of burnout, professional isolation, and diminished purpose among radiologists. To address these challenges and promote workforce sustainability, we propose the C.A.R.E. framework—encompassing Community, Advocacy, Recognition, and Empowerment—as a comprehensive, human-centered model to foster belonging and well-being in radiology. This framework emphasizes intentional strategies to build meaningful professional relationships in hybrid and remote environments; offers pathways to engage in advocacy at personal, interpersonal, and systemic levels; underscores the importance of recognition and appreciation in cultivating engagement and combating burnout; and highlights mentorship, coaching, and sponsorship as pivotal tools for professional and personal growth. By implementing the C.A.R.E. framework, radiology departments and organizations can create inclusive, supportive environments that enhance individual fulfillment, professional resilience, and organizational success. Future initiatives should focus on operationalizing and evaluating these practices to ensure sustained improvements in workforce well-being and patient care outcomes.</div></div>","PeriodicalId":50928,"journal":{"name":"Academic Radiology","volume":"33 3","pages":"Pages 718-725"},"PeriodicalIF":3.9,"publicationDate":"2026-03-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145558193","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"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

Deep Learning-Based Differentiation of DCIS and IDC from Mammographic Microcalcifications 基于深度学习的DCIS和IDC与乳腺微钙化的鉴别。

IF 3.9 2区医学 Q1 RADIOLOGY, NUCLEAR MEDICINE & MEDICAL IMAGING

Academic Radiology

Pub Date : 2026-03-01 Epub Date: 2025-12-22 DOI: 10.1016/j.acra.2025.12.002

Deniz Esin Tekcan Sanli MD , Ahmet Necati Sanli

引用次数: 0

Pixel-level Radiomics and Deep Learning for Predicting Ki-67 Expression in Breast Cancer Based on Dual-modal Ultrasound Images 基于双模超声图像的像素级放射组学和深度学习预测乳腺癌中Ki-67的表达。

IF 3.9 2区医学 Q1 RADIOLOGY, NUCLEAR MEDICINE & MEDICAL IMAGING

Academic Radiology

Pub Date : 2026-03-01 Epub Date: 2026-01-14 DOI: 10.1016/j.acra.2025.12.047

Wei Wei , Fei Xia , Di Zhang , Wang Zhou , Xinjin Wang , Yu Gao , Wenwu Lu , Huijun Feng , Chaoxue Zhang

Rationale and Objectives

This study aimed to develop a deep learning model using a novel pixel-level radiomics approach based on two-dimensional (2D) and strain elastography (SE) ultrasound images to predict Ki-67 expression in breast cancer (BC).

Methods

This multicenter study included 1031 BC patients, who were divided into training (n = 616), internal validation (n = 265), and external test (n = 150) cohorts. An additional 63 patients were prospectively enrolled for further validation. The deep learning model, termed Vision-Mamba, predicts Ki67 expression by integrating ultrasound (2D and SE) images with pixel-level radiomics feature maps (RFMs). A combined model was subsequently constructed by incorporating independent clinical predictors. Model performance was assessed using receiver operating characteristic (ROC) curves, calibration curves, and decision curve analysis (DCA). SHapley Additive exPlanations (SHAP) were applied to enhance interpretability.

Results

We developed a Vision-Mamba-US-RFMs-Clinical (V-MURC) model that integrates ultrasound images, RFMs, and clinical data for accurate prediction of Ki-67 expression in BC. The area under the ROC curve (AUC) values for the internal validation, external test, and prospective validation cohorts were 0.954 (95% CI, 0.929 - 0.975), 0.941 (95% CI, 0.903 - 0.975), and 0.945 (95% CI, 0.883 - 0.989), respectively, demonstrating excellent discrimination and calibration. Compared with individual models, the V-MURC model achieved significantly superior performance across all datasets (Delong test, P < 0.05). Calibration curves and DCA further supported its clinical applicability. SHAP analysis provided visual interpretability of the model's decision-making process.

Conclusion

The V-MURC model based on pixel-level RFMs can accurately predict Ki-67 expression in BC and may serve as a valuable tool for individualized treatment decision-making in clinical practice.

