European Journal of Radiology最新文献

{"title":"Potential values of uEXPLORER total-body PET/CT in nasopharyngeal carcinoma: dynamic imaging and kinetic parameters","authors":"Ying-Ying Hu ,&nbsp;Zhen-Chong Yang ,&nbsp;Li-Ping Liang ,&nbsp;Jia-Yi Shen ,&nbsp;Yu-Mo Zhao ,&nbsp;Jie-Yi Lin ,&nbsp;Wei-Guang Zhang ,&nbsp;Jia-Tai Feng ,&nbsp;Xiao-Fei Lv ,&nbsp;Guo-Dong Jia ,&nbsp;Shan-Shan Guo ,&nbsp;Li-Ting Liu ,&nbsp;Jing-Yi Wang ,&nbsp;Qiu-Yan Chen ,&nbsp;Wei Fan ,&nbsp;Lin-Quan Tang","doi":"10.1016/j.ejrad.2025.112422","DOIUrl":"10.1016/j.ejrad.2025.112422","url":null,"abstract":"<div><h3>Objectives</h3><div>To explore the dynamic imaging and kinetic parameters of total-body positron-emission tomography/computed tomography (TB-PET/CT) in nasopharyngeal carcinoma (NPC).</div></div><div><h3>Methods</h3><div>Fifty-six NPC patients were prospectively enrolled and underwent pretreatment 1-hour dynamic TB-PET/CT (uEXPLORER) acquisition. Standardized uptake values (SUVs) and Patlak-based quantitative kinetic parameters, including glucose metabolic rate (Ki), glucose transport rate of flow-in and flow-out between blood and tissues (K₁ and k₂), and tissue glucosamine-6-phosphatization rate (k₃) were assessed. The suspected metastatic small lymph nodes with mild-to-moderate [¹⁸F] FDG uptake and those important for the N staging of tumor underwent ultrasound-guided biopsy, and the lesion-to-background ratio (LBR) in TB-PET images and K_i images were compared.</div></div><div><h3>Results</h3><div>In total, 97 frame of SUVmean images were collected in each patient. The time‒activity curves (TAC) of primary lesions showed a continuous upward trend. The SUVmean dispersion among primary lesions increased gradually. SUV_max correlated strongly with K_i and k₃ (r = 0.903 and 0.660, P = 0.001 and &lt; 0.001, respectively) in primary lesions. Out of the 34 cervical lymph nodes biopsied using ultrasound, 10 were confirmed as metastases, and 24 involved inflammations. SUV_max, K_i, k_2, and k₃ were significantly different (p &lt; 0.0001, p &lt; 0.01, p &lt; 0.01, and p &lt; 0.0001 respectively) between metastatic lymph nodes and inflamed lymph nodes, while K₁ had no significant difference. K_i images had higher LBR than PET images in metastatic lymph nodes conspicuity (p &lt; 0.05).</div></div><div><h3>Conclusion</h3><div>TB-PET/CT uncovered tumor heterogeneity and metabolic characteristics via [¹⁸F] FDG uptake during the first hour after injection. Kinetic parameters and K_i images may help in discriminating between malignant and inflammatory lymph nodes in Nasopharyngeal Carcinoma.</div></div>","PeriodicalId":12063,"journal":{"name":"European Journal of Radiology","volume":"193 ","pages":"Article 112422"},"PeriodicalIF":3.3,"publicationDate":"2025-09-16","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145097679","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Multimodal radiomic analysis to determine high-Consistency prognostic phenotypes associated with epidermal growth factor receptor mutations in non-small cell lung cancer brain metastases","authors":"Che-Yu Hsu ,&nbsp;Hsin-Han Tsai ,&nbsp;Ting-Li Chen ,&nbsp;Chih-Hsin Yang ,&nbsp;Kao-Lang Liu ,&nbsp;Sung-Hsin Kuo ,&nbsp;Feng-Ming Hsu ,&nbsp;Wei-Wu Chen ,&nbsp;Wei-Hsun Hsu ,&nbsp;Weichung Wang","doi":"10.1016/j.ejrad.2025.112420","DOIUrl":"10.1016/j.ejrad.2025.112420","url":null,"abstract":"<div><h3>Introduction</h3><div>Limited linkage between epidermal growth factor receptor (EGFR) mutations and recurrence–predictive radiomic signatures restricts the application of radiomics–guided therapy for brain metastases (BMs) from non–small–cell lung cancer (NSCLC). This study aimed to establish an EGFR-associated radiomic signature (EGFR-RS), compare its consistency with that of conventional whole radiomic features-based radiomic signature (WF-RS), and evaluate its efficacy in predicting local recurrence for BMs treated with radiosurgery.</div></div><div><h3>Methods</h3><div>Brain magnetic resonance (MR) and computed tomography (CT) images of NSCLC patients with BMs undergoing radiosurgery between 2008 and 2020 were examined. The least absolute shrinkage and selection operator was utilized to select features and develop signatures. Discriminative abilities were assessed using the area under the curve, while univariable and multivariable competing risk regression determined predictors and established a clinical-radiomic model.