Pub Date : 2026-03-01Epub Date: 2025-11-29DOI: 10.1016/j.acra.2025.11.025
Mohamad M. Alzein MD , Andrew C. Gordon MD, PhD , Maha H. Hussain MD , Riad Salem MD , Robert J. Lewandowski MD, FSIR
Rationale and Objectives
This study reports outcomes of patients undergoing transarterial radioembolization (TARE) utilizing yttrium-90 (Y90) for androgen-independent prostate cancer liver metastasis.
Materials and Methods
A retrospective review was conducted for seven patients treated with TARE between 2007 and January 2024. Index tumor response was described through the Response Evaluation Criteria in Solid Tumors (RECIST) version 1.1. Survival analysis was calculated by the Kaplan–Meier method for index tumor, liver, extra-hepatic, and time-to-progression, as well as overall survival from day of TARE. Adverse events within one month of TARE were evaluated by the Common Terminology Criteria for Adverse Events (CTCAE) version 5.
Results
The median age of our cohort was 60.9 years (range, 52.3–79.2 years). The index tumor was treated with a median dose of 95.0 Gy (range, 28.7–300.7 Gy) and activity of 1.2 GBq (range, 0.7–2.5 GBq). Imaging follow up was completed for 86% (n = 6), with one early death. By RECIST criteria, 50% (n = 3) of patients achieved partial response and 50% (n = 3) of patients had stable disease as their final imaging responses. No index tumor progressed based on the RECIST criteria. Median progression was 2.3 (range, 1.3–6.9), 2.8 (range, 1.3–6.9), and 7.0 (range, 1.3–22.1) months for time-to-, hepatic, and extrahepatic progression, respectively. Median overall survival was 16.2 months (range, 1.0–93.2 months). One patient died within 30 days of the procedure. One patient reported CTCAE grade 3 effect: fatigue (n = 1).
Conclusion
TARE demonstrates antitumor activity and manageable toxicity in androgen-independent prostate cancer liver metastasis. However, optimal treatment timing remains uncertain.
{"title":"Yttrium-90 Radioembolization for Androgen-Independent Prostate Cancer Metastasis to the Liver","authors":"Mohamad M. Alzein MD , Andrew C. Gordon MD, PhD , Maha H. Hussain MD , Riad Salem MD , Robert J. Lewandowski MD, FSIR","doi":"10.1016/j.acra.2025.11.025","DOIUrl":"10.1016/j.acra.2025.11.025","url":null,"abstract":"<div><h3>Rationale and Objectives</h3><div>This study reports outcomes of patients undergoing transarterial radioembolization (TARE) utilizing yttrium-90 (Y90) for androgen-independent prostate cancer liver metastasis.</div></div><div><h3>Materials and Methods</h3><div>A retrospective review was conducted for seven patients treated with TARE between 2007 and January 2024. Index tumor response was described through the Response Evaluation Criteria in Solid Tumors (RECIST) version 1.1. Survival analysis was calculated by the Kaplan–Meier method for index tumor, liver, extra-hepatic, and time-to-progression, as well as overall survival from day of TARE. Adverse events within one month of TARE were evaluated by the Common Terminology Criteria for Adverse Events (CTCAE) version 5.</div></div><div><h3>Results</h3><div>The median age of our cohort was 60.9 years (range, 52.3–79.2 years). The index tumor was treated with a median dose of 95.0 Gy (range, 28.7–300.7 Gy) and activity of 1.2 GBq (range, 0.7–2.5 GBq). Imaging follow up was completed for 86% (<em>n</em> = 6), with one early death. By RECIST criteria, 50% (<em>n</em> = 3) of patients achieved partial response and 50% (<em>n</em> = 3) of patients had stable disease as their final imaging responses. No index tumor progressed based on the RECIST criteria. Median progression was 2.3 (range, 1.3–6.9), 2.8 (range, 1.3–6.9), and 7.0 (range, 1.3–22.1) months for time-to-, hepatic, and extrahepatic progression, respectively. Median overall survival was 16.2 months (range, 1.0–93.2 months). One patient died within 30 days of the procedure. One patient reported CTCAE grade 3 effect: fatigue (<em>n</em> = 1).</div></div><div><h3>Conclusion</h3><div>TARE demonstrates antitumor activity and manageable toxicity in androgen-independent prostate cancer liver metastasis. However, optimal treatment timing remains uncertain.</div></div>","PeriodicalId":50928,"journal":{"name":"Academic Radiology","volume":"33 3","pages":"Pages 1005-1011"},"PeriodicalIF":3.9,"publicationDate":"2026-03-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145649771","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}
Pub Date : 2026-03-01Epub Date: 2025-12-27DOI: 10.1016/j.acra.2025.12.012
Bin Zhang , Shumao Zhang , Guang Tian , Haohan Wang , Deyuan Zhu , Yaodan Ban , Benqiang Yang
Rationale and Objectives
Tumor heterogeneity is a main driver of varied treatment responses for neoadjuvant therapy (NAT) in breast cancer, yet conventional radiomics often overlooks intratumoral heterogeneity. This study aims to develop and validate a habitat radiomics model derived from pretreatment MRI for noninvasive prediction of pathologic complete response (pCR) in breast cancer patients undergoing NAT.