基本原理和目的：本研究旨在利用基于二维（2D）和应变弹性成像（SE）超声图像的新型像素级放射组学方法开发一种深度学习模型，以预测乳腺癌（BC）中Ki-67的表达。方法：本多中心研究纳入1031例BC患者，分为训练组（n = 616）、内部验证组（n = 265）和外部测试组（n = 150）。另外63名患者被纳入前瞻性研究以进一步验证。该深度学习模型被称为Vision-Mamba，通过整合超声（2D和SE）图像和像素级放射组学特征图（rfm）来预测Ki67的表达。随后通过合并独立的临床预测因子构建了一个联合模型。采用受试者工作特征（ROC）曲线、校准曲线和决策曲线分析（DCA）评估模型的性能。采用SHapley加性解释（SHAP）提高可解释性。结果：我们建立了一个视觉-曼巴-美国- rfm -临床（V-MURC）模型，该模型整合了超声图像、rfm和临床数据，用于准确预测BC中Ki-67的表达。内部验证队列、外部验证队列和前瞻性验证队列的ROC曲线下面积（AUC）值分别为0.954 （95% CI, 0.929 ~ 0.975）、0.941 （95% CI, 0.903 ~ 0.975）和0.945 (95% CI, 0.883 ~ 0.989)，具有良好的判别和校准能力。与单个模型相比，V-MURC模型在所有数据集上的性能都显著优于单个模型（Delong检验，P < 0.05）。校准曲线和DCA进一步支持了其临床适用性。SHAP分析提供了模型决策过程的可视化可解释性。结论：基于像素级rmrm的V-MURC模型可以准确预测BC中Ki-67的表达，可作为临床个体化治疗决策的重要工具。

{"title":"Pixel-level Radiomics and Deep Learning for Predicting Ki-67 Expression in Breast Cancer Based on Dual-modal Ultrasound Images","authors":"Wei Wei , Fei Xia , Di Zhang , Wang Zhou , Xinjin Wang , Yu Gao , Wenwu Lu , Huijun Feng , Chaoxue Zhang","doi":"10.1016/j.acra.2025.12.047","DOIUrl":"10.1016/j.acra.2025.12.047","url":null,"abstract":"<div><h3>Rationale and Objectives</h3><div>This study aimed to develop a deep learning model using a novel pixel-level radiomics approach based on two-dimensional (2D) and strain elastography (SE) ultrasound images to predict Ki-67 expression in breast cancer (BC).</div></div><div><h3>Methods</h3><div>This multicenter study included 1031 BC patients, who were divided into training (n = 616), internal validation (n = 265), and external test (n = 150) cohorts. An additional 63 patients were prospectively enrolled for further validation. The deep learning model, termed Vision-Mamba, predicts Ki67 expression by integrating ultrasound (2D and SE) images with pixel-level radiomics feature maps (RFMs). A combined model was subsequently constructed by incorporating independent clinical predictors. Model performance was assessed using receiver operating characteristic (ROC) curves, calibration curves, and decision curve analysis (DCA). SHapley Additive exPlanations (SHAP) were applied to enhance interpretability.</div></div><div><h3>Results</h3><div>We developed a Vision-Mamba-US-RFMs-Clinical (V-MURC) model that integrates ultrasound images, RFMs, and clinical data for accurate prediction of Ki-67 expression in BC. The area under the ROC curve (AUC) values for the internal validation, external test, and prospective validation cohorts were 0.954 (95% CI, 0.929 - 0.975), 0.941 (95% CI, 0.903 - 0.975), and 0.945 (95% CI, 0.883 - 0.989), respectively, demonstrating excellent discrimination and calibration. Compared with individual models, the V-MURC model achieved significantly superior performance across all datasets (Delong test, <em>P</em> < 0.05). Calibration curves and DCA further supported its clinical applicability. SHAP analysis provided visual interpretability of the model's decision-making process.</div></div><div><h3>Conclusion</h3><div>The V-MURC model based on pixel-level RFMs can accurately predict Ki-67 expression in BC and may serve as a valuable tool for individualized treatment decision-making in clinical practice.</div></div>","PeriodicalId":50928,"journal":{"name":"Academic Radiology","volume":"33 3","pages":"Pages 900-917"},"PeriodicalIF":3.9,"publicationDate":"2026-03-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145991799","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"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 Cardiac MRI Planning: From Localizers to Cine Images 自动心脏MRI计划：从定位器到电影图像。

IF 3.9 2区医学 Q1 RADIOLOGY, NUCLEAR MEDICINE & MEDICAL IMAGING

Academic Radiology

Pub Date : 2026-03-01 Epub Date: 2026-03-02 DOI: 10.1016/j.acra.2026.01.026

Soheil Kooraki MD, Arash Bedayat MD

引用次数: 0

From Artificial Intelligence to Robotics: How to Navigate Technological Innovation in Radiology With the Gartner Hype Cycle 从人工智能到机器人技术：如何利用高德纳技术成熟度周期引导放射学技术创新。

IF 3.9 2区医学 Q1 RADIOLOGY, NUCLEAR MEDICINE & MEDICAL IMAGING

Academic Radiology

Pub Date : 2026-03-01 Epub Date: 2026-03-02 DOI: 10.1016/j.acra.2025.11.018

Siddhant Dogra M.D. , Safiullah Rifai , Mohiuddin Hadi M.D. , Tarek Hanna M.D. , Anna Rozenshtein M.D. , Michele Retrouvey M.D. , Heng Huang PhD , Florence X. Doo M.D, M.A

Radiology is rapidly evolving across both technology and practice, making it difficult to distinguish innovations with genuine lasting impact from those that may fade after initial enthusiasm. In this narrative review, we use an expert-informed framework and a targeted appraisal of the literature to map key advances in radiology onto the Gartner hype cycle, across three key areas: software and algorithms, advanced imaging tools and techniques, and clinical practice paradigms. We provide strategic considerations for radiologists, trainees, and leaders, including potential future implications for the specialty. Together, radiologists, trainees, and leaders play distinct and complementary roles in guiding the adoption of emerging technologies, from deepening clinical-technical expertise to developing systemic governance and infrastructure that supports safe, evidence-driven implementation.

