Insights into Imaging最新文献

{"title":"Chinese expert concern and consensus on applications of artificial intelligence in clinical cancer imaging.","authors":"Hong Wu, Xiaorui Yin, Jiejun Cheng, Jiayin Zhang, Linfeng Zheng, Lei Zhang, Feiyun Wu, Qiufeng Zhao, Jun Yang, Han Wang","doi":"10.1186/s13244-026-02359-5","DOIUrl":"10.1186/s13244-026-02359-5","url":null,"abstract":"<p><p>Artificial intelligence (AI) demonstrates potential throughout the cancer care continuum, with evidence supporting its application in medical imaging for detection, staging, treatment planning, and prognostic evaluation. However, clinical translation is hindered by challenges, data curation and annotation, model interpretability, generalizability, and integration into workflows. To address these barriers and provide guidance, a national multidisciplinary expert panel in China developed this consensus. A modified Delphi approach was employed to achieve expert consensus, involving 81 specialists in radiology, nuclear medicine, oncology, and imaging AI from university hospitals across China. These experts completed a survey containing 30 core statements addressing AI applications in clinical cancer imaging, spanning cancer screening, diagnosis, staging, treatment planning, response assessment, prognostic prediction, data governance, and implementation. Consensus was defined as a mean score ≥ 7 on a 9-point Likert scale, with ≥ 80% of experts scoring ≥ 7. All 30 statements fulfilled these thresholds, with mean scores ranging from 8.06 to 8.58 and the proportion of experts scoring ≥ 7 ranging from 86% to 98%. This expert consensus summarizes key AI application scenarios in cancer imaging and delivers recommendations on data acquisition and annotation, model development and validation, interpretability, multicenter generalizability, privacy-preserving collaboration, clinical workflow integration, and post-deployment monitoring, while contextualizing these statements across major clinical application domains and key implementation challenges in practice. It further identifies priority research directions, including the integration of multimodal and multi-omics data, longitudinal modeling of treatment response, and prospective validation in clinical settings, to support the safe, effective implementation of AI technologies in cancer imaging. KEY POINTS: Question AI translation in oncologic imaging remains constrained by limitations in rigorous validation, actionable interpretability, standardization, governance, and workflow integration. Findings Eighty-one Chinese experts reached consensus on 30 clinically practical statements covering AI applications from early detection to deployment. Critical relevance statement Recommendations highlight expert-supervised labeling, multicenter validation, subgroup evaluation, interpretable outputs, privacy-secured collaboration, integrated workflows, and post-implementation surveillance.</p>","PeriodicalId":13639,"journal":{"name":"Insights into Imaging","volume":"17 1","pages":""},"PeriodicalIF":5.6,"publicationDate":"2026-07-24","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC13400512/pdf/","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"148578623","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Multi-component radiological model based on intratumoral CT threshold segmentation for predicting visceral pleural invasion in lung adenocarcinoma ≤ 30 mm.","authors":"Yuanxin Sun, Jing Chen, Tingting Wang, Lu Zhang, Tingjia Xue, Weiqiu Jin, Hong Yu, Xiaodan Ye","doi":"10.1186/s13244-026-02351-z","DOIUrl":"10.1186/s13244-026-02351-z","url":null,"abstract":"<p><strong>Objectives: </strong>This retrospective study aims to investigate the value of intratumoral computed tomography (CT) threshold segmentation in radiomics, deep learning (DL), and radiomics-DL combined models for predicting visceral pleural invasion (VPI) in lung adenocarcinoma (LUAD) ≤ 30 mm.