Insights into Imaging最新文献

{"title":"MRI analysis of distal tibiofibular joint and ankle anatomy to assess lateral ankle sprain risk.","authors":"João Vieira, Ana Catarina Vieira, Alberto Vieira","doi":"10.1186/s13244-025-02102-6","DOIUrl":"https://doi.org/10.1186/s13244-025-02102-6","url":null,"abstract":"<p><strong>Objective: </strong>To evaluate the risk of lateral ankle sprain (LAS) related to bone anatomical variations of the distal tibiofibular syndesmosis (DTS) and the height of both malleolar articular surfaces.</p><p><strong>Materials and methods: </strong>This retrospective cohort study included patients undergoing evaluation and assessment of quantitative parameters of the DTS and height of both malleolar articular surfaces in ankle MRI. Of the 216 patients included, 116 suffered LAS (53.7%). The measurements of the DTS were: anterior facet length of the fibular notch (a), posterior facet length of the fibular notch (b), angle between the anterior and posterior facets (c), fibular notch depth (d), tibial thickness (e), and fibular thickness (f). A subjective morphological analysis of the DTS was also assessed. The measurements of the malleolar articular surface length were: medial malleolar articular surface height (h), lateral malleolar articular surface height (i), and width of the talar dome articular surface (j).</p><p><strong>Results: </strong>Evaluating the DTS, patients with LAS showed a greater c (p < 0.001), a higher a/b (p = 0.013), and a lower d/e (p < 0.001). A plane DTS was also found to be a risk factor for sprain. Additionally, patients with LAS exhibited a lower i/j (p = 0.003). Indeed, the values of c, a/b, and i/j were independent predictors of LAS (p < 0.001, p = 0.015, p = 0.011).</p><p><strong>Conclusion: </strong>Anatomical factors of the DTS and lateral malleolus articular surface were predictors of the presence of LAS, mainly the angle and ratio between the anterior and the posterior facets and the ratio between the lateral malleolar height and the width of the talar articular surface.</p><p><strong>Critical relevance statement: </strong>Lateral ankle sprains are one of the most common musculoskeletal injuries, and the predisposing anatomical factors are not clear. This study correlates anatomical variants of the distal tibiofibular syndesmosis and the height of both malleoli with lateral ankle sprain.</p><p><strong>Key points: </strong>What specific bony variations in the distal tibiofibular syndesmosis and malleolar-talus relationship may predispose individuals to lateral ankle sprains? A shallow fibular notch, flat-type syndesmosis, and a lower lateral malleolar height/talar width ratio are more often found in first lateral ankle sprains. Understanding these anatomical variations may aid in injury prevention and improve risk assessment.</p>","PeriodicalId":13639,"journal":{"name":"Insights into Imaging","volume":"16 1","pages":"213"},"PeriodicalIF":4.5,"publicationDate":"2025-10-04","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145225322","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":"3D T1 turbo spin echo improves detection of gadolinium-enhancing multiple-sclerosis lesions.","authors":"Pablo Naval-Baudin, William F Bermúdez Bravo, Vanessa I Pineda-Borja, Pablo Arroyo-Pereiro, Ignacio Martínez-Zalacaín, Lucía Romero-Pinel, Paloma Mora, Nahum Calvo, Antonio Martínez-Yélamos, Sergio Martínez-Yélamos, Mònica Cos, Albert Pons-Escoda, Carles Majós","doi":"10.1186/s13244-025-02093-4","DOIUrl":"https://doi.org/10.1186/s13244-025-02093-4","url":null,"abstract":"<p><strong>Objectives: </strong>To compare the performance of 3D T1 turbo spin echo (3DT1TSE) and 3D T1 turbo field echo (3DT1TFE) MRI in detecting gadolinium-enhancing lesions in multiple sclerosis (MS).</p><p><strong>Materials and methods: </strong>We retrospectively analyzed 255 3-T MRIs from MS patients, each including post-contrast 3DT1TSE and 3DT1TFE sequences. Two blinded readers independently assessed enhancing lesions per sequence. A consensus review, incorporating longitudinal imaging and additional sequences, served as the reference standard.