Insights into Imaging最新文献

{"title":"Contrast-enhanced mammography in the setting of neoadjuvant therapy from a multidisciplinary perspective.","authors":"Lígia Pires-Gonçalves, Conceição Leal, António Guimarães-Santos, Ana Teresa Aguiar, Miguel Abreu, Donzília Brito, Rui Manuel Henrique","doi":"10.1186/s13244-026-02294-5","DOIUrl":"10.1186/s13244-026-02294-5","url":null,"abstract":"<p><p>Contrast-enhanced mammography (CEM) is a morphofunctional imaging technique that is gaining interest and acceptance in clinical practice. With increasing use of neoadjuvant therapy (NAT) in breast cancer and of precision medicine, the need for equitable access to appropriate high-quality advanced imaging for monitoring response has become ever more relevant. CEM is an imaging alternative for NAT monitoring, with encouraging initial results and a good logistical profile. In this educational review from a practical perspective, we explore key aspects for response assessment of NAT and CEM roles in this setting, based on relevant literature and the authors' clinical experience. Pearls and pitfalls, as well as opportunities and challenges for NAT monitoring with CEM, will be approached from the perspectives of the radiologist, the oncologist, the surgeon, the pathologist, and the patient. CRITICAL RELEVANCE STATEMENT: The early experience with CEM in the NAT setting is positive, encouraging its further development, not only from the radiologist's view but also from a broad multidisciplinary perspective. KEY POINTS: CEM seems to be non-inferior to MRI for response evaluation and preoperative re-staging after NAT. CEM may predict and monitor in vivo chemosensitivity during NAT. CEM may allow a greater patient flow rate compared to MRI.</p>","PeriodicalId":13639,"journal":{"name":"Insights into Imaging","volume":"17 1","pages":""},"PeriodicalIF":5.6,"publicationDate":"2026-07-15","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC13373129/pdf/","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"148446320","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":"Development of an expert-annotated chest X-ray dataset to support AI validation in tuberculosis diagnosis.","authors":"Wiwatana Tanomkiat, Shiva Raj Timsina, Thammasin Ingviya, Sitthichok Chaichulee, Patinya Yutchawit, Chiawchan Thuncharoenkanka, Wuttikrai Pannuan, Chureerat Chantharat, Urasri Kerdmeesup, Nantaka Kiranantawat, Sitang Nirattisaikul, Sutarat Tungsagunwattana, Krisna Dissaneevate, Warawut Sukkasem, Chanavee Toh, Suthinee Teerajaruwat","doi":"10.1186/s13244-026-02334-0","DOIUrl":"10.1186/s13244-026-02334-0","url":null,"abstract":"<p><strong>Objectives: </strong>To assess inter-rater agreement and diagnostic performance of six United States National Institute for Occupational Safety and Health-certified B readers (physicians certified in standardized chest X-ray (CXR) classification for pneumoconiosis) in diagnosing tuberculosis (TB) on CXR, with the goal of developing a reliable dataset to support external validation of artificial intelligence (AI) models for TB detection.</p><p><strong>Materials and methods: </strong>1097 CXRs from five institutions were analyzed by six B readers. Patients aged ≥ 15 years were included, excluding those with HIV or opportunistic infections. CXRs were classified as unremarkable or abnormal. Abnormalities were categorized using a modified International Labor Organization classification and classified as consistent or inconsistent with TB. Microbiological references included sputum smears, cultures, or molecular tests. Descriptive statistics summarized the CXR and microbiological findings. Inter-rater agreement was assessed with Fleiss' kappa. Diagnostic performance was evaluated by comparing CXR findings to microbiological references.</p><p><strong>Results: </strong>Of the 3117 readings, 69% of CXRs were abnormal, and 31% were unremarkable. Microbiological results confirmed that 87% of abnormal CXRs were TB cases, while 83% of unremarkable CXRs were non-TB. Inter-rater agreements were κ = 0.83 for all findings, κ = 0.67 for findings consistent with TB, and κ = 0.76 for active TB findings. For findings consistent with TB, the sensitivity, specificity, and accuracy ranged from 77.2% to 91.1%, 87.4% to 98.6%, and 84.1% to 90.1%, respectively.