</p><p><strong>Materials and methods: </strong>Patients with invasive LUAD who underwent surgery and had preoperative thin-slice CT scans within four weeks were enrolled from two centers (n = 816). Patients from center 1 were divided into a training set (TS, n = 591) and an internal test set (ITS, n = 98) based on surgical time. Patients from center 2 constituted the external test set (ETS, n = 127). Solid, ground-glass, and peritumoral components were extracted using intratumoral CT threshold segmentation and peritumoral expansion methods. The radiomics model was a Random Forest Classifier; the DL model was a pre-trained Vision Transformer (ViT) fine-tuned on three components; the combined model integrated features from two pipelines. Model performance was evaluated by the area under the receiver operating characteristic curve (AUC), sensitivity, and specificity. Clinical utility and model interpretability were evaluated using decision curve analysis and SHapley Additive exPlanations (SHAP), respectively.</p><p><strong>Results: </strong>Radiomics, ViT, and radiomics-ViT models achieved AUCs of 0.865, 0.858, and 0.895 in TS; 0.865, 0.844, and 0.852 in ITS; and 0.844, 0.816, and 0.823 in ETS, respectively. Radiomics-ViT model achieved the highest sensitivity, with ViT features contributing the most.</p><p><strong>Conclusion: </strong>A hybrid multi-component feature pipeline could serve as a reliable and highly sensitive tool for VPI prediction in LUAD ≤ 30 mm.</p><p><strong>Critical relevance statement: </strong>The multi-component radiomics-ViT model achieved high sensitivity for VPI prediction in LUAD ≤ 30 mm, which is a promising tool for preoperative treatment design and prognostic assessment.</p><p><strong>Key points: </strong>CT attenuation-defined components remain underexplored in artificial intelligence (AI) models for VPI prediction. Solid, ground-glass, and peritumoral components enable AI models for VPI prediction. Multi-component inputs effectively capture tumor heterogeneity for VPI prediction in LUAD ≤ 30 mm.</p>","PeriodicalId":13639,"journal":{"name":"Insights into Imaging","volume":"17 1","pages":""},"PeriodicalIF":5.6,"publicationDate":"2026-07-24","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC13400552/pdf/","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"148584344","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Radiological pure-solid appearance clinical stage I lung adenocarcinoma: a comparative study of mucinous and non-mucinous adenocarcinoma based on imaging features and survival outcomes.","authors":"Wei Li, He Du, Shixing Wu, Jingyi Wang, Nan Song, Ziwei Wan, Fengying Wu, Jingyun Shi","doi":"10.1186/s13244-026-02353-x","DOIUrl":"10.1186/s13244-026-02353-x","url":null,"abstract":"<p><strong>Objectives: </strong>Our study aims to compare the clinicopathological, genetic, and prognostic differences between pure-solid appearance clinical stage I mucinous and non-mucinous adenocarcinomas.</p><p><strong>Materials and methods: </strong>Five hundred seventy-nine patients with radiological pure-solid appearance clinical stage I lung adenocarcinoma, were retrospectively evaluated. Clinicopathological, radiological, genetic characteristics, and recurrence patterns were compared between the non-mucinous adenocarcinoma and mucinous adenocarcinoma groups. Recurrence-free survival (RFS) and overall survival (OS) were assessed by Kaplan-Meier curves and compared between groups using log-rank tests. Cox proportional hazard models were used to analyze prognostic factors.</p><p><strong>Results: </strong>Of 579 patients (mean age, 59.9 years ± 9.7 [SD]; 298 women), 442 had non-mucinous adenocarcinoma, and 137 had mucinous adenocarcinoma. The non-mucinous adenocarcinoma group, in comparison with mucinous adenocarcinoma, showed higher CEA levels (p = 0.003), higher CT values (36.88 vs 20.89 HU, p < 0.001), a higher incidence of emphysema (25.1% vs 3.6%, p < 0.001), a lower clinical T stage (51.8% vs 42.3%, p < 0.001), and the absence of air bronchogram (61.3% vs 27%, p < 0.001). Recurrences occurred more frequently in the non-mucinous adenocarcinoma group (38.5%, 170/442) compared to the mucinous adenocarcinoma group (13.1%, 18/137; p < 0.001). Brain metastasis was significantly more common in the non-mucinous adenocarcinoma group (27.1% vs 0%; p = 0.024). Compared with the non-mucinous adenocarcinoma group, the mucinous adenocarcinoma group had better 5-year RFS (86.9% vs 62.7%; p < 0.001) and 5-year OS (97.1% vs 83.2%, p < 0.001).</p><p><strong>Conclusion: </strong>Imaging features, survival outcomes, and recurrence patterns differed significantly between mucinous and non-mucinous lung adenocarcinomas.