</div></div><div><h3>Results</h3><div>In total, 318 patients with 759 BMs were enrolled. The EGFR-RS, incorporating 11 MR and six CT EGFR-associated prognostic radiomic features, displayed better consistency, and superior predictive performance than the WF-RS, with C-indices of 0.746 (95 %CI 0.616, 0.876) in the test cohort, compared with 0.655 (95 %CI 0.527, 0.784) for the WF-RS. Multivariable analysis indicated EGFR-RS as the sole significant predictor of local recurrence in both the discovery and test sets (P &lt; 0.001, hazard ratio [HR] = 2.75; and P = 0.01, HR = 2.13, respectively). The clinical-radiomic model (EGFR-RS + <em>EGFR</em> mutation status + BM size) outperformed the clinical model in identifying high-risk lesions with local recurrence (discovery: P &lt; 0.001; HR = 4.54; test: P = 0.002; HR = 5.1).</div></div><div><h3>Conclusion</h3><div>The multimodal EGFR-RS, demonstrating better consistency than the WF-RS, effectively predicted the local recurrence of NSCLC BMs.</div></div>","PeriodicalId":12063,"journal":{"name":"European Journal of Radiology","volume":"193 ","pages":"Article 112420"},"PeriodicalIF":3.3,"publicationDate":"2025-09-15","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145156449","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Deep learning based multi-shot breast diffusion MRI: Improving imaging quality and reduced distortion","authors":"Ning Chien ,&nbsp;Yi-Hsuan Cho ,&nbsp;Ming-Yang Wang ,&nbsp;Li-Wei Tsai ,&nbsp;Cheng-Ya Yeh ,&nbsp;Chia-Wei Li ,&nbsp;Patricia Lan ,&nbsp;Xinzeng Wang ,&nbsp;Kao-Lang Liu ,&nbsp;Yeun-Chung Chang","doi":"10.1016/j.ejrad.2025.112419","DOIUrl":"10.1016/j.ejrad.2025.112419","url":null,"abstract":"<div><h3>Objective</h3><div>To investigate the imaging performance of deep-learning reconstruction on multiplexed sensitivity encoding (MUSE DL) compared to single-shot diffusion-weighted imaging (SS-DWI) in the breast.</div></div><div><h3>Materials and Methods</h3><div>In this prospective, institutional review board-approved study, both single-shot (SS-DWI) and multi-shot MUSE DWI were performed on patients. MUSE DWI was processed using deep-learning reconstruction (MUSE DL). Quantitative analysis included calculating apparent diffusion coefficients (ADCs), signal-to-noise ratio (SNR) within fibroglandular tissue (FGT), adjacent pectoralis muscle, and breast tumors. The Hausdorff distance (HD) was used as a distortion index to compare breast contours between T2-weighted anatomical images, SS-DWI, and MUSE images. Subjective visual qualitative analysis was performed using Likert scale. Quantitative analyses were assessed using Friedman’s rank-based analysis with Bonferroni correction.</div></div><div><h3>Results</h3><div>Sixty-one female participants (mean age 49.07 years ± 11.0 [standard deviation]; age range 23–75 years) with 65 breast lesions were included in this study. All data were acquired using a 3 T MRI scanner. The MUSE DL yielded significant improvement in image quality compared with non-DL MUSE in both 2-shot and 4-shot settings (SNR enhancement FGT 2-shot DL 207.8 % [125.5–309.3],4- shot DL 175.1 % [102.2–223.5]). No significant difference was observed in the ADC between MUSE, MUSE DL, and SS-DWI in both benign (<em>P</em> = 0.154) and malignant tumors (<em>P</em> = 0.167). There was significantly less distortion in the 2- and 4-shot MUSE DL images (HD 3.11 mm, 2.58 mm) than in the SS-DWI images (4.15 mm, <em>P</em> &lt; 0.001).</div></div><div><h3>Conclusions</h3><div>MUSE DL enhances SNR, minimizes image distortion, and preserves lesion diagnosis accuracy and ADC values.