Materials and Methods
This retrospective multicenter study included 301 patients with breast cancer who underwent NAT followed by surgery. Patients were assigned to a training set (n=143), internal validation set (n=62), and two external testing sets (n=96). Habitat subregions were generated via K-means clustering, and concentric peritumoral regions (3 mm, 5 mm, and 7 mm) were delineated. Several models were developed: a clinical model, an intratumoral radiomics model, a habitat radiomics model, three peritumoral models, and a combined model. Models were evaluated using the area under the receiver operating characteristic curve (AUC), while calibration and decision curves were used to assess model reliability and clinical applicability. Model interpretability was assessed using Shapley Additive exPlanations (SHAP).
Results
The habitat radiomics model achieved AUCs of 0.931 (training), 0.850 (internal validation), 0.811 (external test 1), and 0.802 (external test 2), outperforming both global and peritumoral radiomics models. The combined model integrating 3 mm peritumoral features and clinical factors achieved higher AUCs of 0.957, 0.871, 0.842, and 0.853, respectively. SHAP analysis revealed that four of the top five contributing features originated from one dominant intratumoral habitat, highlighting habitat subregions as an important predictive marker of pCR.
Conclusion
MRI-based habitat radiomics enables noninvasive prediction of pCR by capturing spatial heterogeneity within and around tumors. This approach may improve individualized treatment planning in breast cancer.
{"title":"MRI-based Intratumoral and Peritumoral Habitat Radiomics for Early Prediction of Pathologic Complete Response in Breast Cancer: A Multicenter Study","authors":"Bin Zhang , Shumao Zhang , Guang Tian , Haohan Wang , Deyuan Zhu , Yaodan Ban , Benqiang Yang","doi":"10.1016/j.acra.2025.12.012","DOIUrl":"10.1016/j.acra.2025.12.012","url":null,"abstract":"<div><h3>Rationale and Objectives</h3><div>Tumor heterogeneity is a main driver of varied treatment responses for neoadjuvant therapy (NAT) in breast cancer, yet conventional radiomics often overlooks intratumoral heterogeneity. This study aims to develop and validate a habitat radiomics model derived from pretreatment MRI for noninvasive prediction of pathologic complete response (pCR) in breast cancer patients undergoing NAT.</div></div><div><h3>Materials and Methods</h3><div>This retrospective multicenter study included 301 patients with breast cancer who underwent NAT followed by surgery. Patients were assigned to a training set (n=143), internal validation set (n=62), and two external testing sets (n=96). Habitat subregions were generated via K-means clustering, and concentric peritumoral regions (3 mm, 5 mm, and 7 mm) were delineated. Several models were developed: a clinical model, an intratumoral radiomics model, a habitat radiomics model, three peritumoral models, and a combined model. Models were evaluated using the area under the receiver operating characteristic curve (AUC), while calibration and decision curves were used to assess model reliability and clinical applicability. Model interpretability was assessed using Shapley Additive exPlanations (SHAP).</div></div><div><h3>Results</h3><div>The habitat radiomics model achieved AUCs of 0.931 (training), 0.850 (internal validation), 0.811 (external test 1), and 0.802 (external test 2), outperforming both global and peritumoral radiomics models. The combined model integrating 3 mm peritumoral features and clinical factors achieved higher AUCs of 0.957, 0.871, 0.842, and 0.853, respectively. SHAP analysis revealed that four of the top five contributing features originated from one dominant intratumoral habitat, highlighting habitat subregions as an important predictive marker of pCR.</div></div><div><h3>Conclusion</h3><div>MRI-based habitat radiomics enables noninvasive prediction of pCR by capturing spatial heterogeneity within and around tumors. This approach may improve individualized treatment planning in breast cancer.</div></div>","PeriodicalId":50928,"journal":{"name":"Academic Radiology","volume":"33 3","pages":"Pages 889-899"},"PeriodicalIF":3.9,"publicationDate":"2026-03-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145851516","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}
Pub Date : 2026-03-01Epub Date: 2026-03-02DOI: 10.1016/j.acra.2026.01.046
Ceylan Z. Cankurtaran MD , Joseph Fotos MD , Melis Ozkan BS , Vidya Sankar Viswanathan MBBS , Daphne Zhu BS , Jessica M. Sin MD, PhD , Monica Cheng MD , Nicole Brofman MD , Anna Rozenshtein MD , Michele Retrouvey MD
Recent advancements in imaging technologies are poised to redefine the diagnostic and therapeutic role of radiology. In part 2 of this review series, we outline emerging innovations with transformative clinical potential, including photon-counting detector CT for ultra-high-resolution, spectral imaging; ultra-high-field MRI systems (≥7 T); and the expanding field of theranostics. We also examine nanotechnology-enhanced imaging agents, the integration of AI-driven opportunistic screening, and radiogenomics to enable precision diagnostics and early disease detection. Together, these developments represent a paradigm shift toward more personalized, data-rich, and preventative imaging approaches. Challenges in implementation, such as safety, cost, and workflow integration, are discussed with an emphasis on collaboration and infrastructure to support sustained clinical translation.