</p><p><strong>Key points: </strong>Question For radiological pure-solid appearance clinical stage I lung adenocarcinoma, there remains controversy regarding the clinicopathological characteristics and prognosis of mucinous versus non-mucinous subtypes. Findings There were significant differences in clinicopathological features, imaging findings, and recurrence patterns between mucinous and non-mucinous adenocarcinomas. Compared with the non-mucinous adenocarcinoma, the mucinous group had better 5-year RFS and 5-year OS. Critical relevance statement Lung mucinous adenocarcinoma, Lung mucinous adenocarcinoma, due to its unique imaging and clinicopathological features, was a special independent subtype of adenocarcinoma. For radiological pure-solid appearance lung adenocarcinoma, mucinous adenocarcinoma had a better survival prognosis than non-mucinous adenocarcinoma.</p>","PeriodicalId":13639,"journal":{"name":"Insights into Imaging","volume":"17 1","pages":""},"PeriodicalIF":5.6,"publicationDate":"2026-07-23","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC13396331/pdf/","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"148578573","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Assessment of inter-reader agreement and diagnostic performance of Node-RADS in colon cancer.","authors":"Hui Wang, Bailing Dai, Na Feng, Hongxia Zhang, Xiaoxiang Ning, Kefeng Zhou, Guoqun Mao, Qiang Li, Yongyu An","doi":"10.1186/s13244-026-02349-7","DOIUrl":"10.1186/s13244-026-02349-7","url":null,"abstract":"<p><strong>Objectives: </strong>To assess the inter-reader agreement and diagnostic performance of Node-reporting and data system (Node-RADS) in the evaluation of colon cancer among radiologists with varied experience levels.</p><p><strong>Materials and methods: </strong>A total of 247 patients with colon cancer were included in this retrospective study. Lymph node evaluations were performed by six radiologists with diverse experience levels (three junior and three senior readers) using Node-RADS. Inter-reader agreement for Node-RADS classification and specific imaging features was assessed using the kappa statistic and compared between junior and senior readers. The diagnostic performance of Node-RADS was evaluated at a cutoff of 3 and 4 score.</p><p><strong>Results: </strong>Lymph node metastasis was observed in 47.8% of patients. The overall inter-reader agreement was substantial for Node-RADS classification (k = 0.69), almost perfect for size (k = 0.83), moderate for texture and border (k = 0.41-0.45), and fair for shape (k = 0.37). Senior readers demonstrated higher inter-reader agreement in both Node-RADS and specific features than junior readers. Senior readers achieved higher AUC (0.79-0.80) than junior readers (0.72-0.78). At a cutoff of 3, sensitivity, specificity, and accuracy across all readers were 52.5-70.3%, 84.5-93.0%, and 69.2-79.8%, respectively. The corresponding values were 33.9-44.9%, 92.3-98.4%, and 64.4-72.5%, respectively, at a cutoff of 4. Senior readers showed significantly higher pooled sensitivity and accuracy, with comparable specificity compared to junior readers at both Node-RADS cutoffs.</p><p><strong>Conclusion: </strong>The Node-RADS demonstrates high agreement between observers. Senior readers showed higher inter-reader agreement and diagnostic performance than junior readers. Node-RADS exhibits high specificity but relatively low sensitivity for detecting lymph node metastasis in colon cancer.</p><p><strong>Critical relevance statement: </strong>This study demonstrated good inter-observer agreement for Node-reporting and data system (Node-RADS) in colon cancer across readers of varying experience, supporting its clinical reliability. However, its limited sensitivity for predicting lymph node metastasis highlights the need for further improvement.</p><p><strong>Key points: </strong>Assessing inter-observer agreement among different readers for Node-RADS assignment is critical before its clinical application. Substantial inter-observer agreement for Node-RADS was observed across readers of varying experience levels.