As part 6 of the Radiology Research Alliance (RRA) review series on emerging technologies in collaboration with the University of Maryland Institute for Health Computing (UM-IHC) and the Medical Intelligent Imaging (UM2ii) Center, this paper provides a critical perspective on how radiologists, trainees, and leaders can navigate the hype cycle to identify meaningful innovations and guide strategic adoption in practice.

放射学在技术和实践方面都在迅速发展，这使得很难区分具有真正持久影响的创新与那些可能在最初的热情之后消退的创新。在这篇叙述性回顾中，我们使用专家知情的框架和对文献的有针对性的评估，将放射学的关键进展映射到Gartner炒作周期中，涉及三个关键领域：软件和算法，先进的成像工具和技术，以及临床实践范例。我们为放射科医生、培训生和领导者提供战略考虑，包括该专业未来的潜在影响。从深化临床技术专业知识到发展支持安全、循证实施的系统治理和基础设施，放射科医生、受训人员和领导者在指导采用新兴技术方面发挥着独特而互补的作用。作为放射学研究联盟（RRA）与马里兰大学健康计算研究所（UM-IHC）和医学智能成像（UM2ii）中心合作的新兴技术回顾系列的第6部分，本文提供了一个重要的视角，说明放射科医生、学员和领导者如何驾驭炒作周期，以确定有意义的创新并指导实践中的战略采用。

{"title":"From Artificial Intelligence to Robotics: How to Navigate Technological Innovation in Radiology With the Gartner Hype Cycle","authors":"Siddhant Dogra M.D. , Safiullah Rifai , Mohiuddin Hadi M.D. , Tarek Hanna M.D. , Anna Rozenshtein M.D. , Michele Retrouvey M.D. , Heng Huang PhD , Florence X. Doo M.D, M.A","doi":"10.1016/j.acra.2025.11.018","DOIUrl":"10.1016/j.acra.2025.11.018","url":null,"abstract":"<div><div>Radiology is rapidly evolving across both technology and practice, making it difficult to distinguish innovations with genuine lasting impact from those that may fade after initial enthusiasm. In this narrative review, we use an expert-informed framework and a targeted appraisal of the literature to map key advances in radiology onto the Gartner hype cycle, across three key areas: software and algorithms, advanced imaging tools and techniques, and clinical practice paradigms. We provide strategic considerations for radiologists, trainees, and leaders, including potential future implications for the specialty. Together, radiologists, trainees, and leaders play distinct and complementary roles in guiding the adoption of emerging technologies, from deepening clinical-technical expertise to developing systemic governance and infrastructure that supports safe, evidence-driven implementation.</div><div>As part 6 of the Radiology Research Alliance (RRA) review series on emerging technologies in collaboration with the University of Maryland Institute for Health Computing (UM-IHC) and the Medical Intelligent Imaging (UM2ii) Center, this paper provides a critical perspective on how radiologists, trainees, and leaders can navigate the hype cycle to identify meaningful innovations and guide strategic adoption in practice.</div></div>","PeriodicalId":50928,"journal":{"name":"Academic Radiology","volume":"33 3","pages":"Pages 662-679"},"PeriodicalIF":3.9,"publicationDate":"2026-03-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"147357290","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"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

Preoperative Prediction of Occult Lymph Node Metastasis in Clinically Node-Negative Early-Stage Lung Adenocarcinoma: A Multicenter Machine Learning Study 临床淋巴结阴性早期肺腺癌隐匿淋巴结转移的术前预测：一项多中心机器学习研究。

IF 3.9 2区医学 Q1 RADIOLOGY, NUCLEAR MEDICINE & MEDICAL IMAGING

Academic Radiology

Pub Date : 2026-03-01 Epub Date: 2025-11-15 DOI: 10.1016/j.acra.2025.10.058

Fengnian Zhao MD , Yunqing Zhao MD , Zhaoxiang Ye MD, PhD , Haoran Sun MD , Yanbo Li MD , Guiming Zhou MD

Background

Occult lymph node metastasis (OLNM) in cN0 early-stage lung adenocarcinoma (LUAD) leads to pathological understaging and suboptimal surgical management. Current prediction tools exhibit limited robustness. This study aimed to develop and validate a machine learning model that integrates CT semantic and radiomic features for the preoperative prediction of OLNM.