</div></div>","PeriodicalId":12063,"journal":{"name":"European Journal of Radiology","volume":"193 ","pages":"Article 112419"},"PeriodicalIF":3.3,"publicationDate":"2025-09-15","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145097678","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Fully automatic bile duct segmentation in magnetic resonance cholangiopancreatography for biliary surgery planning using deep learning","authors":"Haisu Tao ,&nbsp;Junfeng Wang ,&nbsp;Kangwei Guo ,&nbsp;Wang Luo ,&nbsp;Xiaojun Zeng ,&nbsp;Mingjun Lu ,&nbsp;Jinyu Lin ,&nbsp;Baihong Li ,&nbsp;Yinling Qian ,&nbsp;Jian Yang","doi":"10.1016/j.ejrad.2025.112415","DOIUrl":"10.1016/j.ejrad.2025.112415","url":null,"abstract":"<div><h3>Objectives</h3><div>To automatically and accurately perform three-dimensional reconstruction of dilated and non-dilated bile ducts based on magnetic resonance cholangiopancreatography (MRCP) data, assisting in the formulation of optimal surgical plans and guiding precise bile duct surgery.</div></div><div><h3>Methods</h3><div>A total of 249 consecutive patients who underwent standardized 3D-MRCP scans were randomly divided into a training cohort (n = 208) and a testing cohort (n = 41). Ground truth segmentation was manually delineated by two hepatobiliary surgeons or radiologists following industry certification procedures and reviewed by two expert-level physicians for biliary surgery planning. The deep learning semantic segmentation model was constructed using the nnU-Net framework. Model performance was assessed by comparing model predictions with ground truth segmentation as well as real surgical scenarios. The generalization of the model was tested on a dataset of 10 3D-MRCP scans from other centers, with ground truth segmentation of biliary structures.</div></div><div><h3>Results</h3><div>The evaluation was performed on 41 internal test sets and 10 external test sets, with mean Dice Similarity Coefficient (DSC) values of respectively 0.9403 and 0.9070. The correlation coefficient between the 3D model based on automatic segmentation predictions and the ground truth results exceeded 0.95. The 95 % limits of agreement (LoA) for biliary tract length ranged from −4.456 to 4.781, and for biliary tract volume ranged from −3.404 to 3.650 ml. Furthermore, the intraoperative Indocyanine green (ICG) fluorescence imaging and operation situation validated that this model can accurately reconstruct biliary landmarks.</div></div><div><h3>Conclusion</h3><div>By leveraging a deep learning algorithmic framework, an AI model can be trained to perform automatic and accurate 3D reconstructions of non-dilated bile ducts, thereby providing guidance for the preoperative planning of complex biliary surgeries.</div></div>","PeriodicalId":12063,"journal":{"name":"European Journal of Radiology","volume":"193 ","pages":"Article 112415"},"PeriodicalIF":3.3,"publicationDate":"2025-09-15","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145091684","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Cerebral collateral cascade associated with infarct growth rate in ischemic stroke patients undergoing endovascular treatment","authors":"Liping Huang ,&nbsp;Hongfa Zhang ,&nbsp;Wenze Li ,&nbsp;Chen Gong ,&nbsp;Shuyu Jiang ,&nbsp;Zhipeng Li ,&nbsp;Jinxian Yuan ,&nbsp;Tao Xu ,&nbsp;Yangmei Chen ,&nbsp;Lina Zhang ,&nbsp;You Wang","doi":"10.1016/j.ejrad.2025.112418","DOIUrl":"10.1016/j.ejrad.2025.112418","url":null,"abstract":"<div><h3>Background</h3><div>Variability in infarct growth rate(IGR) is strongly associated with clinical outcomes in acute ischemic stroke(AIS) patients receiving endovascular treatment(EVT). Recently, the cerebral collateral cascade(CCC) has been shown to be related to imaging and clinical outcomes in AIS. Therefore, we investigated the association between CCC and IGR.</div></div><div><h3>Methods</h3><div>This was a multicenter retrospective study for AIS patients receiving EVT. IGR was calculated as the ischemic core volume on perfusion computed tomography divided by the time from stroke onset to imaging. Cerebral collateral circulation was assessed using the CCC, which integrated arterial collaterals, tissue-level collaterals, and venous outflow. Multivariable regression was applied to determine factors associated with IGR.