</p><p><strong>Conclusion: </strong>B readers exhibited strong agreement and high accuracy in diagnosing TB on CXR, providing a robust dataset that could be used for external validation of AI models in TB diagnosis.</p><p><strong>Critical relevance statement: </strong>With substantial inter-rater agreement and strong diagnostic performance, B readers' interpretations of CXRs provide a reliable dataset to support external validation of AI models for TB diagnosis.</p><p><strong>Key points: </strong>B readers can provide a reliable reference for validating AI models in TB diagnosis. Less variability and high accuracy were observed among B readers in CXR-based TB diagnosis.</p>","PeriodicalId":13639,"journal":{"name":"Insights into Imaging","volume":"17 1","pages":""},"PeriodicalIF":5.6,"publicationDate":"2026-07-10","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC13354736/pdf/","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"148420846","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":"Integrating MRI-based tumour staging within the TNM classification system in modern prostate cancer management.","authors":"Georgios Agrotis, Andrei Purysko, Anwar Padhani, Fredrik Jäderling, Geert Villeirs, Samuel J Withey, Maarten de Rooij, Tristan Barrett, Ivo G Schoots","doi":"10.1186/s13244-026-02346-w","DOIUrl":"10.1186/s13244-026-02346-w","url":null,"abstract":"<p><p>MRI has transformed prostate cancer diagnosis, risk stratification, and treatment planning. However, the tumour-node-metastasis (TNM) classification continues to rely exclusively on digital rectal examination (DRE) for clinical T-staging, despite DRE's limited diagnostic accuracy and prognostic performance. This mismatch between modern MRI practice and outdated staging criteria undermines prognostication and treatment decision-making. In this review, we synthesise current evidence on DRE- and MRI-based local staging, discuss the limitations of the current TNM framework, and present a proposal for parallel reporting of MRI-based T-staging (mrT) alongside clinical T-staging (cT). We further outline the rationale for developing MRI-derived prognostic groups, benchmarked against pathology and long-term oncologic outcomes, and present a framework intended to support future evidence-based revisions of the official TNM classification. CRITICAL RELEVANCE STATEMENT: Clinical adoption of prostate MRI, TNM T-staging still relies on digital rectal examination, a method with limited sensitivity and prognostic value. This discordance leads to stage misclassification and suboptimal risk stratification. Parallel reporting of MRI-based T-staging and development of MRI-derived prognostic groups are essential to align staging with contemporary imaging practice, support evidence-based treatment decisions, and inform future revisions of the TNM classification system. KEY POINTS: Digital rectal examination (DRE) is inferior to MRI for detecting extraprostatic extension. MRI-based and DRE-based T-staging should be reported separately to evaluate stage migration, as MRI-based T-staging may shift patients into higher stages without reflecting true biological risk. Parallel reporting of MRI-based T-staging and development of MRI-derived prognostic groups are essential to align staging with contemporary imaging practice, support evidence-based treatment decisions, and inform future revisions of the TNM classification system.</p>","PeriodicalId":13639,"journal":{"name":"Insights into Imaging","volume":"17 1","pages":""},"PeriodicalIF":5.6,"publicationDate":"2026-07-09","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC13350576/pdf/","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"148420835","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":"Evaluation of the objective structured clinical examination in the assessment of nuclear medicine residency training.","authors":"Lijuan Di, Jianhua Zhang, Guangyu Zhao, Yan Fan","doi":"10.1186/s13244-026-02340-2","DOIUrl":"10.1186/s13244-026-02340-2","url":null,"abstract":"<p><strong>Objectives: </strong>We evaluated the capacity of the Objective Structured Clinical Examination (OSCE) to certify resident competency in nuclear medicine (NM).</p><p><strong>Materials and methods: </strong>A retrospective analysis was conducted on the OSCE scores of residents who participated in the examination for the certification of NM physicians in Beijing from 2019 to 2024, with a focus on difficulty, discrimination, reliability, and validity.