</p>","PeriodicalId":13639,"journal":{"name":"Insights into Imaging","volume":"17 1","pages":""},"PeriodicalIF":5.6,"publicationDate":"2026-07-22","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC13391975/pdf/","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"148561981","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Temporal changes in breast cancer lifetime risk and implications for personalised surveillance in high-risk women.","authors":"Machteld Keupers, Sam Nijssen, Willem Sarkol, Yao-Kuan Wang, Lesley Cockmartin, Hans Wildiers, Hilde Bosmans, Chantal Van Ongeval","doi":"10.1186/s13244-026-02355-9","DOIUrl":"10.1186/s13244-026-02355-9","url":null,"abstract":"<p><strong>Objectives: </strong>To evaluate temporal changes in breast cancer lifetime risk (BC-LTR) over a 10-year period among women with clinically elevated risk but without known pathogenic genetic mutations, using the International Breast Cancer Intervention Study (IBIS) breast cancer risk evaluation tool, and to assess implications for personalised surveillance strategies.</p><p><strong>Materials and methods: </strong>In this retrospective study, women referred for elevated risk assessment in 2014 underwent BC-LTR estimation using IBIS (v8.0b). Risk was recalculated in 2024 using the same model without recalibration. Women who developed breast cancer, underwent bilateral mastectomy, died during follow-up, were older than 85 years in 2024, or carried pathogenic genetic mutations were excluded, as IBIS BC-LTR recalculation is not applicable in these contexts. Risk factor temporal changes were evaluated using paired statistics. A linear mixed model assessed predictors of IBIS BC-LTR and their association with temporal changes.</p><p><strong>Results: </strong>Among 362 eligible women, mean IBIS BC-LTR decreased from 23% in 2014 to 19% in 2024 (mean absolute change = 6.5, p < 0.001). Risk group reclassification occurred in 44%: 36% shifted to a lower risk category and 8% to a higher category. Increase in IBIS BC-LTR was associated with increasing density (p < 0.0001), higher BMI (p = 0.002), current HRT use (p = 0.007), and additional affected relatives (p < 0.03), whereas downward reclassification was mainly linked to decreasing density (p < 0.0001).</p><p><strong>Conclusion: </strong>IBIS BC-LTR estimates change over time in women with elevated clinical risk. Periodic reassessment may better align estimated risk with tailored screening strategies, supporting more precise, individualised surveillance.</p><p><strong>Critical relevance statement: </strong>Variation in IBIS BC-LTR scores over time highlights the need for periodic recalculation to improve individual risk stratification and guide more appropriate, personalised imaging and screening pathways.</p><p><strong>Key points: </strong>Breast cancer risk models identify women most likely to benefit from supplemental screening. IBIS BC-LTR estimates changed over 10 years, reclassifying over 40% of women with elevated baseline risk into a different risk category. Periodic IBIS BC-LTR recalculation is essential to optimise risk stratification and screening pathways.</p>","PeriodicalId":13639,"journal":{"name":"Insights into Imaging","volume":"17 1","pages":""},"PeriodicalIF":5.6,"publicationDate":"2026-07-22","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC13391461/pdf/","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"148561992","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Multidimensional deep learning for grading and prognostic assessment of intrahepatic mass-forming cholangiocarcinoma.","authors":"Liyong Zhuo, Wenjing Chen, Zijun Song, Lihong Xing, Xiaomeng Li, Jiawei Hao, Zimei Yang, Xuechun Wang, Caiying Li, Jianing Wang, Xiaoping Yin","doi":"10.1186/s13244-026-02350-0","DOIUrl":"10.1186/s13244-026-02350-0","url":null,"abstract":"<p><strong>Objective: </strong>To develop an interpretable magnetic resonance imaging (MRI)-based framework for preoperative histologic grading of intrahepatic mass-forming cholangiocarcinoma (IMCC) and exploratory prognostic stratification.