{"title":"A RRA Perspective on Advanced Imaging in Radiology: Emerging Technologies for Precision and Personalized Care","authors":"Ceylan Z. Cankurtaran MD , Joseph Fotos MD , Melis Ozkan BS , Vidya Sankar Viswanathan MBBS , Daphne Zhu BS , Jessica M. Sin MD, PhD , Monica Cheng MD , Nicole Brofman MD , Anna Rozenshtein MD , Michele Retrouvey MD","doi":"10.1016/j.acra.2026.01.046","DOIUrl":"10.1016/j.acra.2026.01.046","url":null,"abstract":"<div><div>Recent advancements in imaging technologies are poised to redefine the diagnostic and therapeutic role of radiology. In part 2 of this review series, we outline emerging innovations with transformative clinical potential, including photon-counting detector CT for ultra-high-resolution, spectral imaging; ultra-high-field MRI systems (≥7<!--> <!-->T); and the expanding field of theranostics. We also examine nanotechnology-enhanced imaging agents, the integration of AI-driven opportunistic screening, and radiogenomics to enable precision diagnostics and early disease detection. Together, these developments represent a paradigm shift toward more personalized, data-rich, and preventative imaging approaches. Challenges in implementation, such as safety, cost, and workflow integration, are discussed with an emphasis on collaboration and infrastructure to support sustained clinical translation.</div></div>","PeriodicalId":50928,"journal":{"name":"Academic Radiology","volume":"33 3","pages":"Pages 630-638"},"PeriodicalIF":3.9,"publicationDate":"2026-03-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"147356790","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}
{"title":"Bridging the Transparency Gap: Enhancing Disclosure of Generative AI Tools in Radiology Research Manuscripts","authors":"Herlina Uinarni , Laith Saheb , Nigora Djuraeva , Aria Diba","doi":"10.1016/j.acra.2025.10.053","DOIUrl":"10.1016/j.acra.2025.10.053","url":null,"abstract":"","PeriodicalId":50928,"journal":{"name":"Academic Radiology","volume":"33 3","pages":"Pages 716-717"},"PeriodicalIF":3.9,"publicationDate":"2026-03-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145530399","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}
Pub Date : 2026-03-01Epub Date: 2025-11-17DOI: 10.1016/j.acra.2025.10.050
Quan Yuan , Zhipeng Hong , Rongjie Ye , Peidong Yang , Juli Lin , Xinyu Jiang , Huilei Qiu , Aoyu Liu , Hao Yu , Fei Gao , Pengfei He , Kaizhou Chen , Jian Cai , Xinru Xie , Wenkang You , Haiyu Yuan , Kejie Zhang , Shuqi Yang , Boqian Yu , Xinfa Huang , Ming Niu
Rationale and Objectives
Breast cancer (BC) remains a leading contributor to the global cancer burden among women, with neoadjuvant chemotherapy (NAC) established as the standard of care for early-stage disease. However, substantial interpatient variability in treatment outcomes persists, primarily driven by inherent tumor biological heterogeneity. This underscores an urgent need for more precise prognostic tools to optimize clinical decision-making.
Materials and Methods
This multicenter study included 216 BC patients who completed NAC, with no overlap in datasets with previous research. We extracted four-dimensional data: clinical characteristics, pathomics features, deep learning-derived pathological features (via ResNet50), and multiparametric MRI (mpMRI) radiomics. A multimodal Cox model integrating deep feature representations and radiomic variables was constructed to combine these data. Notably, this approach differs from prior studies, which have predominantly focused on single-modality inputs (eg, radiomics or pathomics alone) or short-term endpoints such as pathological complete response (pCR).
Results
The proposed model, leveraging deep feature representations derived from CNNs and radiomic fusion, achieved superior prognostic accuracy in predicting 5-year and 7-year overall survival (OS) compared to both single-modality models and findings from previous research. For 5-year OS, it achieved an area under the receiver operating characteristic curve (AUC) of 0.890 in the training set and 0.820 in the validation set; for 7-year OS, the AUC values were 0.910 (training) and 0.870 (validation), with statistically significant superiority over unidimensional models. Calibration curves and decision curve analyses further confirmed its robust clinical utility.
Conclusion
The multimodal integration of imaging, pathology, and clinical data, particularly the inclusion of CNN-derived deep features, provides complementary information that improves survival prediction in NAC-treated BC patients. This represents a meaningful advancement over existing models that rely on single-modality data or focus on short-term outcomes.
Research registration unique identifying number
The study is registered at https://www.chictr.org.cn and has acquired only Identifier: ChiCTR2500098023.