</p><p><strong>Results: </strong>The OSCE is comprised of six stations, including high-radioactivity room operations, imaging room operations, PET/CT case analysis, SPECT/CT case analysis, patient interviews and report writing. The mean total OSCE scores (mean ± SD) were 88.58 ± 2.37 points (2019), 89.13 ± 2.72 points (2020), 86.15 ± 2.75 points (2021), 92.96 ± 2.31 points (2022), 91.11 ± 1.37 points (2023), and 89.08 ± 2.75 points (2024), respectively. The overall difficulty coefficients of the OSCE were 0.89, 0.89, 0.86, 0.88, 0.91, and 0.89, respectively. The internal consistency reliability values (Cronbach's α) were 0.66, 0.76, 0.57, 0.28, 0.05, and 0.69, respectively, across the years. The station-level discrimination indices ranged from 0.46 to 0.83. Among the 30 stations that were double-scored by two examiners, 73.33% (22/30) of the stations demonstrated excellent interrater consistency.</p><p><strong>Conclusion: </strong>The OSCE effectively simulates clinical environments and provides a comprehensive assessment of the clinical competence of NM residents.</p><p><strong>Critical relevance statement: </strong>This study conducted a multidimensional evaluation of 6-year OSCE data, focusing on difficulty, discrimination, reliability, and validity, demonstrating its effectiveness in simulating clinical environments and assessing the clinical competence of NM residents.</p><p><strong>Key points: </strong>Compared to the numerous studies on medical students, relatively few studies on the OSCE target NM residents. The OSCE exhibited acceptable station difficulty and discrimination, moderate reliability, and satisfactory content validity for assessing NM residents.</p>","PeriodicalId":13639,"journal":{"name":"Insights into Imaging","volume":"17 1","pages":""},"PeriodicalIF":5.6,"publicationDate":"2026-07-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC13323698/pdf/","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"148368273","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":"ECR 2026 Book of Abstracts.","authors":"","doi":"10.1186/s13244-026-02345-x","DOIUrl":"https://doi.org/10.1186/s13244-026-02345-x","url":null,"abstract":"","PeriodicalId":13639,"journal":{"name":"Insights into Imaging","volume":"17 Suppl 1","pages":""},"PeriodicalIF":5.6,"publicationDate":"2026-07-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"148360625","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":"Slow-flow, high-impact: radiologic features in venous malformations of the female genital tract.","authors":"Maëlle Guihard, Antoine Fraissenon, Sophie Collardeau-Frachon, Pierre-Paul Bringuier, Pierre-Adrien Bolze, Laurent Guibaud, Pascal Rousset","doi":"10.1186/s13244-026-02343-z","DOIUrl":"10.1186/s13244-026-02343-z","url":null,"abstract":"<p><p>Gynaecological venous malformations are rare, congenital slow-flow vascular anomalies arising from early developmental errors. They may occur as isolated lesions or within syndromic contexts (associated with somatic gene variants such as PIK3CA and TEK). They remain poorly documented and frequently misclassified due to persistent outdated terminology and diagnostic challenges-requiring integration of clinical, radiological, and histopathological data. Imaging plays a pivotal role in the diagnosis of gynaecological venous malformations: ultrasound serves as the first-line imaging modality, while Magnetic Resonance Imaging offers superior anatomical detail, confirms the diagnosis, and guides treatment planning and follow-up. However, the lack of standardised protocols and consensual imaging features complicates differentiation from other gynaecological vascular conditions with overlapping radiological appearances. Given diagnostic challenges and concerns regarding fertility, a multidisciplinary approach in specialised vascular anomaly centres combining clinical, radiological, histological, and genetic data is essential. This review aims to provide a complete sonographic and MRI framework aligned with current nomenclature, with key clinical and histopathological insights to support radiologists in their diagnostic process. CRITICAL RELEVANCE STATEMENT: Gynaecological venous malformations are frequently overlooked or misdiagnosed due to shared imaging features and a lack of standardised protocols, making accurate imaging description and multidisciplinary assessment essential for a positive diagnosis and appropriate management in dedicated centres. KEY POINTS: Gynaecological venous malformations are rare and often misdiagnosed congenital anomalies. Diagnostic features include an organ enlargement with a sponge-like appearance and slow-flow.