</p><p><strong>Results: </strong>The consensus identified 70 enhancing lesions in 31 patients. All 70 were visible on 3DT1TSE, while 64 (91%) were detectable on 3DT1TFE. Reader sensitivity was higher for 3DT1TSE (84% and 90%) than 3DT1TFE (45% and 40%) (p < 0.01). Inter-reader agreement was excellent for 3DT1TSE (ICC = 0.90) and moderate for 3DT1TFE (intraclass correlation coefficient = 0.69). Although false positives were more common with 3DT1TSE, they were readily excluded during consensus reading. In six patients, enhancing lesions were detected only on 3DT1TSE, with treatment escalation in two.</p><p><strong>Conclusion: </strong>3DT1TSE outperformed 3DT1TFE in sensitivity and reader agreement for enhancing lesion detection in MS. Incorporating 3DT1TSE into standard MRI protocols may improve disease activity assessment and clinical decision-making.</p><p><strong>Critical relevance statement: </strong>Replacing 3D gradient-echo with post-contrast 3D T1 turbo spin-echo brain MRI greatly improves the detection of gadolinium-enhancing multiple-sclerosis lesions, boosting diagnostic sensitivity and reader agreement and directly influencing treatment-escalation decisions in routine practice.</p><p><strong>Key points: </strong>Detecting and enhancing MS lesions is limited by standard 3D T1 turbo field echo (3DT1TFE) MRI. 3D T1 turbo spin echo detects significantly more gadolinium-enhancing MS lesions than conventional 3DT1TFE. Greater lesion detection allows more precise activity assessment and optimal treatment management.</p>","PeriodicalId":13639,"journal":{"name":"Insights into Imaging","volume":"16 1","pages":"212"},"PeriodicalIF":4.5,"publicationDate":"2025-10-03","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145225293","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":"Machine learning combined with CT-based radiomics predicts the prognosis of oesophageal squamous cell carcinoma.","authors":"Mingyu Liu, Rongxin Lu, Bo Wang, Jun Fan, Yuheng Wang, Jiashan Zhu, Jinhua Luo","doi":"10.1186/s13244-025-02049-8","DOIUrl":"10.1186/s13244-025-02049-8","url":null,"abstract":"<p><strong>Objectives: </strong>This retrospective study aims to develop a machine learning model integrating preoperative CT radiomics and clinicopathological data to predict 3-year recurrence and recurrence patterns in postoperative oesophageal squamous cell carcinoma.</p><p><strong>Materials and methods: </strong>Tumour regions were segmented using 3D-Slicer, and radiomic features were extracted via Python. LASSO regression selected prognostic features for model integration. Clinicopathological data include tumour length, lymph node positivity, differentiation grade, and neurovascular infiltration. Ultimately, a machine learning model was established by combining the screened imaging feature data and clinicopathological data and validating model performance. A nomogram was constructed for survival prediction, and risk stratification was carried out through the prediction results of the machine learning model and the nomogram. Survival analysis was performed for stage-based patient subgroups across risk stratifications to identify adjuvant therapy-benefiting cohorts.</p><p><strong>Results: </strong>Patients were randomly divided into a 7:3 ratio of 368 patients in the training cohorts and 158 patients in the validation cohorts. The LASSO regression screens out 6 recurrence prediction and 9 recurrence pattern prediction features, respectively. Among 526 patients (mean age 63; 427 males), the model achieved high accuracy in predicting recurrence (training cohort AUC: 0.826 [logistic regression]/0.820 [SVM]; validation cohort: 0.830/0.825) and recurrence patterns (training:0.801/0.799; validation:0.806/0.798). Risk stratification based on a machine learning model and nomogram predictions revealed that adjuvant therapy significantly improved disease-free survival in stages II-III patients with predicted recurrence and low survival (HR 0.372, 95% CI: 0.206-0.669; p < 0.001).</p><p><strong>Conclusion: </strong>Machine learning models exhibit excellent performance in predicting recurrence after surgery for squamous oesophageal cancer.</p><p><strong>Critical relevance statement: </strong>Radiomic features of contrast-enhanced CT imaging can predict the prognosis of patients with oesophageal squamous cell carcinoma, which in turn can help clinicians stratify risk and screen out patient populations that could benefit from adjuvant therapy, thereby aiding medical decision-making.</p><p><strong>Key points: </strong>There is a lack of prognostic models for oesophageal squamous cell carcinoma in current research. The prognostic prediction model that we have developed has high accuracy by combining radiomics features and clinicopathologic data. This model aids in risk stratification of patients and aids clinical decision-making through predictive outcomes.