Methods

In this retrospective multicenter study, an interpretable machine learning model was developed. A cohort of 752 patients (training: n = 495; internal validation: n = 124; external validation: n = 133) underwent rigorous feature processing: ComBat harmonization for scanner variability, PCA dimensionality reduction, and LASSO regression for feature selection. Seven classifiers were optimized using SMOTE-balanced training data.

Results

The XGBoost model demonstrated robust performance, achieving an ROCAUC of 0.814 (0.666–0.925) and a PR-AUC of 0.502 (0.328–0.803) in the internal validation cohort. It maintained strong generalizability in the external validation cohort, with an ROC-AUC of 0.826 (0.746–0.897) and a PR-AUC of 0.486 (0.331–0.714). The model was well-calibrated (Brier scores: 0.135 and 0.128, respectively). Risk stratification identified five clinically actionable tiers: "very-low-risk" to "very-high-risk" patients exhibited monotonically increasing rates of OLNM (internal validation: 3.7% to 40.0%; external validation: 4.1-fold increase in metastasis). SHAP analysis identified consolidation level, radiomics-derived Rad-score, and lobulation as the top three predictors.

Conclusion

This validated model integrates physician-interpreted semantics with data-driven radiomics, providing a non-invasive tool for personalized surgical planning. It enables tailored lymph node dissection strategies while enhancing accessibility in resource-limited settings.

Data availability

The datasets used and/or analyzed during the current study are available from the corresponding author on reasonable request.

背景：cN0早期肺腺癌（LUAD）的隐性淋巴结转移（OLNM）导致病理性分期不足和手术治疗不理想。目前的预测工具表现出有限的稳健性。本研究旨在开发和验证一种整合CT语义和放射学特征的机器学习模型，用于OLNM的术前预测。方法：在这项回顾性的多中心研究中，开发了一个可解释的机器学习模型。752名患者（训练：n = 495，内部验证：n = 124，外部验证：n = 133）进行了严格的特征处理：针对扫描仪变异性的战斗协调，PCA降维，以及用于特征选择的LASSO回归。使用SMOTE-balanced训练数据对7个分类器进行了优化。结果：XGBoost模型表现出稳健的性能，在内部验证队列中ROCAUC为0.814 (0.666-0.925)，PR-AUC为0.502（0.28 -0.803）。它在外部验证队列中保持了很强的通用性，ROC-AUC为0.826 (0.746-0.897)，PR-AUC为0.486（0.331-0.714）。该模型校正良好（Brier评分分别为0.135和0.128）。风险分层确定了五个临床可操作的级别：“极低风险”至“极高风险”患者的OLNM发生率单调增加（内部验证：3.7%至40.0%；外部验证：转移增加4.1倍）。SHAP分析确定了巩固水平、放射组学衍生的rad评分和分叶化是最重要的三个预测因素。结论：该验证模型将医生解释的语义与数据驱动的放射组学相结合，为个性化手术计划提供了一种非侵入性工具。它使量身定制的淋巴结清扫策略，同时提高在资源有限的设置可及性。数据可用性：当前研究中使用和/或分析的数据集可根据通讯作者的合理要求从其处获取。

{"title":"Preoperative Prediction of Occult Lymph Node Metastasis in Clinically Node-Negative Early-Stage Lung Adenocarcinoma: A Multicenter Machine Learning Study","authors":"Fengnian Zhao MD , Yunqing Zhao MD , Zhaoxiang Ye MD, PhD , Haoran Sun MD , Yanbo Li MD , Guiming Zhou MD","doi":"10.1016/j.acra.2025.10.058","DOIUrl":"10.1016/j.acra.2025.10.058","url":null,"abstract":"<div><h3>Background</h3><div>Occult lymph node metastasis (OLNM) in cN0 early-stage lung adenocarcinoma (LUAD) leads to pathological understaging and suboptimal surgical management. Current prediction tools exhibit limited robustness. This study aimed to develop and validate a machine learning model that integrates CT semantic and radiomic features for the preoperative prediction of OLNM.</div></div><div><h3>Methods</h3><div>In this retrospective multicenter study, an interpretable machine learning model was developed. A cohort of 752 patients (training: <em>n</em> = 495; internal validation: <em>n</em> = 124; external validation: <em>n</em> = 133) underwent rigorous feature processing: ComBat harmonization for scanner variability, PCA dimensionality reduction, and LASSO regression for feature selection. Seven classifiers were optimized using SMOTE-balanced training data.</div></div><div><h3>Results</h3><div>The XGBoost model demonstrated robust performance, achieving an ROCAUC of 0.814 (0.666–0.925) and a PR-AUC of 0.502 (0.328–0.803) in the internal validation cohort. It maintained strong generalizability in the external validation cohort, with an ROC-AUC of 0.826 (0.746–0.897) and a PR-AUC of 0.486 (0.331–0.714). The model was well-calibrated (Brier scores: 0.135 and 0.128, respectively). Risk stratification identified five clinically actionable tiers: \"very-low-risk\" to \"very-high-risk\" patients exhibited monotonically increasing rates of OLNM (internal validation: 3.7% to 40.0%; external validation: 4.1-fold increase in metastasis). SHAP analysis identified consolidation level, radiomics-derived Rad-score, and lobulation as the top three predictors.</div></div><div><h3>Conclusion</h3><div>This validated model integrates physician-interpreted semantics with data-driven radiomics, providing a non-invasive tool for personalized surgical planning. It enables tailored lymph node dissection strategies while enhancing accessibility in resource-limited settings.</div></div><div><h3>Data availability</h3><div>The datasets used and/or analyzed during the current study are available from the corresponding author on reasonable request.</div></div>","PeriodicalId":50928,"journal":{"name":"Academic Radiology","volume":"33 3","pages":"Pages 1152-1166"},"PeriodicalIF":3.9,"publicationDate":"2026-03-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145530767","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"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