{"title":"Integrating Deep Feature Extraction and MRI Radiomics for Survival Prediction in Breast Cancer After Neoadjuvant Chemotherapy","authors":"Quan Yuan , Zhipeng Hong , Rongjie Ye , Peidong Yang , Juli Lin , Xinyu Jiang , Huilei Qiu , Aoyu Liu , Hao Yu , Fei Gao , Pengfei He , Kaizhou Chen , Jian Cai , Xinru Xie , Wenkang You , Haiyu Yuan , Kejie Zhang , Shuqi Yang , Boqian Yu , Xinfa Huang , Ming Niu","doi":"10.1016/j.acra.2025.10.050","DOIUrl":"10.1016/j.acra.2025.10.050","url":null,"abstract":"<div><h3>Rationale and Objectives</h3><div>Breast cancer (BC) remains a leading contributor to the global cancer burden among women, with neoadjuvant chemotherapy (NAC) established as the standard of care for early-stage disease. However, substantial interpatient variability in treatment outcomes persists, primarily driven by inherent tumor biological heterogeneity. This underscores an urgent need for more precise prognostic tools to optimize clinical decision-making.</div></div><div><h3>Materials and Methods</h3><div>This multicenter study included 216 BC patients who completed NAC, with no overlap in datasets with previous research. We extracted four-dimensional data: clinical characteristics, pathomics features, deep learning-derived pathological features (via ResNet50), and multiparametric MRI (mpMRI) radiomics. A multimodal Cox model integrating deep feature representations and radiomic variables was constructed to combine these data. Notably, this approach differs from prior studies, which have predominantly focused on single-modality inputs (eg, radiomics or pathomics alone) or short-term endpoints such as pathological complete response (pCR).</div></div><div><h3>Results</h3><div>The proposed model, leveraging deep feature representations derived from CNNs and radiomic fusion, achieved superior prognostic accuracy in predicting 5-year and 7-year overall survival (OS) compared to both single-modality models and findings from previous research. For 5-year OS, it achieved an area under the receiver operating characteristic curve (AUC) of 0.890 in the training set and 0.820 in the validation set; for 7-year OS, the AUC values were 0.910 (training) and 0.870 (validation), with statistically significant superiority over unidimensional models. Calibration curves and decision curve analyses further confirmed its robust clinical utility.</div></div><div><h3>Conclusion</h3><div>The multimodal integration of imaging, pathology, and clinical data, particularly the inclusion of CNN-derived deep features, provides complementary information that improves survival prediction in NAC-treated BC patients. This represents a meaningful advancement over existing models that rely on single-modality data or focus on short-term outcomes.</div></div><div><h3>Research registration unique identifying number</h3><div>The study is registered at https://www.chictr.org.cn and has acquired only Identifier: ChiCTR2500098023.</div></div>","PeriodicalId":50928,"journal":{"name":"Academic Radiology","volume":"33 3","pages":"Pages 872-888"},"PeriodicalIF":3.9,"publicationDate":"2026-03-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145551486","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}
Pub Date : 2026-03-01Epub Date: 2025-11-19DOI: 10.1016/j.acra.2025.10.063
Pranay Chetan Uppuluri MD (Chief Radiologist) , Carina W. Yang MD (Associate Professor)
Foundation Models (FMs) mark a significant evolution in medical AI, enabling multimodal and multitask performance across text and imaging. Radiology, with its structured data formats and early adoption of AI, is uniquely positioned to benefit from FM capabilities. However, despite promising technical advances, questions remain about their clinical readiness, safety, and regulatory oversight. This narrative review explores the development, utility, and implementation challenges of FMs in U.S. radiology. Literature from PubMed, Scopus, arXiv, and IEEE Xplore (January 2022 to May 2025) was synthesized to highlight architectural trends, clinical applications, evaluation methods, and regulatory developments. U.S.-based models like CheXzero, BioMedCLIP, and Med-PaLM demonstrate strong diagnostic and reporting performance but face key limitations—including lack of FDA clearance, limited external validation, and integration barriers with PACS/RIS systems. Safety issues such as hallucination, automation bias, and underperformance in edge cases persist. While human-in-the-loop frameworks, federated learning, and emerging reporting standards show promise, institutional readiness and regulatory clarity remain fragmented. We propose a roadmap that includes continuous monitoring, equity-focused design, and a national FM registry to guide responsible deployment. Radiology’s digital maturity makes it a critical testbed for foundational AI integration—offering lessons for broader clinical adoption.