</p>","PeriodicalId":13639,"journal":{"name":"Insights into Imaging","volume":"17 1","pages":""},"PeriodicalIF":5.6,"publicationDate":"2026-06-30","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC13319296/pdf/","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"148360683","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":"Assessing the reliability of Response Evaluation Criteria In Solid Tumors (RECIST): a systematic review of the factors contributing to inter-observer variability.","authors":"Illaa Smesseim, Teresa T Bucho, Renaud Tissier, Stevie van der Mierden, Ferry Lalezari, Jacobus A Burgers, Stefano Trebeschi","doi":"10.1186/s13244-026-02320-6","DOIUrl":"10.1186/s13244-026-02320-6","url":null,"abstract":"<p><strong>Background: </strong>The Response Evaluation Criteria in Solid Tumors (RECIST) have been the standard for assessing tumor response in oncology trials since 2000. Despite the use of these criteria, interobserver variability (IOV) remains a significant concern, as it can affect patient management and clinical trial outcomes. We aimed to review the current literature on factors contributing to IOV in RECIST.</p><p><strong>Materials and methods: </strong>We conducted a systematic review to summarize potential factors that can contribute to variability, variance, or reproducibility in RECIST assessments: (I) manual measurements, (II) selection of lesions, (III) 1D diameters, (IV) intra-radiologist variability, (V) the experience of the readers, (VI) local vs centralized assessment, (VII) variations of criteria used, (VIII) type of imaging, (IX) follow-up schedule. Inclusion criteria required studies to use RECIST (1.0, 1.1, or modified versions), involve multiple response evaluations and provide comparative data on the same test subjects.</p><p><strong>Results: </strong>We identified 246 studies, of which 88 met our inclusion criteria. Median sample size across studies was 50 patients (IQR: 25.8-90.8). Most studies (68.2%) used CT scans as the sole imaging modality. The most common primary tumors studied were lung and hepatocellular carcinoma (each 14.8%). Most studies showed an IOV being affected by measurement methods, lesion selection, and imaging quality. Statistical methods for IOV vary: Kappa for categorical and intraclass correlation for continuous data, with calculation differences limiting comparability and risking misleading conclusions.</p><p><strong>Conclusion: </strong>Standardized protocols are needed to study interobserver variability in RECIST and enable us to understand the current criteria and develop better ones.</p><p><strong>Key points: </strong>We should standardize the assessment of interobserver variability in RECIST to improve reproducibility, comparability, and clinical trial decision reliability. Interobserver variability in RECIST depends on measurement methods, lesion selection, and imaging quality. Inconsistent statistics hinder study comparability. Standardized RECIST protocols can reduce variability, improving evaluation consistency, clinical decisions, and treatment outcomes in oncology trials.</p>","PeriodicalId":13639,"journal":{"name":"Insights into Imaging","volume":"17 1","pages":""},"PeriodicalIF":5.6,"publicationDate":"2026-06-30","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC13319625/pdf/","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"148360709","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":"How to optimise breast cancer staging with contrast-enhanced mammography: current evidence and clinical implications.","authors":"Chiara Bellini, Giulia Bicchierai, Diego de Benedetto, Federica di Naro, Chiara Maiello, Sofia Vidali, Paolina Tonelli, Ermanno Vanzi, Cecilia Boeri, Vittorio Miele, Jacopo Nori","doi":"10.1186/s13244-026-02331-3","DOIUrl":"10.1186/s13244-026-02331-3","url":null,"abstract":"<p><p>Contrast-enhanced mammography (CEM) is a functional breast imaging technique that combines conventional mammographic morphology with contrast-based assessment of tumour vascularity. In the presurgical setting, CEM has gained interest as a tool to improve local staging by refining