</p><p><strong>Materials and methods: </strong>A retrospective analysis was conducted on preoperative MRI from 333 IMCC patients across three centers (training cohort, n = 240; external validation cohort, n = 93). An ensemble deep learning (DL) framework synergizing 2.5D and 3D ResNet-50 architectures was constructed. Significant variables from clinical-laboratory-imaging (ClinLabImag) features, radiomics, and DL outputs were integrated into a combined model. Discrimination was assessed using the area under the receiver operating characteristic curve (AUC). Model interpretability was evaluated with Gradient-weighted Class Activation Mapping (Grad-CAM) and SHapley Additive exPlanations (SHAP), and the Kaplan-Meier method was used to compare overall survival (OS) between risk groups.</p><p><strong>Results: </strong>The DL model achieved an external validation AUC of 0.804 (95% confidence interval (CI): 0.712-0.896), significantly outperforming standalone 2.5D (p = 0.025) and 3D architectures (p = 0.030). The Combined model (AUC: 0.843 [95% CI, 0.749-0.938]) showed better external validation performance than the Radiomics (p = 0.019) and ClinLabImag models (p = 0.017), with only modest, non-significant improvement over the DL model (p = 0.355). SHAP analysis showed that DL features contributed most to model predictions. The Combined model showed exploratory OS differences between risk groups.</p><p><strong>Conclusion: </strong>An interpretable multidimensional MRI-based DL framework supports noninvasive preoperative grading in IMCC and provides exploratory prognostic information.</p><p><strong>Critical relevance statement: </strong>This study critically evaluates an interpretable multidimensional MRI-based deep learning framework for preoperative grading and exploratory prognostic stratification of intrahepatic mass-forming cholangiocarcinoma, supporting individualized radiologic risk assessment before treatment.</p><p><strong>Key points: </strong>Reliable noninvasive MRI biomarkers are needed for preoperative grading and prognostic assessment of intrahepatic mass-forming cholangiocarcinoma to guide individualized treatment planning. The combined multidimensional MRI model achieved favorable external validation performance and showed exploratory overall survival differences between IMCC risk groups.</p>","PeriodicalId":13639,"journal":{"name":"Insights into Imaging","volume":"17 1","pages":""},"PeriodicalIF":5.6,"publicationDate":"2026-07-22","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC13391990/pdf/","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"148548696","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Spectral CT parameters of perivascular adipose tissue as non-invasive biomarkers for identifying symptomatic carotid atherosclerosis.","authors":"Xiaohan Zheng, Xuehuan Liu, Jinyu Song, Endong Zhao, Weiwei Cui, Gouling Zhan, Zuoxi Li, Hong Yu, Xiao Gao, Jun Liu","doi":"10.1186/s13244-026-02354-w","DOIUrl":"10.1186/s13244-026-02354-w","url":null,"abstract":"<p><strong>Objectives: </strong>To evaluate the association between computed tomography (CT) spectral parameters of perivascular adipose tissue (PVAT) and carotid plaque composition, and to assess their value for identifying symptomatic plaques.</p><p><strong>Materials and methods: </strong>In this study, 306 consecutive patients with computed tomography angiography (CTA)-confirmed carotid atherosclerosis who underwent head and neck spectral CT angiography were analyzed. Quantitative plaque parameters and PVAT spectral metrics were extracted. Correlations between quantitative plaque parameters and PVAT spectral indices were analyzed, and logistic regression was used to identify independent predictors of symptomatic plaques. Diagnostic performance was evaluated by generating receiver operating characteristic (ROC) curves.</p><p><strong>Results: </strong>Compared with asymptomatic patients, symptomatic carotid plaques showed higher PVAT effective atomic number (Zeff) and iodine concentration (IC) but lower fat fraction (FF) (all p < 0.05). Zeff and IC correlated positively with fibrous fatty and necrotic core volumes, whereas FF correlated negatively with necrotic core volume (all p < 0.001). In multivariable models, Zeff, IC, FF, and virtual monoenergetic image attenuation at 70 keV (CT_70keV) were independently associated with symptomatic plaques (all p < 0.05). CT_70keV and FF performed best for identifying symptomatic plaques (AUC 0.857 and 0.820). The AUC increased from 0.716 for stenosis severity alone to 0.821 after the addition of necrotic core volume, and further to 0.897 and 0.916 after the addition of FF and CT_70keV, respectively.