{"title":"The Emergence of Foundation Models in U.S. Radiology: A Narrative Review of Clinical Utility, Safety, and Evaluation","authors":"Pranay Chetan Uppuluri MD (Chief Radiologist) , Carina W. Yang MD (Associate Professor)","doi":"10.1016/j.acra.2025.10.063","DOIUrl":"10.1016/j.acra.2025.10.063","url":null,"abstract":"<div><div>Foundation Models (FMs) mark a significant evolution in medical AI, enabling multimodal and multitask performance across text and imaging. Radiology, with its structured data formats and early adoption of AI, is uniquely positioned to benefit from FM capabilities. However, despite promising technical advances, questions remain about their clinical readiness, safety, and regulatory oversight. This narrative review explores the development, utility, and implementation challenges of FMs in U.S. radiology. Literature from PubMed, Scopus, arXiv, and IEEE Xplore (January 2022 to May 2025) was synthesized to highlight architectural trends, clinical applications, evaluation methods, and regulatory developments. U.S.-based models like CheXzero, BioMedCLIP, and Med-PaLM demonstrate strong diagnostic and reporting performance but face key limitations—including lack of FDA clearance, limited external validation, and integration barriers with PACS/RIS systems. Safety issues such as hallucination, automation bias, and underperformance in edge cases persist. While human-in-the-loop frameworks, federated learning, and emerging reporting standards show promise, institutional readiness and regulatory clarity remain fragmented. We propose a roadmap that includes continuous monitoring, equity-focused design, and a national FM registry to guide responsible deployment. Radiology’s digital maturity makes it a critical testbed for foundational AI integration—offering lessons for broader clinical adoption.</div></div>","PeriodicalId":50928,"journal":{"name":"Academic Radiology","volume":"33 3","pages":"Pages 797-814"},"PeriodicalIF":3.9,"publicationDate":"2026-03-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145566280","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}
Pub Date : 2026-03-01Epub Date: 2025-11-24DOI: 10.1016/j.acra.2025.11.009
Xinhang Gu , Junfeng Zhao , Jiaxiao Geng , Ying Li , Chengxin Liu
Rationale and Objectives
It remains uncertain whether all patients with oligometastatic non-small cell lung cancer (NSCLC) benefit from the combination of third-generation EGFR-TKIs and TRT. This study aimed to identify which patients are most likely to benefit from combined third-generation EGFR-TKI and TRT, and which patients may safely omit TRT, thereby guiding clinical decision-making and optimizing prognosis.
Materials and Methods
A total of 338 patients with EGFR-mutated oligometastatic NSCLC who received first-line third-generation EGFR-TKI treatment were included. These patients were divided into training and validation cohorts. Univariate and multivariate analyses incorporating clinicopathological variables and radiomic features were performed to identify independent prognostic factors for progression-free survival (PFS) and overall survival (OS). A predictive nomogram was developed based on these factors and validated using receiver operating characteristic curves, calibration curves, and decision curve analysis.
Results
EGFR exon 21 L858R mutation, brain metastasis, neutrophil-to-lymphocyte ratio ≥ 4.39, and the radiomic score (Rad-score) were identified as independent risk factors for PFS. Age > 60 years, EGFR exon 21 L858R mutation, brain metastasis, monocyte-to-lymphocyte ratio > 0.26, and Rad-score were OS independent predictors. In the training cohort, the nomogram achieved excellent predictive performance with AUCs of 0.858, 0.834, and 0.785 for 1-, 2-, and 3-year PFS, and 0.882, 0.868 and 0.877 for 2-, 3-, and 4-year OS, respectively. In the validation cohort, respective AUCs were 0.800, 0.740, and 0.734,and 0.835, 0.729, 0.766, confirming good discrimination. The model successfully stratified patients into low- and high-risk groups. High-risk patients derived significant PFS (p < 0.001) and OS (p < 0.001) benefits from TRT, whereas low-risk patients did not show significant improvements in PFS (p = 0.056) or OS (p = 0.093) with TRT.
Conclusion
We established and confirmed a robust predictive nomogram that integrates clinicopathological and radiomic factors to stratify patients with first-line therapy for EGFR-mutant oligometastatic NSCLC involving third-generation EGFR-TKIs. This approach helps determine which patients may gain the greatest benefit from combined TRT and help avoid unnecessary TRT in low-risk patients, supporting precision treatment strategies.