</div></div><div><h3>Results</h3><div>A total of 321 patients were included. The median ischemic core volume was 12.5 mL, and the median IGR was 3.0 mL/h. Multivariable regression analysis showed that compared with patients in the CCC_– group, those in the CCC_mixed(β = −21.43, 95 % CI −35.08 to −7.79; P = 0.002) and CCC₊(β = −34.15, 95 % CI −49.90 to −18.40; P &lt; 0.001) groups had significantly lower IGR. Similar associations were observed in both the elderly and late-window cohorts. Additionally, the NIHSS score at admission and ICA occlusion were significantly associated with IGR (P &lt; 0.05). Furthermore, the diagnostic performance of CCC was significantly better than that of Tan scale (P = 0.001), VO (P &lt; 0.001), or HIR (P = 0.001) as assessed by DeLong’s test.</div></div><div><h3>Conclusion</h3><div>CCC profiles were strongly associated with infarct progression in AIS patients undergoing EVT. These findings might help us understand the rule of occurrence and development of infarction progression.</div></div>","PeriodicalId":12063,"journal":{"name":"European Journal of Radiology","volume":"193 ","pages":"Article 112418"},"PeriodicalIF":3.3,"publicationDate":"2025-09-14","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145085648","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Chat GPT-4 shows high agreement in MRI protocol selection compared to board-certified neuroradiologists","authors":"Zeynep Bendella ,&nbsp;Barbara Daria Wichtmann ,&nbsp;Ralf Clauberg ,&nbsp;Vera C. Keil ,&nbsp;Nils C. Lehnen ,&nbsp;Robert Haase ,&nbsp;Laura C. Sáez ,&nbsp;Isabella C. Wiest ,&nbsp;Jakob Nikolas Kather ,&nbsp;Christoph Endler ,&nbsp;Alexander Radbruch ,&nbsp;Daniel Paech ,&nbsp;Katerina Deike","doi":"10.1016/j.ejrad.2025.112416","DOIUrl":"10.1016/j.ejrad.2025.112416","url":null,"abstract":"<div><h3>Objectives</h3><div>The aim of this study was to determine whether ChatGPT-4 can correctly suggest MRI protocols and additional MRI sequences based on real-world Radiology Request Forms (RRFs) as well as to investigate the ability of ChatGPT-4 to suggest time saving protocols.</div></div><div><h3>Material &amp; methods</h3><div>Retrospectively, 1,001 RRFs of our Department of Neuroradiology (in-house dataset), 200 RRFs of an independent Department of General Radiology (independent dataset) and 300 RRFs from an external, foreign Department of Neuroradiology (external dataset) were included. Patients’ age, sex, and clinical information were extracted from the RRFs and used to prompt ChatGPT- 4 to choose an adequate MRI protocol from predefined institutional lists. Four independent raters then assessed its performance. Additionally, ChatGPT-4 was tasked with creating case-specific protocols aimed at saving time.</div></div><div><h3>Results</h3><div>Two and 7 of 1,001 protocol suggestions of ChatGPT-4 were rated “unacceptable” in the in-house dataset for reader 1 and 2, respectively. No protocol suggestions were rated “unacceptable” in both the independent and external dataset. When assessing the inter-reader agreement, Coheńs weighted ĸ ranged from 0.88 to 0.98 (each p &lt; 0.001).</div><div>ChatGPT-4′s freely composed protocols were approved in 766/1,001 (76.5 %) and 140/300 (46.67 %) cases of the in-house and external dataset with mean time savings (standard deviation) of 3:51 (minutes:seconds) (±2:40) minutes and 2:59 (±3:42) minutes per adopted in-house and external MRI protocol.</div></div><div><h3>Conclusion</h3><div>ChatGPT-4 demonstrated a very high agreement with board-certified (neuro-)radiologists in selecting MRI protocols and was able to suggest approved time saving protocols from the set of available sequences.</div></div>","PeriodicalId":12063,"journal":{"name":"European Journal of Radiology","volume":"193 ","pages":"Article 112416"},"PeriodicalIF":3.3,"publicationDate":"2025-09-13","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145079386","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Large language models in radiology reporting: Bridging semantics, education, and safety","authors":"Eren Çamur ,&nbsp;Turay Cesur ,&nbsp;Yasin Celal Güneş","doi":"10.1016/j.ejrad.2025.112417","DOIUrl":"10.1016/j.ejrad.2025.112417","url":null,"abstract":"","PeriodicalId":12063,"journal":{"name":"European Journal of Radiology","volume":"192 ","pages":"Article 112417"},"PeriodicalIF":3.3,"publicationDate":"2025-09-11","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145060033","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"AI radiomics predicts spatial glioma recurrence on preoperative MRI: a systematic