Artificial Intelligence Education in Radiology Training: A Systematic Review of Effectiveness, Barriers, and Future Directions 放射学培训中的人工智能教育：有效性、障碍和未来方向的系统回顾。

IF 3.9 2区医学 Q1 RADIOLOGY, NUCLEAR MEDICINE & MEDICAL IMAGING

Academic Radiology

Pub Date : 2026-03-01 Epub Date: 2025-11-15 DOI: 10.1016/j.acra.2025.10.049

Pedram Keshavarz , Zahra Mohammadigoldar , Arash Bedayat , Steven S. Raman , Ryan Tai

Rationale and Objectives

The purpose of this systematic review study was to characterize the current landscape of various artificial intelligence (AI) education in radiology, summarizing existing curricula, outcomes, challenges, and future directions for effective integration into residency training.

Materials and Methods

A comprehensive search of PubMed, Web of Science, Embase, and Google Scholar identified relevant published studies up to June 19, 2025.

Results

Of the 2646 studies screened, 14 studies evaluated the performance of AI-based training programs for radiology trainees; among these, 92.9% (13/14) reported improvements in trainees’ performance, including better diagnostic precision and interpretation (57.2%, 8/14), greater trainee confidence (57.2%, 8/14), hands-on experience with AI platforms (85.7%, 12/14), increased AI knowledge (85.7%, 12/14), engagement with AI-based case learning (35.7%, 5/14), understanding of AI ethics and bias (7.1%, 1/14), and acceptance of AI-assisted learning (78.6%, 11/14), whereas one study (7.1%, 1/14) found no significant benefit. Performance evaluation metrics varied across studies, with 35.7% (5/14) reporting a higher median of sensitivity, specificity, and accuracy (72%, 80%, and 81.3%) after AI training compared to before AI training (62.2%, 78.9%, and 76.5%, respectively), and 28.6% (4/14) showing improved AI knowledge scores. Hands-on simulations and didactic lectures were the most common AI training formats (78.6% and 71.4%). Risks and concerns included over-reliance on AI, limited exposure to complex or rare cases, and a lack of feedback. Recommendations highlighted the need for AI-faculty teaching, broader content coverage, and standardized multi-center AI-training programs to facilitate wider adoption.

Conclusion

92.9% of studies showed that AI-based training can enhance radiology trainees’ knowledge, interpretive skills, or diagnostic performance, especially for junior trainees; however, its safe adoption requires standardized curricula with diverse cases, mentorship, workflow integration, and robust evaluation, with larger studies needed to confirm generalizability.

基本原理和目标：本系统综述研究的目的是描述放射学中各种人工智能（AI）教育的现状，总结现有课程、结果、挑战和有效整合到住院医师培训中的未来方向。材料和方法：综合检索PubMed， Web of Science， Embase和谷歌Scholar，确定了截至2025年6月19日的相关已发表研究。结果：在筛选的2646项研究中，14项研究评估了基于人工智能的放射学受训人员培训计划的表现；其中，92.9%（13/14）报告了受训者绩效的改善，包括更好的诊断精度和解释（57.2%,8/14），更大的受训者信心（57.2%,8/14），人工智能平台的实践经验（85.7%,12/14），人工智能知识的增加（85.7%,12/14），参与基于人工智能的案例学习（35.7%,5/14），对人工智能伦理和偏见的理解（7.1%,1/14），以及接受人工智能辅助学习（78.6%,11/14）。1/14)没有发现显著的益处。各研究的绩效评估指标各不相同，35.7%（5/14）报告人工智能训练后的灵敏度、特异性和准确性中位数（72%、80%和81.3%）高于人工智能训练前（分别为62.2%、78.9%和76.5%），28.6%（4/14）显示人工智能知识得分有所提高。实践模拟和教学讲座是最常见的人工智能训练形式（78.6%和71.4%）。风险和担忧包括过度依赖人工智能，接触复杂或罕见病例的机会有限，以及缺乏反馈。建议强调了人工智能教师教学、更广泛的内容覆盖和标准化的多中心人工智能培训计划的必要性，以促进更广泛的采用。结论：92.9%的研究表明，基于人工智能的培训可以提高放射科实习生的知识、解释技能或诊断表现，尤其是对初级实习生；然而，它的安全采用需要标准化的课程，包括不同的案例、指导、工作流集成和可靠的评估，需要更大的研究来确认其普遍性。