{"title":"Identifying Patients with EGFR-Mutated Oligometastatic NSCLC Suitable for Third-Generation EGFR-TKI Combined with Thoracic Radiotherapy Using Nomograms Based on CT Radiomic and Clinicopathological Factors","authors":"Xinhang Gu , Junfeng Zhao , Jiaxiao Geng , Ying Li , Chengxin Liu","doi":"10.1016/j.acra.2025.11.009","DOIUrl":"10.1016/j.acra.2025.11.009","url":null,"abstract":"<div><h3>Rationale and Objectives</h3><div>It remains uncertain whether all patients with oligometastatic non-small cell lung cancer (NSCLC) benefit from the combination of third-generation EGFR-TKIs and TRT. This study aimed to identify which patients are most likely to benefit from combined third-generation EGFR-TKI and TRT, and which patients may safely omit TRT, thereby guiding clinical decision-making and optimizing prognosis.</div></div><div><h3>Materials and Methods</h3><div>A total of 338 patients with EGFR-mutated oligometastatic NSCLC who received first-line third-generation EGFR-TKI treatment were included. These patients were divided into training and validation cohorts. Univariate and multivariate analyses incorporating clinicopathological variables and radiomic features were performed to identify independent prognostic factors for progression-free survival (PFS) and overall survival (OS). A predictive nomogram was developed based on these factors and validated using receiver operating characteristic curves, calibration curves, and decision curve analysis.</div></div><div><h3>Results</h3><div>EGFR exon 21 L858R mutation, brain metastasis, neutrophil-to-lymphocyte ratio ≥ 4.39, and the radiomic score (Rad-score) were identified as independent risk factors for PFS. Age > 60 years, EGFR exon 21 L858R mutation, brain metastasis, monocyte-to-lymphocyte ratio > 0.26, and Rad-score were OS independent predictors. In the training cohort, the nomogram achieved excellent predictive performance with AUCs of 0.858, 0.834, and 0.785 for 1-, 2-, and 3-year PFS, and 0.882, 0.868 and 0.877 for 2-, 3-, and 4-year OS, respectively. In the validation cohort, respective AUCs were 0.800, 0.740, and 0.734,and 0.835, 0.729, 0.766, confirming good discrimination. The model successfully stratified patients into low- and high-risk groups. High-risk patients derived significant PFS (p < 0.001) and OS (p < 0.001) benefits from TRT, whereas low-risk patients did not show significant improvements in PFS (p = 0.056) or OS (p = 0.093) with TRT.</div></div><div><h3>Conclusion</h3><div>We established and confirmed a robust predictive nomogram that integrates clinicopathological and radiomic factors to stratify patients with first-line therapy for EGFR-mutant oligometastatic NSCLC involving third-generation EGFR-TKIs. This approach helps determine which patients may gain the greatest benefit from combined TRT and help avoid unnecessary TRT in low-risk patients, supporting precision treatment strategies.</div></div>","PeriodicalId":50928,"journal":{"name":"Academic Radiology","volume":"33 3","pages":"Pages 1167-1179"},"PeriodicalIF":3.9,"publicationDate":"2026-03-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145607075","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}
Pub Date : 2026-03-01Epub Date: 2025-11-24DOI: 10.1016/j.acra.2025.10.045
Mina Dawod MD , Mensur Koso MD , Bidhu Sharma BS , Haris Mujovic BS , Amel Lilic BS , Matthew Yoder MD , Mina S. Makary MD
<div><h3>Rationale and Objectives</h3><div>To determine if student exposure to an endovascular simulator can increase confidence performing procedures and interest in interventional radiology (IR) and to assess if past and present video game experience confers improved procedural skills.</div></div><div><h3>Materials and Methods</h3><div>This IRB-approved prospective randomized control study evaluated medical student specialty interest and procedural performance on an endovascular stimulator before and after video game simulation. Participating students were required to complete a pre-procedure survey and an initial simulated procedure utilizing an endovascular simulator before being randomized to either a Video Game (VG) or control arm. Students then proceeded to complete a second simulated procedure and a post-procedure survey. Before starting the procedure, a standardized explanation was read to each participant. Students randomized to the VG arm would play a video game within a ten-minute slot between the two procedures, while control arm students would rest. Survey data included demographic information, history of video gaming, confidence performing endovascular procedures, and interest in procedural specialties. Primary outcomes included self-reported procedural confidence and specialty interest on a five-point scale (five being the highest). Secondary outcomes included procedural skill as measured by time to completion during the simulated procedures.</div></div><div><h3>Results</h3><div>A total of 52 medical students (mean age, 25.9 years; male 58%) participated in this study, with 26 students randomized to the VG arm and the remainder to the control arm. Of the total cohort, 31% (n=16) identified as first year medical students, 31% (n=16) as third year, 23% (n=12) as fourth year, 10% (n=5) as second year, and 6% (n=3) as leave of absence students. The cohort’s average confidence performing procedures prior to participation was 2.4 (out of five), interest in procedural specialties was 4.1, and interest in IR was 2.8. After participation in the study, confidence performing procedures rose by 57% to 3.8 (<em>p</em><0.0001), interest in procedural specialties rose by 8% to 4.4 (<em>p</em>=0.0016), and interest in IR increased by 28% to 3.6 (<em>p</em><0.0001). The collective cohort improved by an average of 23% between procedure one completion time and procedure two completion time (3 min 56 s to 3 min one second). The differences in the rate of improvement between the VG group and control group was not significant. Gender was found to be the only background variable to significantly correlate with procedural times (<em>p</em>=0.01).</div></div><div><h3>Conclusion</h3><div>Student confidence performing procedures significantly increased after participating in the study, as did student interest in procedural specialties in general and in IR specifically. A history of video games and prospective VG group participation did not confer proc