tumour extent evaluation and detecting additional disease beyond conventional imaging. This narrative review critically summarises current evidence on the role of CEM in preoperative breast cancer staging. Technical aspects, diagnostic performance for index lesion size assessment and detection of additional malignant foci, comparison with breast MRI, and impact on surgical planning are discussed, with attention to histological subtypes and clinically relevant scenarios encountered in daily practice. Available studies suggest that CEM provides tumour extent assessment comparable to MRI in selected settings, with higher specificity for additional lesions and meaningful influence on surgical decision-making. However, limitations related to contrast administration, radiation dose, evaluation of posterior structures, and nodal staging remain. Heterogeneity in protocols and outcome definitions also limits generalizability. CEM represents a valuable adjunct for presurgical breast cancer staging when appropriately integrated into multimodality workflows. Prospective multicenter studies are needed to define its impact on surgical outcomes and clarify its role within personalised breast cancer management. CRITICAL RELEVANCE STATEMENT: This review provides a critical overview of CEM in preoperative breast cancer staging, highlighting its role as a complementary imaging tool when MRI is unavailable or unsuitable, and emphasising the need for prospective multicentre studies to validate its clinical impact. KEY POINTS: To define the role of CEM in preoperative breast cancer staging, focusing on tumour extent, detection of additional disease, and implications for surgical planning. CEM improves tumour extent assessment and detection of additional lesions, showing performance similar to MRI in selected clinical scenarios and influencing surgical planning in approximately 20-30% of patients.</p>","PeriodicalId":13639,"journal":{"name":"Insights into Imaging","volume":"17 1","pages":""},"PeriodicalIF":5.6,"publicationDate":"2026-06-29","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC13315364/pdf/","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"148345110","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":"MRI-based quantification of intratumoral heterogeneity for predicting recurrence risk in ER+/HER2- breast cancer.","authors":"Yang Chen, Jie Shi, Jing Chen, Lizhi Xie, Jin Ye, Wei Tang, Qin Xiao, Yan Huang, Yajia Gu, Weijun Peng","doi":"10.1186/s13244-026-02342-0","DOIUrl":"10.1186/s13244-026-02342-0","url":null,"abstract":"<p><strong>Objectives: </strong>The need for a cost-effective, rapid, and increasingly accessible alternative to the 21-gene assay prompted this study, which developed a novel MRI-based intratumoral heterogeneity score (ITHscore) to quantify tumor heterogeneity and integrated it with radiomic and clinical features to predict the 21-gene recurrence score (RS).</p><p><strong>Materials and methods: </strong>This retrospective study included ER+/HER2- breast cancer patients who underwent 21-gene assay and preoperative MRI at our institution (April 2017-March 2019). Patients were randomly split into training (70%) and internal test (30%) cohorts, with an external test cohort from the public Duke-Breast-Cancer-MRI dataset. Tumor volumes of interest were automatically segmented using a pre-trained Scalable and Transferable U-Net framework, followed by k-means clustering to compute the ITHscore. Predictive models for the RS were built using clinical, radiomics, and ITHscore features with the support vector machine method, and evaluated by receiver operating characteristic curves.</p><p><strong>Results: </strong>The institutional dataset comprised 452 patients (training: 316 (187 high-risk, 129 low-risk); internal test: 136 (80 high-risk, 56 low-risk)), while the external Duke cohort included 230 patients (44 high-risk, 186 low-risk). The ITHscore was significantly elevated in high-risk patients (p < 0.001), and its incorporation into the clinical-radiomic model improved RS prediction, yielding AUCs of 0.86 for the internal test cohort and 0.82 for the external test cohort.</p><p><strong>Conclusions: </strong>In this exploratory study, the ITHscore, which held promise as a noninvasive and intuitive means of characterizing intratumoral heterogeneity, demonstrated a potential incremental value for predicting RS in patients with ER+/HER2- breast cancer.