</p><p><strong>Conclusion: </strong>Spectral CT-derived PVAT parameters differ between symptomatic and asymptomatic carotid atherosclerosis and are associated with plaque composition, providing complementary noninvasive information for identifying symptomatic plaques.</p><p><strong>Key points: </strong>Question: Can spectral CT-derived PVAT parameters improve noninvasive identification of symptomatic carotid atherosclerosis beyond luminal stenosis?</p><p><strong>Findings: </strong>PVAT spectral parameters differed between symptomatic and asymptomatic plaques, and adding FF or CT_70keV improved diagnostic performance. Critical relevance statement: spectral parameters of carotid PVAT differ significantly between symptomatic and asymptomatic carotid atherosclerosis and are associated with plaque components.</p>","PeriodicalId":13639,"journal":{"name":"Insights into Imaging","volume":"17 1","pages":""},"PeriodicalIF":5.6,"publicationDate":"2026-07-22","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC13391981/pdf/","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"148561984","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Indeterminate CT reporting of adult acute appendicitis: radiologist versus lexicon-based diagnostic certainty.","authors":"Piyachai Siriphiphatcharoen, Rathachai Kaewlai, Sasima Tongsai, Jitti Chatpuwaphat, Shanigarn Thiravit, Papasorn Wattanakul, Patcharaporn Thaisuriyo, Napakadol Noppakunsomboon","doi":"10.1186/s13244-026-02352-y","DOIUrl":"10.1186/s13244-026-02352-y","url":null,"abstract":"<p><strong>Objectives: </strong>To evaluate report indeterminacy, diagnostic performance, and agreement with final diagnosis in CT performed for suspected adult acute appendicitis (AA), comparing original radiologist reports, adjudicated impressions, and certainty derived from structured lexicon-based schemes. We also aimed to identify factors associated with indeterminate reporting.</p><p><strong>Materials and methods: </strong>A total of 983 consecutive CT reports between August 2021 and July 2025 were reviewed. Diagnostic certainty was categorized from original radiology reports (negative, indeterminate, positive), with indeterminate cases adjudicated by two radiologists. Two objective lexicon-based schemes were applied using extracted CT findings (appendiceal diameter (6-mm and 7-mm cutoffs), wall thickening, wall hyperenhancement, and periappendiceal fat stranding) to generate four certainty categories (negative, possible, probable, positive). Surgical, pathological, and clinical outcomes served as reference standards.</p><p><strong>Results: </strong>Acute appendicitis was confirmed in 393 patients (40%). Indeterminate reporting occurred in 15.5% of original reports, 12% of adjudicated reports, and 27.9-46.1% of lexicon-based schemes. Across all approaches, AA prevalence increased consistently with higher diagnostic certainty. Adjudicated reports demonstrated the highest diagnostic performance (sensitivity 85.2-98.7%, specificity 85.9-97.1%, accuracy 91.0-92.4%) and excellent agreement with final diagnosis (kappa = 0.819-0.838). Several CT features were independently associated with indeterminate reporting.</p><p><strong>Conclusions: </strong>Radiologist-adjudicated impressions conveyed diagnostic certainty more effectively than original reports and lexicon-based schemes, with lower indeterminacy and stronger agreement with the final diagnosis. These findings underscore the practical importance of careful wording and judicious use of indeterminate categories when communicating CT findings for suspected adult AA.</p><p><strong>Critical relevance statement: </strong>Careful wording of CT impressions, grounded in radiologists' integrated judgment, can reduce diagnostic ambiguity and potentially improve clinical decision-making, supporting standardized languages to express diagnostic certainty in suspected adult acute appendicitis.</p><p><strong>Key points: </strong>CT reports for suspected appendicitis vary in wording, leading to indeterminate impressions, potentially affecting diagnostic performance and agreement with final diagnosis. Adjudicated radiologist impressions reduced indeterminate reports and achieved higher accuracy and agreement with final diagnosis than original reports or lexicon-based schemes.