{"title":"Impact of an Endovascular Simulator and Video Games on Medical Student Procedural Outcomes and Interventional Radiology Interest","authors":"Mina Dawod MD , Mensur Koso MD , Bidhu Sharma BS , Haris Mujovic BS , Amel Lilic BS , Matthew Yoder MD , Mina S. Makary MD","doi":"10.1016/j.acra.2025.10.045","DOIUrl":"10.1016/j.acra.2025.10.045","url":null,"abstract":"<div><h3>Rationale and Objectives</h3><div>To determine if student exposure to an endovascular simulator can increase confidence performing procedures and interest in interventional radiology (IR) and to assess if past and present video game experience confers improved procedural skills.</div></div><div><h3>Materials and Methods</h3><div>This IRB-approved prospective randomized control study evaluated medical student specialty interest and procedural performance on an endovascular stimulator before and after video game simulation. Participating students were required to complete a pre-procedure survey and an initial simulated procedure utilizing an endovascular simulator before being randomized to either a Video Game (VG) or control arm. Students then proceeded to complete a second simulated procedure and a post-procedure survey. Before starting the procedure, a standardized explanation was read to each participant. Students randomized to the VG arm would play a video game within a ten-minute slot between the two procedures, while control arm students would rest. Survey data included demographic information, history of video gaming, confidence performing endovascular procedures, and interest in procedural specialties. Primary outcomes included self-reported procedural confidence and specialty interest on a five-point scale (five being the highest). Secondary outcomes included procedural skill as measured by time to completion during the simulated procedures.</div></div><div><h3>Results</h3><div>A total of 52 medical students (mean age, 25.9 years; male 58%) participated in this study, with 26 students randomized to the VG arm and the remainder to the control arm. Of the total cohort, 31% (n=16) identified as first year medical students, 31% (n=16) as third year, 23% (n=12) as fourth year, 10% (n=5) as second year, and 6% (n=3) as leave of absence students. The cohort’s average confidence performing procedures prior to participation was 2.4 (out of five), interest in procedural specialties was 4.1, and interest in IR was 2.8. After participation in the study, confidence performing procedures rose by 57% to 3.8 (<em>p</em><0.0001), interest in procedural specialties rose by 8% to 4.4 (<em>p</em>=0.0016), and interest in IR increased by 28% to 3.6 (<em>p</em><0.0001). The collective cohort improved by an average of 23% between procedure one completion time and procedure two completion time (3 min 56 s to 3 min one second). The differences in the rate of improvement between the VG group and control group was not significant. Gender was found to be the only background variable to significantly correlate with procedural times (<em>p</em>=0.01).</div></div><div><h3>Conclusion</h3><div>Student confidence performing procedures significantly increased after participating in the study, as did student interest in procedural specialties in general and in IR specifically. A history of video games and prospective VG group participation did not confer proc","PeriodicalId":50928,"journal":{"name":"Academic Radiology","volume":"33 3","pages":"Pages 707-715"},"PeriodicalIF":3.9,"publicationDate":"2026-03-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145607068","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}
Pub Date : 2026-03-01Epub Date: 2025-12-26DOI: 10.1016/j.acra.2025.12.006
Zhang Jie , Li Xiaodan , Gao Mingjie , Sa Rina , Qu Xin , Wang Lijuan MD, PhD (Professor)
Objectives
This study aimed to develop and validate an integrated model incorporating clinical, radiomic, and deep learning features to predict long-term ipsilateral ischemic stroke after carotid artery stenting (CAS).
Methods
We analyzed 802 patients who underwent CAS at three centers between 2018 and 2024. Clinical and ultrasound data were collected, and radiomic and deep learning features were extracted from preoperative plaque images. A combined model was built using Cox regression and random survival forest models, and then presented as a nomogram. Model performance was assessed using the C-index and Kaplan–Meier analysis.
Results
Over a median follow-up of 62 months, 213 patients (26.6%) experienced ipsilateral stroke. We integrated the significant clinical, radiomics, and deep learning features (p<0.05) into a nomogram. The model demonstrated strong discriminative ability for stroke, with C-indices of 0.800 (95% CI: 0.759–0.841) in the training set, 0.751 (95% CI: 0.677–0.828) in the internal validation set, and 0.708 (95% CI: 0.634–0.782) in the external validation set. It significantly outperformed the conventional ultrasound model across all datasets (p<0.05). Moreover, Kaplan–Meier analysis confirmed that patients stratified into the high-risk group based on the nomogram had a substantially higher probability of ipsilateral stroke compared to the low-risk group (p<0.05).
Conclusion
A deep learning radiomics model based on preoperative ultrasound can predict long-time risk of ipsilateral stroke for patients with CAS, which may help in risk stratification and guide treatment decision-making and follow-up management.