</p><p><strong>Critical relevance statement: </strong>The MRI-derived quantification of intratumoral heterogeneity facilitates simple and quantitative assessment of tumor heterogeneity and demonstrates potential incremental value in providing a rapid, cost-effective, and accessible prediction of the recurrence score in ER+/HER2- breast cancer.</p><p><strong>Key points: </strong>There is a need for a cost-effective, rapid, and increasingly accessible alternative to the 21-gene assay for predicting recurrence risk in ER+/HER2- breast cancer. The intratumoral heterogeneity score demonstrated potential as a noninvasive, intuitive, and quantitative biomarker for characterizing intratumoral heterogeneity and predicting recurrence score.</p>","PeriodicalId":13639,"journal":{"name":"Insights into Imaging","volume":"17 1","pages":""},"PeriodicalIF":5.6,"publicationDate":"2026-06-27","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC13310210/pdf/","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"148338476","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":"A perioperative multi-modal fusion and deep learning-based prognostic system for upper tract urothelial carcinoma: a multi-institutional study.","authors":"Xiang Peng, Yang Li, Wei Shi, Bangxin Xiao, Xiao Xiao, Xiaofeng Yue, Qiao Xv, Qing Jiang, Weiyang He, Yingjie Xv, Mingzhao Xiao","doi":"10.1186/s13244-026-02337-x","DOIUrl":"10.1186/s13244-026-02337-x","url":null,"abstract":"<p><strong>Objectives: </strong>Precise perioperative risk stratification for upper tract urothelial carcinoma (UTUC) is essential. We developed a multimodal prognostic model integrating perioperative clinical data, radiomics, and deep learning (DL) features from baseline CT urography to improve survival prediction and guide adjuvant management.</p><p><strong>Materials and methods: </strong>We retrospectively enrolled 623 patients from six institutions, divided into training, internal validation, and independent external validation sets. Four single-modal models (clinical, radiomics, 2D DL, and 2.5D DL) were developed, and an integrated combined model was constructed by fusing their prognostic scores. Performance was evaluated using the C-index, area under the curve (AUC), calibration curves, and decision curve analysis (DCA).</p><p><strong>Results: </strong>The combined model consistently outperformed all single-modal models across all cohorts. C-indices reached 0.758 (95% CI: 0.712-0.804), 0.725 (95% CI: 0.651-0.798), and 0.704 (95% CI: 0.631-0.777) in the training, internal validation, and external validation sets, respectively, numerically surpassing the best single-modal models. Notably, our 2.5D DL model (C-index: 0.705) demonstrated a consistent incremental improvement over the 2D DL model (C-index: 0.681) in capturing prognostic information. In external validation, the combined model achieved a 3-year AUC of 0.766. DCA indicated the comprehensive model exhibited excellent calibration and provided the highest net benefits.</p><p><strong>Conclusion: </strong>This multimodal system, featuring a robust 2.5D DL strategy, improves overall survival prediction in UTUC. It offers a valuable tool for accurate perioperative risk stratification immediately after radical nephroureterectomy, demonstrating particularly reliable value for 3-year intermediate-term clinical decision-making.</p><p><strong>Critical relevance statement: </strong>This multimodal system advances clinical radiology by fusing perioperative clinical data, radiomics, and DL features from CTU images, enhancing risk stratification accuracy to guide postoperative adjuvant management for UTUC.</p><p><strong>Key points: </strong>Current prognostic models for UTUC lack accuracy, creating an unmet clinical need for precise perioperative risk stratification to guide adjuvant management. A multimodal prognostic model fusing clinical, radiomic, and DL features from baseline CT urography consistently outperformed single-modality models in predicting overall survival.</p>","PeriodicalId":13639,"journal":{"name":"Insights into Imaging","volume":"17 1","pages":""},"PeriodicalIF":5.6,"publicationDate":"2026-06-25","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC13294415/pdf/","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"148317048","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}