{"title":"Artificial Intelligence Education in Radiology Training: A Systematic Review of Effectiveness, Barriers, and Future Directions","authors":"Pedram Keshavarz , Zahra Mohammadigoldar , Arash Bedayat , Steven S. Raman , Ryan Tai","doi":"10.1016/j.acra.2025.10.049","DOIUrl":"10.1016/j.acra.2025.10.049","url":null,"abstract":"<div><h3>Rationale and Objectives</h3><div>The purpose of this systematic review study was to characterize the current landscape of various artificial intelligence (AI) education in radiology, summarizing existing curricula, outcomes, challenges, and future directions for effective integration into residency training.</div></div><div><h3>Materials and Methods</h3><div>A comprehensive search of PubMed, Web of Science, Embase, and Google Scholar identified relevant published studies up to June 19, 2025.</div></div><div><h3>Results</h3><div>Of the 2646 studies screened, 14 studies evaluated the performance of AI-based training programs for radiology trainees; among these, 92.9% (13/14) reported improvements in trainees’ performance, including better diagnostic precision and interpretation (57.2%, 8/14), greater trainee confidence (57.2%, 8/14), hands-on experience with AI platforms (85.7%, 12/14), increased AI knowledge (85.7%, 12/14), engagement with AI-based case learning (35.7%, 5/14), understanding of AI ethics and bias (7.1%, 1/14), and acceptance of AI-assisted learning (78.6%, 11/14), whereas one study (7.1%, 1/14) found no significant benefit. Performance evaluation metrics varied across studies, with 35.7% (5/14) reporting a higher median of sensitivity, specificity, and accuracy (72%, 80%, and 81.3%) after AI training compared to before AI training (62.2%, 78.9%, and 76.5%, respectively), and 28.6% (4/14) showing improved AI knowledge scores. Hands-on simulations and didactic lectures were the most common AI training formats (78.6% and 71.4%). Risks and concerns included over-reliance on AI, limited exposure to complex or rare cases, and a lack of feedback. Recommendations highlighted the need for AI-faculty teaching, broader content coverage, and standardized multi-center AI-training programs to facilitate wider adoption.</div></div><div><h3>Conclusion</h3><div>92.9% of studies showed that AI-based training can enhance radiology trainees’ knowledge, interpretive skills, or diagnostic performance, especially for junior trainees; however, its safe adoption requires standardized curricula with diverse cases, mentorship, workflow integration, and robust evaluation, with larger studies needed to confirm generalizability.</div></div>","PeriodicalId":50928,"journal":{"name":"Academic Radiology","volume":"33 3","pages":"Pages 695-706"},"PeriodicalIF":3.9,"publicationDate":"2026-03-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145534992","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"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

Deep Learning-Based Cardiac MRI Planning from Localizers to Cine Views Using Landmark Detection 基于深度学习的心脏MRI规划从定位器到使用地标检测的电影视图。

IF 3.9 2区医学 Q1 RADIOLOGY, NUCLEAR MEDICINE & MEDICAL IMAGING

Academic Radiology

Pub Date : 2026-03-01 Epub Date: 2025-12-05 DOI: 10.1016/j.acra.2025.11.028

Durjoy D. Dhruba MS , Sawyer Goetz MD , Otavio Ferreira Dalla Pria MD , Thomas Reith MD , Abigail Reutzel MD , Pritish Y. Aher MD , Prashant Nagpal MD , Sarv Priya Dr, MD

Rationale and Objectives

This study evaluates a fully automated deep learning framework to enhance the efficiency and accuracy of cardiac MRI planning.

Materials and Methods

In this retrospective study, data from 1023 patients (ages 8–90 years) who underwent cardiac MRI were analyzed, including coronal, sagittal, axial localizers, and short-axis (SAX) and long-axis cine images. Experts manually annotated landmarks, serving as the ground truth for developing deep learning models. The models were assessed using 5-fold cross-validation. Performance metrics included median landmark distances and plane angle differences.

Results

The model achieved robust performance in landmark localization across all cardiac MRI planes. For localizer images, median distances were 5.1 mm (superior) and 7.2 mm (inferior) on coronal views, and 5.6 mm (superior) and 7.5 mm (inferior) on sagittal views. Median distances for axial, 2-chamber, and 4-chamber landmarks were 5.2 mm, 5.2 mm, and 5.6 mm, respectively. In short-axis mid slices, annotations based on the left ventricular center, right ventricular insertion points, and right ventricle obtuse angle had a median error of 5.2 mm, while basal slice valve-based annotations had 4.6 mm error. Angular deviations for SAX planning were 2.0° (2CH) and 1.5° (4CH). For long-axis views, angulation errors were lower using SAX mid slices (3.3° for 2CH, 2.6° for 4CH) compared to SAX base (4.0° and 3.9°, respectively).