{"title":"Deep Learning Radiomics Based on Preoperational Ultrasound Images for Predicting Ipsilateral Ischemic Stroke in Patients with Carotid Artery Stenting: A Multicenter Study","authors":"Zhang Jie , Li Xiaodan , Gao Mingjie , Sa Rina , Qu Xin , Wang Lijuan MD, PhD (Professor)","doi":"10.1016/j.acra.2025.12.006","DOIUrl":"10.1016/j.acra.2025.12.006","url":null,"abstract":"<div><h3>Objectives</h3><div>This study aimed to develop and validate an integrated model incorporating clinical, radiomic, and deep learning features to predict long-term ipsilateral ischemic stroke after carotid artery stenting (CAS).</div></div><div><h3>Methods</h3><div>We analyzed 802 patients who underwent CAS at three centers between 2018 and 2024. Clinical and ultrasound data were collected, and radiomic and deep learning features were extracted from preoperative plaque images. A combined model was built using Cox regression and random survival forest models, and then presented as a nomogram. Model performance was assessed using the C-index and Kaplan–Meier analysis.</div></div><div><h3>Results</h3><div>Over a median follow-up of 62 months, 213 patients (26.6%) experienced ipsilateral stroke. We integrated the significant clinical, radiomics, and deep learning features (<em>p</em><0.05) into a nomogram. The model demonstrated strong discriminative ability for stroke, with C-indices of 0.800 (95% CI: 0.759–0.841) in the training set, 0.751 (95% CI: 0.677–0.828) in the internal validation set, and 0.708 (95% CI: 0.634–0.782) in the external validation set. It significantly outperformed the conventional ultrasound model across all datasets (<em>p</em><0.05). Moreover, Kaplan–Meier analysis confirmed that patients stratified into the high-risk group based on the nomogram had a substantially higher probability of ipsilateral stroke compared to the low-risk group (<em>p</em><0.05).</div></div><div><h3>Conclusion</h3><div>A deep learning radiomics model based on preoperative ultrasound can predict long-time risk of ipsilateral stroke for patients with CAS, which may help in risk stratification and guide treatment decision-making and follow-up management.</div></div>","PeriodicalId":50928,"journal":{"name":"Academic Radiology","volume":"33 3","pages":"Pages 1083-1094"},"PeriodicalIF":3.9,"publicationDate":"2026-03-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145847009","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}
Pub Date : 2026-03-01Epub Date: 2026-03-02DOI: 10.1016/j.acra.2025.12.027
Joseph Fotos MD , Nicole Brofman MD , Melis Ozkan BS , Jessica M. Sin MD, PhD , Anna Rozenshtein MD , Michele Retrouvey MD
Advancements in three-dimensional (3D) visualization technologies, including virtual reality, augmented reality, and 3D printing, are opening new avenues for innovation within radiology. These tools enable immersive educational experiences, patient-specific procedural planning, and improved communication across interdisciplinary teams. In medical education, virtual and augmented reality have shown measurable benefits for both cognitive and procedural skills, providing scalable and cost-effective alternatives to traditional approaches. Similarly, 3D printing has advanced anatomical education, procedural preparation, and patient engagement through the creation of tangible, individualized models. Despite these promising applications, challenges such as time-intensive workflows, financial constraints, and the need for technical expertise continue to limit widespread clinical integration. Radiologists, with their expertise in image acquisition, segmentation, and interpretation, are well-positioned to lead adoption and implementation. As part 3 of the Radiology Research Alliance (RRA) review series on emerging technologies, this paper explores the applications, benefits, and barriers of visualization tools, emphasizing their growing potential to transform both education and patient care in an increasingly digital and value-driven healthcare environment.
{"title":"Visualization Tools in Radiology: A RRA Perspective on Virtual Reality, Augmented Reality, and 3D Printing","authors":"Joseph Fotos MD , Nicole Brofman MD , Melis Ozkan BS , Jessica M. Sin MD, PhD , Anna Rozenshtein MD , Michele Retrouvey MD","doi":"10.1016/j.acra.2025.12.027","DOIUrl":"10.1016/j.acra.2025.12.027","url":null,"abstract":"<div><div>Advancements in three-dimensional (3D) visualization technologies, including virtual reality, augmented reality, and 3D printing, are opening new avenues for innovation within radiology. These tools enable immersive educational experiences, patient-specific procedural planning, and improved communication across interdisciplinary teams. In medical education, virtual and augmented reality have shown measurable benefits for both cognitive and procedural skills, providing scalable and cost-effective alternatives to traditional approaches. Similarly, 3D printing has advanced anatomical education, procedural preparation, and patient engagement through the creation of tangible, individualized models. Despite these promising applications, challenges such as time-intensive workflows, financial constraints, and the need for technical expertise continue to limit widespread clinical integration. Radiologists, with their expertise in image acquisition, segmentation, and interpretation, are well-positioned to lead adoption and implementation. As part 3 of the Radiology Research Alliance (RRA) review series on emerging technologies, this paper explores the applications, benefits, and barriers of visualization tools, emphasizing their growing potential to transform both education and patient care in an increasingly digital and value-driven healthcare environment.</div></div>","PeriodicalId":50928,"journal":{"name":"Academic Radiology","volume":"33 3","pages":"Pages 639-645"},"PeriodicalIF":3.9,"publicationDate":"2026-03-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"147357225","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}
扫码关注我们
求助内容：
应助结果提醒方式：
京公网安备 11010802042870号