</p>","PeriodicalId":13639,"journal":{"name":"Insights into Imaging","volume":"16 1","pages":"211"},"PeriodicalIF":4.5,"publicationDate":"2025-10-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC12488548/pdf/","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145199320","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":"Representation of women authorship in the top-cited articles published in the medical imaging literature.","authors":"Hyung Jin Lee, Dae Young Yoon, Sora Baek, Kyoung Ja Lim, Young Lan Seo, Eun Joo Yun","doi":"10.1186/s13244-025-02085-4","DOIUrl":"10.1186/s13244-025-02085-4","url":null,"abstract":"<p><strong>Objective: </strong>To investigate the representation of women among the authors of top-cited articles published in the medical imaging literature.</p><p><strong>Materials and methods: </strong>This retrospective bibliometric study queried the Web of Science database to identify the top-cited articles (citation number ≥ 300) in the medical imaging literature. The gender of the first and last (senior) authors was determined based on online databases. The year of publication, country of origin, document type, and subspecialty for each article were also collected. We analyzed the proportion of women authors and the relationships between author gender and article characteristics.</p><p><strong>Results: </strong>Among 596 top-cited articles, women accounted for 132 (22.1%) of first authors and 84 (14.1%) of last authors. Women as last authors were more likely to publish with women first authors compared to male first authors (odds ratio: 1.35). Women's first authorship was significantly more frequent in articles from South Korea (44.4%; phi = 0.095) and in radiation oncology (38.1%; phi = 0.106) and significantly less frequent in articles from France (0.0%; phi = -0.102). Women's last authorship was significantly more frequent in articles from the Netherlands (30.6%; phi = 0.120), in breast (38.9%; phi = 0.126), and in radiation oncology (28.6%; phi = 0.115), and significantly less frequent in nuclear medicine (4.3%; phi = -0.083).</p><p><strong>Conclusion: </strong>Women authors remain underrepresented in top-cited articles published in the medical imaging literature, with country of origin and subspecialty identified as factors of influence.</p><p><strong>Critical relevance statement: </strong>Women are still underrepresented among the authors of the top-cited articles in the medical imaging literature. The findings highlight the gender disparities in the highest academic achievement in this biomedical field and provide valuable insight into this ongoing issue.</p><p><strong>Key points: </strong>Women authors remain underrepresented in top-cited articles in the medical imaging literature. Women accounted for 22.1% of first authors and 14.1% of last authors. There were variations in the proportion of women authors between countries and subspecialties.</p>","PeriodicalId":13639,"journal":{"name":"Insights into Imaging","volume":"16 1","pages":"210"},"PeriodicalIF":4.5,"publicationDate":"2025-09-27","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC12476343/pdf/","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145175095","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":"Intratumoral heterogeneity score enhances invasiveness prediction in pulmonary ground-glass nodules via stacking ensemble machine learning.","authors":"Zhichao Zuo, Ying Zeng, Jinqiu Deng, Shanyue Lin, Wanyin Qi, Xiaohong Fan, Yujie Feng","doi":"10.1186/s13244-025-02097-0","DOIUrl":"10.1186/s13244-025-02097-0","url":null,"abstract":"<p><strong>Objectives: </strong>The preoperative differentiation of adenocarcinomas in situ, minimally invasive adenocarcinoma, and invasive adenocarcinoma using computed tomography (CT) is crucial for guiding clinical management decisions. However, accurately classifying ground-glass nodules poses a significant challenge. Incorporating quantitative intratumoral heterogeneity scores may improve the accuracy of this ternary classification.