Conclusion

A deep learning-based automated workflow for cardiac MRI planning is feasible with improved precision.

基本原理和目的：本研究评估了一个全自动深度学习框架，以提高心脏MRI计划的效率和准确性。材料和方法：在这项回顾性研究中，分析了1023例（8-90岁）接受心脏MRI检查的患者的数据，包括冠状、矢状、轴向定位、短轴（SAX）和长轴电影图像。专家手动标注地标，作为开发深度学习模型的基础。采用5倍交叉验证对模型进行评估。性能指标包括中位数地标距离和平面角度差。结果：该模型在所有心脏MRI平面的地标定位方面取得了稳健的表现。对于定位器图像，冠状位上的中位距离为5.1 mm（上）和7.2 mm（下），矢状位上的中位距离为5.6 mm（上）和7.5 mm（下）。轴向、2室和4室标志的中位距离分别为5.2 mm、5.2 mm和5.6 mm。在短轴正中切片中，基于左心室中心、右心室插入点和右心室钝角的标注中值误差为5.2 mm，而基于基底片瓣的标注中值误差为4.6 mm。SAX规划的角度偏差分别为2.0°（2CH）和1.5°（4CH）。对于长轴视图，与SAX基片（分别为4.0°和3.9°）相比，使用SAX中片（2CH为3.3°，4CH为2.6°）的角度误差更低。结论：一种基于深度学习的心脏MRI规划自动化工作流程是可行的，并且精度更高。

{"title":"Deep Learning-Based Cardiac MRI Planning from Localizers to Cine Views Using Landmark Detection","authors":"Durjoy D. Dhruba MS , Sawyer Goetz MD , Otavio Ferreira Dalla Pria MD , Thomas Reith MD , Abigail Reutzel MD , Pritish Y. Aher MD , Prashant Nagpal MD , Sarv Priya Dr, MD","doi":"10.1016/j.acra.2025.11.028","DOIUrl":"10.1016/j.acra.2025.11.028","url":null,"abstract":"<div><h3>Rationale and Objectives</h3><div>This study evaluates a fully automated deep learning framework to enhance the efficiency and accuracy of cardiac MRI planning.</div></div><div><h3>Materials and Methods</h3><div>In this retrospective study, data from 1023 patients (ages 8–90 years) who underwent cardiac MRI were analyzed, including coronal, sagittal, axial localizers, and short-axis (SAX) and long-axis cine images. Experts manually annotated landmarks, serving as the ground truth for developing deep learning models. The models were assessed using 5-fold cross-validation. Performance metrics included median landmark distances and plane angle differences.</div></div><div><h3>Results</h3><div>The model achieved robust performance in landmark localization across all cardiac MRI planes. For localizer images, median distances were 5.1 mm (superior) and 7.2 mm (inferior) on coronal views, and 5.6 mm (superior) and 7.5 mm (inferior) on sagittal views. Median distances for axial, 2-chamber, and 4-chamber landmarks were 5.2 mm, 5.2 mm, and 5.6 mm, respectively. In short-axis mid slices, annotations based on the left ventricular center, right ventricular insertion points, and right ventricle obtuse angle had a median error of 5.2 mm, while basal slice valve-based annotations had 4.6 mm error. Angular deviations for SAX planning were 2.0° (2CH) and 1.5° (4CH). For long-axis views, angulation errors were lower using SAX mid slices (3.3° for 2CH, 2.6° for 4CH) compared to SAX base (4.0° and 3.9°, respectively).</div></div><div><h3>Conclusion</h3><div>A deep learning-based automated workflow for cardiac MRI planning is feasible with improved precision.</div></div>","PeriodicalId":50928,"journal":{"name":"Academic Radiology","volume":"33 3","pages":"Pages 924-935"},"PeriodicalIF":3.9,"publicationDate":"2026-03-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145696480","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"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

Comparison of Postoperative Adjuvant CTACE vs DEB-TACE Combined with Tislelizumab and Lenvatinib in BCLC Stage B/C Hepatocellular Carcinoma: A Propensity Score-Matched Analysis of Survival Outcomes BCLC期B/C肝细胞癌术后辅助CTACE与DEB-TACE联合Tislelizumab和Lenvatinib的比较：生存结果的倾向评分匹配分析

IF 3.9 2区医学 Q1 RADIOLOGY, NUCLEAR MEDICINE & MEDICAL IMAGING

Academic Radiology

Pub Date : 2026-03-01 Epub Date: 2025-12-16 DOI: 10.1016/j.acra.2025.11.046

Yuan Shi , Liping Yang , Xiaodong Li , Han Zhang , Honghan Jiang , Kai Chen , Zhengrong Shi

Purpose

This single-center, retrospective study aimed to compare the efficacy and safety of two adjuvant triple-therapy regimens—conventional transarterial chemoembolization (CTACE) versus drug-eluting bead TACE (DEB-TACE), each combined with tislelizumab and lenvatinib—in patients with BCLC stage B/C hepatocellular carcinoma (HCC) following curative resection.