review","authors":"Sebastiaan Bijlsma ,&nbsp;Thomas Maal ,&nbsp;Christian Rubbert ,&nbsp;Manoj Mannil ,&nbsp;Anton Meijer ,&nbsp;Anja van der Kolk ,&nbsp;Guido de Jong ,&nbsp;Dylan Henssen","doi":"10.1016/j.ejrad.2025.112412","DOIUrl":"10.1016/j.ejrad.2025.112412","url":null,"abstract":"<div><h3>Background</h3><div>Adult-type diffuse gliomas are highly infiltrative primary brain tumors in which, after a combination of surgery and chemoradiation therapy, tumor recurrence is inevitable. Artificial Intelligence (AI) models have been found capable to predict local and distant tumor recurrence at baseline, with the potential to guide surgical margins and enable focal dose escalation in radiotherapy. This study systematically reviews the literature on the performance of AI models in predicting local or distant tumor recurrence in glioma patients using preoperative MRI data.</div></div><div><h3>Methods</h3><div>A systematic literature search was conducted across PubMed, EMBASE, and the Cochrane Library. Studies evaluating AI-based models for spatial recurrence prediction in gliomas using preoperative MRI were included. Study quality and methodological rigor were assessed using the PROBAST + AI tool.</div></div><div><h3>Findings</h3><div>Eight studies, comprising 1004 high grade glioma patients, were included. A variety of machine learning and deep learning model architectures (e.g., Random Forest classifiers, Support Vector Machines and custom Convolutional Neural Networks) were employed. Input data were a heterogeneous combination of conventional MRI (e.g., T1CE, FLAIR) and more advanced imaging modalities (e.g., diffusion-weighted imaging). Considerable variability was reported with regard to sensitivity and specificity rates (ranging between 40 %-97 % and 29 %-98 %, respectively) for predicting tumor recurrence. The odds ratios for predicting regions of tumor recurrence, however, were generally high (ranging between 8.13–19.48). External validation was performed in 4 studies with one study using a multicenter cohort of 6 different institutions, demonstrating high generalizability in predictive performance. Risk of bias analysis was performed using the recently published PROBAST + AI tool and revealed generally low to unclear concern for risk of bias and low concern for applicability.</div></div><div><h3>Interpretation</h3><div>AI models have been shown capable of predicting local and distant tumor recurrence in glioma patients from baseline MRI data. While the high odds ratios reported from the multicenter study are encouraging, the evidence comes mainly from small, single-center, retrospective cohorts, so larger prospective multicenter studies are needed before clinical adoption.</div></div>","PeriodicalId":12063,"journal":{"name":"European Journal of Radiology","volume":"193 ","pages":"Article 112412"},"PeriodicalIF":3.3,"publicationDate":"2025-09-11","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145110060","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Synthetic MRI as a T2WI alternative in bp-MRI: comparable image quality and improved PI-RADS performance","authors":"You Yun ,&nbsp;Jinxia Guo ,&nbsp;Nannan Shao ,&nbsp;Xiaoxian Zhang ,&nbsp;Lifeng Wang ,&nbsp;Wentao Liu ,&nbsp;Yue Wu ,&nbsp;Chunmiao Xu ,&nbsp;Renzhi Zhang ,&nbsp;Xuejun Chen","doi":"10.1016/j.ejrad.2025.112414","DOIUrl":"10.1016/j.ejrad.2025.112414","url":null,"abstract":"<div><h3>Objectives</h3><div>Prostate cancer significantly impacts men’s health, highlighting the necessity for precise diagnosis. This study evaluates combining Magnetic Resonance Image Compilation (MAGiC) parameters with Prostate Imaging Reporting and Data System (PI-RADS) scores to enhance diagnostic accuracy among radiologists with varied experience.</div></div><div><h3>Methods</h3><div>In this retrospective study, 174 patients with suspected prostate cancer were recruited from February 2023 to May 2024. Synthetic MRI-derived T1, T2, and proton density (PD) maps were generated, and synthetic T2-weighted imaging (T2WI) was reconstructed. Two radiologists of varying experience independently assessed lesions using PI-RADS based on both synthetic and conventional T2WI. Image quality was evaluated using the prostate imaging quality (PI-QUAL) scoring v2 system, and diagnostic performance was analyzed using receiver operating characteristic (ROC) curve analysis.