</p><p><strong>Materials and methods: </strong>In this multicenter retrospective study, we developed ternary classification models by leveraging insights from both base and stacking ensemble machine learning models, incorporating intratumoral heterogeneity scores along with clinical-radiological features to distinguish adenocarcinomas in situ, minimally invasive adenocarcinoma, and invasive adenocarcinoma. The machine learning models were trained, and final model selection depended on maximizing the macro-average area under the curve (macro-AUC) in both the internal and external validation sets.</p><p><strong>Results: </strong>Data from 802 patients from three centers were divided into a training set (n = 477) and an internal test set (n = 205), in a 7:3 ratio, with an additional external validation set comprising 120 patients. The stacking classifier exhibited superior performance relative to the other models, achieving macro-AUC values of 0.7850 and 0.7717 for the internal and external validation sets, respectively. Moreover, an interpretability analysis utilizing the Shapley Additive Explanation identified four key features of this ternary classification: intratumoral heterogeneity score, nodule size, nodule type, and age.</p><p><strong>Conclusion: </strong>The stacking classifier, recognized as the optimal algorithm for integrating the intratumoral heterogeneity score and clinical-radiological features, effectively served as a ternary classification model for assessing the invasiveness of lung adenocarcinoma in chest CT images.</p><p><strong>Critical relevance statement: </strong>Our stacking classifier integrated intratumoral heterogeneity scores and clinical-radiological features to improve the ternary classification of lung adenocarcinoma invasiveness (adenocarcinomas in situ/minimally invasive adenocarcinoma/invasive adenocarcinoma), aiding in precise diagnosis and clinical decision-making for ground-glass nodules.</p><p><strong>Key points: </strong>The intratumoral heterogeneity score effectively assessed the invasiveness of lung adenocarcinoma. The stacking classifier outperformed other methods for this ternary classification task. Intratumoral heterogeneity score, nodule size, nodule type, and age predict invasiveness.</p>","PeriodicalId":13639,"journal":{"name":"Insights into Imaging","volume":"16 1","pages":"209"},"PeriodicalIF":4.5,"publicationDate":"2025-09-26","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC12474818/pdf/","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145175027","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":"Image-guided biopsy of breast lesions-when to use what biopsy technique.","authors":"Wendelien B G Sanderink, Julia Camps-Herrero, Alexandra Athanasiou, Henrique L Couto, Kirti Mehta, Popat Palak Bhavesh Thakkar, Pooja Jagmohan, Sara E Vázquez-Manjarrez, Seigo Nakamura, Jelle Wesseling, Ritse M Mann","doi":"10.1186/s13244-025-02084-5","DOIUrl":"10.1186/s13244-025-02084-5","url":null,"abstract":"<p><p>In recent years, minimally invasive diagnostic options for breast lesions have expanded, but consensus on optimal biopsy techniques and imaging combinations remains lacking. This study, driven by an adapted RAND-UCLA Appropriateness Method and insights from eight experts in breast biopsy from across the world, aims to create consensus for selecting biopsy techniques. Highlighted findings suggest Vacuum-Assisted Biopsy (VAB) for lesions visible exclusively at mammography/tomosynthesis (with or without contrast enhancement) or MRI. Core-needle biopsy (CNB) takes precedence for masses over 5 mm visible under US. The selection of other biopsy techniques during US-guided procedures depends on lesion type, size, and sampling indication. VAB is preferred for smaller masses (< 5 mm), complex cystic and solid lesions with small solid parts, small intraductal masses, architectural distortions, and calcifications visible on US. In re-biopsy scenarios for inconclusive findings or high-risk lesions, the panel suggests two VAB extensions: Extended Vacuum-Assisted Biopsy (EVAB) for unambiguous lesion classification and Vacuum-Assisted Excision (VAE) for complete lesion removal. Furthermore, the panel provides detailed input on how to handle specific cases, such as re-biopsy for lobular neoplasia, flat epithelial atypia and atypical ductal hyperplasia. Surgical excision is advised for DCIS and benign or borderline phyllodes tumors found through initial CNB or VAB. In conclusion, an international expert group formulated recommendations on diagnostic breast biopsies under image guidance, aiming to ensure accurate diagnosis worldwide by providing practical advice on needle selection and biopsy approach. KEY POINTS: Evidence-based literature on the preferred biopsy technique and imaging combination for the diagnosis of breast lesions is sparse, and a general consensus is not available. The selection of biopsy technique for different image-guided procedures depends on lesion type, size, and sampling indication. This international expert panel consensus statement addresses standard approaches for varying biopsy indications.