Methods

Using propensity score matching to minimize selection bias, we analyzed data from 211 eligible patients treated between March 2021 and June 2024. Key endpoints included overall survival (OS), progression-free survival (PFS), and adverse events (AEs).

Results

After matching, the DEB-TACE-based regimen (DEB-TACE-T-L) was associated with significantly longer OS and PFS compared to the CTACE-based regimen (CTACE-T-L). In an exploratory subgroup analysis of 49 patients with microvascular invasion (MVI), those receiving DEB-TACE-T-L showed a promising trend toward improved survival outcomes; however, this finding should be interpreted as hypothesis-generating due to the limited sample size and lack of a statistically significant interaction between treatment type and MVI status. The safety profile was comparable between the two groups.

Conclusion

Our preliminary results suggest that adjuvant DEB-TACE-T-L may offer a survival benefit over CTACE-T-L in resected BCLC B/C HCC, with acceptable toxicity. The potential enhanced effect in MVI-positive patients observed here warrants further validation in larger, prospective studies.

目的：这项单中心、回顾性研究旨在比较两种辅助三联治疗方案——常规经动脉化疗栓塞（CTACE）和药物洗脱头TACE (debe -TACE)，分别联合替利单抗和lenvatinib在BCLC期B/C肝细胞癌（HCC）根治切除后的疗效和安全性。方法：采用倾向评分匹配来最小化选择偏差，我们分析了211例2021年3月至2024年6月期间接受治疗的符合条件的患者的数据。主要终点包括总生存期（OS）、无进展生存期（PFS）和不良事件（ae）。结果：匹配后，与基于ctace的方案（CTACE-T-L）相比，基于debe - tace的方案（debe - tace - t - l）的OS和PFS明显更长。在49例微血管侵犯（MVI）患者的探索性亚组分析中，接受DEB-TACE-T-L治疗的患者显示出改善生存结果的良好趋势；然而，由于样本量有限，且治疗类型和MVI状态之间缺乏统计学上显著的相互作用，这一发现应被解释为假设产生。两组之间的安全性具有可比性。结论：我们的初步结果表明，在切除的BCLC B/C HCC中，辅助剂DEB-TACE-T-L可能比cace - t - l提供生存优势，毒性可接受。在mvi阳性患者中观察到的潜在增强效应值得在更大规模的前瞻性研究中进一步验证。

{"title":"Comparison of Postoperative Adjuvant CTACE vs DEB-TACE Combined with Tislelizumab and Lenvatinib in BCLC Stage B/C Hepatocellular Carcinoma: A Propensity Score-Matched Analysis of Survival Outcomes","authors":"Yuan Shi , Liping Yang , Xiaodong Li , Han Zhang , Honghan Jiang , Kai Chen , Zhengrong Shi","doi":"10.1016/j.acra.2025.11.046","DOIUrl":"10.1016/j.acra.2025.11.046","url":null,"abstract":"<div><h3>Purpose</h3><div>This single-center, retrospective study aimed to compare the efficacy and safety of two adjuvant triple-therapy regimens—conventional transarterial chemoembolization (CTACE) versus drug-eluting bead TACE (DEB-TACE), each combined with tislelizumab and lenvatinib—in patients with BCLC stage B/C hepatocellular carcinoma (HCC) following curative resection.</div></div><div><h3>Methods</h3><div>Using propensity score matching to minimize selection bias, we analyzed data from 211 eligible patients treated between March 2021 and June 2024. Key endpoints included overall survival (OS), progression-free survival (PFS), and adverse events (AEs).</div></div><div><h3>Results</h3><div>After matching, the DEB-TACE-based regimen (DEB-TACE-T-L) was associated with significantly longer OS and PFS compared to the CTACE-based regimen (CTACE-T-L). In an exploratory subgroup analysis of 49 patients with microvascular invasion (MVI), those receiving DEB-TACE-T-L showed a promising trend toward improved survival outcomes; however, this finding should be interpreted as hypothesis-generating due to the limited sample size and lack of a statistically significant interaction between treatment type and MVI status. The safety profile was comparable between the two groups.</div></div><div><h3>Conclusion</h3><div>Our preliminary results suggest that adjuvant DEB-TACE-T-L may offer a survival benefit over CTACE-T-L in resected BCLC B/C HCC, with acceptable toxicity. The potential enhanced effect in MVI-positive patients observed here warrants further validation in larger, prospective studies.</div></div>","PeriodicalId":50928,"journal":{"name":"Academic Radiology","volume":"33 3","pages":"Pages 990-1004"},"PeriodicalIF":3.9,"publicationDate":"2026-03-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145769729","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"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