</div></div><div><h3>Results</h3><div>Synthetic T2WI exhibited comparable image quality to conventional T2WI (P = 0.065). After excluding low-quality images, 99 lesions were analyzed. In the peripheral zone, higher T1 values were significantly linked to non-cancerous lesions (R1: OR = 0.989, P = 0.009; R2: OR = 0.990, P = 0.004). The integration of T1 values with PI-RADS scores improved diagnostic performance, achieving area under the curve (AUC) values of 0.960 for R1 and 0.944 for R2.</div></div><div><h3>Conclusion</h3><div>The integration of MAGiC parameters, particularly T1 values, with PI-RADS scores significantly enhances diagnostic accuracy for clinically significant prostate cancer, especially benefitting less experienced radiologists. Additionally, synthetic T2WI demonstrates comparable image quality to conventional T2WI, supporting the clinical implementation of MAGiC parameters in prostate MRI assessments.</div></div>","PeriodicalId":12063,"journal":{"name":"European Journal of Radiology","volume":"193 ","pages":"Article 112414"},"PeriodicalIF":3.3,"publicationDate":"2025-09-10","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145085632","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Quantification of tumor heterogeneity based on fractal dimension for predicting the response to neoadjuvant chemotherapy in triple-negative breast cancer","authors":"Jiamin Guo ,&nbsp;Ying Liu ,&nbsp;Wei Ren ,&nbsp;Yichen Zheng ,&nbsp;Tonghui Ren ,&nbsp;Ji Ma ,&nbsp;Shuang Zhao","doi":"10.1016/j.ejrad.2025.112413","DOIUrl":"10.1016/j.ejrad.2025.112413","url":null,"abstract":"<div><h3>Background</h3><div>Triple-negative breast cancer (TNBC) exhibits high heterogeneity, leading to variable responses to neoadjuvant chemotherapy (NAC) among patients. Noninvasive quantification of intratumoral heterogeneity (ITH) may be valuable in predicting treatment response. This study aims to investigate whether fractal dimension (FD) based on pre-treatment contrast-enhanced magnetic resonance imaging (MRI), combined with clinicopathological data, can predict NAC response in TNBC patients.</div></div><div><h3>Methods</h3><div>We retrospectively collected clinicopathological data and pre-treatment contrast-enhanced breast MRI scans of TNBC patients who underwent NAC followed by surgery at our institution from January 2012 to September 2021. Patients were classified into a pathological complete response (pCR) group and a non-pCR group based on postoperative pathological specimens. Regions of interest (ROIs) were delineated on enhanced MRI lesions, and FD analysis was performed using the box-counting method to assess ITH. Univariate and multivariate regression analyses were used to identify variables associated with pCR. A predictive model incorporating relevant clinicopathological variables and FD was constructed, and model performance was evaluated using the area under the ROI curve (AUC).</div></div><div><h3>Results</h3><div>Among 122 evaluated TNBC patients, 28.7 % (n = 35/122) achieved pCR. Multivariate regression analysis identified tumor T stage (OR = 1.595, 95 %CI:1.032–2.467, p = 0.036), changes in Ki-67 before and after NAC (OR = 0.099, 95 %CI:0.044–0.227, p &lt; 0.001), and pre-treatment FD (adjusted OR = 18.032, 95 %CI:0.749–434.041, p = 0.075) as independent predictors of pCR. In the test set, the AUC of the clinical model based on T stage and Ki-67 changes was 0.846, while the FD model achieved an AUC of 0.867. The combined model, which integrated clinical data with FD, further improved predictive performance, reaching an AUC of 0.895.</div></div><div><h3>Conclusion</h3><div>FD derived from pre-treatment MRI can quantify ITH and serves as a noninvasive imaging biomarker. The combined model integrating FD with clinical data further enhances predictive accuracy.</div></div>","PeriodicalId":12063,"journal":{"name":"European Journal of Radiology","volume":"192 ","pages":"Article 112413"},"PeriodicalIF":3.3,"publicationDate":"2025-09-08","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145043983","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}