</p>","PeriodicalId":13639,"journal":{"name":"Insights into Imaging","volume":"16 1","pages":"208"},"PeriodicalIF":4.5,"publicationDate":"2025-09-25","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC12463797/pdf/","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145137361","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":"Preoperative prediction of microvascular invasion in pancreatic neuroendocrine tumors through analysis of portal venous phase CT images.","authors":"Hai-Yan Chen, Yao Pan, Yu-Wei Li, Li-Ting Shi, Jie-Yu Chen, Yun-Ying Liu, Ri-Sheng Yu, Lei Shi","doi":"10.1186/s13244-025-02091-6","DOIUrl":"10.1186/s13244-025-02091-6","url":null,"abstract":"<p><strong>Objectives: </strong>To evaluate clinical and CT imaging features on portal venous-phase for predicting microvascular invasion (MVI) in patients with pancreatic neuroendocrine tumors (PNETs) and compare survival outcomes.</p><p><strong>Materials and methods: </strong>In this retrospective study, 160 patients (training group) and 28 (validation group) who underwent surgical resection for PNETs were included. Demographic data and CT features were collected. The independent predictive factors for predicting MVI were confirmed through univariate and multivariate logistic regression analyses. The predictive performance was assessed by employing the receiver operating characteristic curve for predicting MVI. An R/shiny app based on logistic regression was developed. A Kaplan-Meier survival analysis with a log-rank test was conducted.</p><p><strong>Results: </strong>In the training group, invasion of surrounding tissues (odds ratio [OR]: 4.12), absolute enhancement (OR: 0.84), and relative enhancement ratio (OR: 16.1) were identified as independent predictors for predicting MVI in PNET patients, with an area under the curve of 0.819 and 0.891 in the training and validation groups, respectively. We have successfully developed a user-friendly web-based R/shiny app for real-time prediction of MVI in patients with PNETs. The median overall survival for patients with MVI was 12 months, compared to 37.5 months for those without MVI (log-rank p = 0.034).</p><p><strong>Conclusions: </strong>Imaging features from portal venous-phase CT images can be used to accurately predict the presence of MVI in patients with PNETs. Patients with MVI are associated with worse survival compared to those without MVI. The web-based R/shiny app for predicting MVI provides real-time data-driven estimates of predictive value to facilitate informed decision-making.</p><p><strong>Critical relevance statement: </strong>Imaging features can accurately predict MVI in patients with PNETs, and the web-based R/shiny app provides real-time, data-driven estimates to enhance decision-making, thereby streamlining clinical practice.</p><p><strong>Key points: </strong>The presence of microvascular invasion (MVI) in patients was associated with worse survival. Surrounding tissue invasion and absolute/relative enhancement ratio were identified as independent predictors for MVI. This web-based app predicts MVI and provides real-time data-driven estimates of predictive value.</p>","PeriodicalId":13639,"journal":{"name":"Insights into Imaging","volume":"16 1","pages":"206"},"PeriodicalIF":4.5,"publicationDate":"2025-09-25","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC12463792/pdf/","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145137325","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":"Segmentation-model-based framework to detect aortic dissection on non-contrast CT images: a retrospective study.","authors":"Qidong Wang, Shan Huang, Weifeng Pan, Zhan Feng, Lei Lv, Dongwei Guan, Zhiwen Yang, Yimin Huang, Wei Liu, Weiwei Shui, Mingliang Ying, Wenbo Xiao","doi":"10.1186/s13244-025-02098-z","DOIUrl":"10.1186/s13244-025-02098-z","url":null,"abstract":"<p><strong>Objectives: </strong>To develop an automated deep learning framework for detecting aortic dissection (AD) and visualizing its morphology and extent on non-contrast CT (NCCT) images.