</p>","PeriodicalId":13639,"journal":{"name":"Insights into Imaging","volume":"17 1","pages":""},"PeriodicalIF":5.6,"publicationDate":"2026-07-22","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC13391980/pdf/","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"148562006","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Diagnostic performance of an artificial intelligence algorithm for detecting pneumoperitoneum on abdominal CT scans.","authors":"Yuwan Hu, Zhigang Sun, Haoyu Li, Shuya Gao, Haoran Du, Tongyin Zhang, Sheng Xie, Peng Wu, Pengbo Jiang, Dijia Wu, Xuemeng Wu, Hongliang Sun","doi":"10.1186/s13244-026-02348-8","DOIUrl":"10.1186/s13244-026-02348-8","url":null,"abstract":"<p><strong>Objectives: </strong>This study aims to evaluate the diagnostic performance of an artificial intelligence (AI) algorithm for detection, segmentation, and volumetric quantification of pneumoperitoneum on abdominal CT scans.</p><p><strong>Materials and methods: </strong>We developed and validated a deep learning-based model for automated pneumoperitoneum detection on CT. Multi-center CT imaging series from 2072 patients were collected and randomly divided into training and testing sets at an approximate 7:3 ratio. The external validation set included 214 emergency CT scans collected between April 2022 and December 2024. Diagnostic reports served as the reference standard. Primary outcome included the area under the curve (AUC), sensitivity, specificity, accuracy, positive predictive value (PPV), and negative predictive value (NPV). Quantitative agreement between AI and reference volumes was assessed using the intraclass correlation coefficient (ICC).</p><p><strong>Results: </strong>In the test set (n = 607), the model demonstrated excellent performance: sensitivity 91.4%, specificity 93.1%, and AUC 0.97 (95% CI: 0.95-0.99). In the external validation cohort (n = 214), the model maintained robust performance with sensitivity 84.3% (95% CI: 76.2-90.5%), specificity 89.6% (95% CI: 82.3-94.6%), accuracy 86.9% (81.6-91.2%), PPV 89.2% (95% CI: 81.8-94.3%), and NPV 84.8% (95% CI: 77.1-90.7%). After excluding cases with minimal free gas (<1 mL), the model's sensitivity improved to 96%. AI-derived volumes showed strong agreement with the reference standard (ICC 0.996, 95% CI: 0.994-0.997).</p><p><strong>Conclusion: </strong>The AI model attained high diagnostic accuracy for pneumoperitoneum on abdominal CT scans, promising to expedite emergency workflows.</p><p><strong>Key points: </strong>Question Reliable AI detection of pneumoperitoneum, particularly for small-volume free air, on emergency CT remains an unmet need for rapid and accurate emergency triage. Findings The AI model shows high sensitivity and specificity for clinically relevant pneumoperitoneum volumes, although trace-volume detection on CT scans remains challenging. Critical relevance statement This study evaluates the diagnostic performance and volume-dependent variability of an AI model for pneumoperitoneum detection on CT scans, with the potential to aid emergency radiology workflow prioritization and decision support, though prospective studies are needed to confirm clinical impact.</p>","PeriodicalId":13639,"journal":{"name":"Insights into Imaging","volume":"17 1","pages":""},"PeriodicalIF":5.6,"publicationDate":"2026-07-18","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC13380604/pdf/","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"148497527","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Correction: Toward a standardized and interoperable imaging biomarker catalog.","authors":"Pablo Rodríguez-Belenguer, Paula Doria-Borrell, Ángel Alberich-Bayarri, Raquel Perez-Lopez, Vicky Goh, Konstantin Nikolaou, Aad van der Lugt, Marion Smits, Luis Rodríguez-Rodríguez, Leonor Cerdá-Alberich, Luis Martí-Bonmatí","doi":"10.1186/s13244-026-02347-9","DOIUrl":"10.1186/s13244-026-02347-9","url":null,"abstract":"","PeriodicalId":13639,"journal":{"name":"Insights into Imaging","volume":"17 1","pages":""},"PeriodicalIF":5.6,"publicationDate":"2026-07-16","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC13376274/pdf/","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"148470420","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}