</p><p><strong>Materials and methods: </strong>This retrospective study included patients who underwent aortic CTA from January 2021 to January 2023 at two tertiary hospitals. Demographic data, medical history, and CT scans were collected. A segmentation-based deep learning model was trained to identify true and false lumens on NCCT images, with performance evaluated on internal and external test sets. Segmentation accuracy was measured using the Dice coefficient, while the intraclass correlation coefficient (ICC) assessed consistency between predicted and ground-truth false lumen volumes. Receiver operating characteristic (ROC) analysis evaluated the model's predictive performance.</p><p><strong>Results: </strong>Among 701 patients (median age, 53 years, IQR: 41-64, 486 males), data from Center 1 were split into training (439 cases: 318 non-AD, 121 AD) and internal test sets (106 cases: 77 non-AD, 29 AD) (8:2 ratio), while Center 2 served as the external test set (156 cases: 80 non-AD, 76 AD). The ICC for false lumen volume was 0.823 (95% CI: 0.750-0.880) internally and 0.823 (95% CI: 0.760-0.870) externally. The model achieved an AUC of 0.935 (95% CI: 0.894-0.968) in the external test set, with an optimal cutoff of 7649 mm³ yielding 88.2% sensitivity, 91.3% specificity, and 89.0% negative predictive value.</p><p><strong>Conclusions: </strong>The proposed deep learning framework accurately detects AD on NCCT and effectively visualizes its morphological features, demonstrating strong clinical potential.</p><p><strong>Critical relevance statement: </strong>This deep learning framework helps reduce the misdiagnosis of AD in emergencies with limited time. The satisfactory results of presenting true/false lumen on NCCT images benefit patients with contrast media contraindications and promote treatment decisions.</p><p><strong>Key points: </strong>False lumen volume was used as an indicator for AD. NCCT detects AD via this segmentation model. This framework enhances AD diagnosis in emergencies, reducing unnecessary contrast use.</p>","PeriodicalId":13639,"journal":{"name":"Insights into Imaging","volume":"16 1","pages":"207"},"PeriodicalIF":4.5,"publicationDate":"2025-09-25","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC12463791/pdf/","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145137342","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 risk stratification for predicting overall survival in pancreatic ductal adenocarcinoma.","authors":"Haitao Sun, Jianbo Li, Lei Li, Peng Wang, Qiying Tang, Tianyu Lu, Jun Han, Zongyu Xie, Yiheng Zhou, Kai Liu, Mengsu Zeng, Minping Hong, Yaolin Xu, Jianjun Zhou","doi":"10.1186/s13244-025-02088-1","DOIUrl":"10.1186/s13244-025-02088-1","url":null,"abstract":"<p><strong>Purpose: </strong>To develop and validate a magnetic resonance imaging (MRI)-based pancreatic risk stratification model (M-PRiSM) for prognosis prediction and therapy guidance in pancreatic ductal adenocarcinoma (PDAC).</p><p><strong>Materials and methods: </strong>In this retrospective multicenter study, a total of 540 patients with PDAC were enrolled. A Cox proportional hazards model was used to identify significant clinical-radiological predictors and develop an M-PRiSM risk score system. The performance of the model was assessed using Harrell's C-index, time-dependent area under the curve (AUC), and calibration curves, and was compared with existing clinical prediction metrics, including the 8^th American Joint Committee on Cancer (AJCC) staging and CA19-9 levels.</p><p><strong>Results: </strong>540 PDAC patients were split into a development cohort (n = 282) and an internal validation cohort (n = 122), with 136 patients forming external cohort A (n = 80) and B (n = 56). The model, incorporating CA19-9 (hazard ratio (HR), 1.604, p = 0.047), margin (HR, 1.918, p = 0.001), tumor size (HR, 1.308, p = 0.003), and venous enhancement ratio (VER) (HR, 0.605, p = 0.047), was developed, showing superior predictive accuracy for OS with C-indexes of 0.73, 0.70, and 0.70, and 0.68 across four cohorts. The model outperformed the 8th AJCC staging and CA19-9 alone across validation cohorts (p < 0.001). High-risk patients exhibit the worst overall survival (OS) and a higher frequency of adverse pathological features, but benefited significantly from adjuvant therapy, with improved survival outcomes.</p><p><strong>Conclusion: </strong>The M-PRiSM model effectively predicts postoperative OS and identifies PDAC patients likely to benefit from adjuvant treatment.</p><p><strong>Critical relevance statement: </strong>The M-PRiSM model, utilizing MRI and clinical features, enables preoperative risk stratification for overall survival in pancreatic ductal adenocarcinoma, offering a generalizable tool to guide personalized adjuvant therapy and enhance prognostic assessment.</p><p><strong>Key points: </strong>The M-PRiSM model can effectively predict pancreatic ductal adenocarcinoma survival. The M-PRiSM model showed strong generalizability in external validations. The M-PRiSM model stratifies patients for guiding adjuvant therapy for improved outcomes.</p>","PeriodicalId":13639,"journal":{"name":"Insights into Imaging","volume":"16 1","pages":"205"},"PeriodicalIF":4.5,"publicationDate":"2025-09-24","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC12460861/pdf/","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145130677","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":"3D gadolinium-enhanced high-resolution near-isotropic pancreatic imaging at 3.0-T MR using deep-learning reconstruction.","authors":"Sylvie Guan, Julie Poujol, Elodie Gouhier, Caroline Touloupas, Alexandre Delpla, Isabelle Boulay-Coletta, Marc Zins","doi":"10.1186/s13244-025-02066-7","DOIUrl":"10.1186/s13244-025-02066-7","url":null,"abstract":"<p><strong>Objectives: </strong>To compare overall image quality, lesion conspicuity and detectability on 3D-T1w-GRE arterial phase high-resolution MR images with deep learning reconstruction (3D-DLR) against standard-of-care reconstruction (SOC-Recon) in patients with suspected pancreatic disease.</p><p><strong>Materials and methods: </strong>Patients who underwent a pancreatic MR exam with a high-resolution 3D-T1w-GRE arterial phase acquisition on a 3.0-T MR system between December 2021 and June 2022 in our center were retrospectively included. A new deep learning-based reconstruction algorithm (3D-DLR) was used to additionally reconstruct arterial phase images. Two radiologists blinded to the reconstruction type assessed images for image quality, artifacts and lesion conspicuity using a Likert scale and counted the lesions. Signal-to-noise ratio and lesion contrast-to-noise ratio were calculated for each reconstruction. Quantitative data were evaluated using paired t-tests. Ordinal data such as image quality, artifacts and lesions conspicuity were analyzed using paired-Wilcoxon tests. Interobserver agreement for image quality and artifact assessment was evaluated using Cohen's kappa.</p><p><strong>Results: </strong>Thirty-two patients (mean age 62 years ± 12, 16 female) were included. 3D-DLR significantly improved SNR for each pancreatic segment and lesion CNR compared to SOC-Recon (p < 0.01), and demonstrated significantly higher average image quality score (3.34 vs 2.68, p < 0.01). 3D DLR also significantly reduced artifacts compared to SOC-Recon (p < 0.01) for one radiologist. 3D-DLR exhibited significantly higher average lesion conspicuity (2.30 vs 1.85, p < 0.01). The sensitivity was increased with 3D-DLR compared to SOC-Recon for both reader 1 and reader 2 (1 vs 0.88 and 0.88 vs 0.83, p = 0.62 for both results).</p><p><strong>Conclusion: </strong>3D-DLR images demonstrated higher overall image quality, leading to better lesion conspicuity.</p><p><strong>Critical relevance statement: </strong>3D deep learning reconstruction can be applied to gadolinium-enhanced pancreatic 3D-T1w arterial phase high-resolution images without additional acquisition time to further improve image quality and lesion conspicuity.</p><p><strong>Key points: </strong>3D DLR has not yet been applied to pancreatic MRI high-resolution sequences. This method improves SNR, CNR, and overall 3D T1w arterial pancreatic image quality. Enhanced lesion conspicuity may improve pancreatic lesion detectability.</p>","PeriodicalId":13639,"journal":{"name":"Insights into Imaging","volume":"16 1","pages":"204"},"PeriodicalIF":4.5,"publicationDate":"2025-09-24","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC12460215/pdf/","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145130711","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}